lvmcache: Remove VG metadata caches (text, cfg, volume_group) for now.

The toollib code for processing PVs given on commandline tried to be smart and
use pv_read, avoiding real metadata scanning. The premise of the optimisation
was that if there is an active MDA on the PV, it can be reliably used to
establish VG membership. This assumption is however wrong, since it's entirely
possible that this particular MDA is simply out of date, and the PV is not
actually in any VG. While this is a subtle and rare bug, it's still a bug. One
way to trigger it would be to hotplug an old disk which at some point was a PV
in an existing VG, having a seemingly valid MDA but with an old sequence
number. All commands that use normal scanning realize that this PV does not
belong into the VG anymore, but anything based on pv_read alone will believe it
does.

.gitignore

@@ -0,0 +1,27 @@
+*.5
+*.8
+*.a
+*.d
+*.o
+*.pc
+*.pot
+*.so
+*.so.*
+*.swp
+*~
+
+.export.sym
+.exported_symbols_generated
+.gdb_history
+
+Makefile
+make.tmpl
+
+configure.h
+version.h
+
+/autom4te.cache/
+/config.log
+/config.status
+/cscope.out
+/tmp/

COPYING

@@ -0,0 +1,340 @@
+		    GNU GENERAL PUBLIC LICENSE
+		       Version 2, June 1991
+
+ Copyright (C) 1989, 1991 Free Software Foundation, Inc.
+     59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ Everyone is permitted to copy and distribute verbatim copies
+ of this license document, but changing it is not allowed.
+
+			    Preamble
+
+  The licenses for most software are designed to take away your
+freedom to share and change it.  By contrast, the GNU General Public
+License is intended to guarantee your freedom to share and change free
+software--to make sure the software is free for all its users.  This
+General Public License applies to most of the Free Software
+Foundation's software and to any other program whose authors commit to
+using it.  (Some other Free Software Foundation software is covered by
+the GNU Library General Public License instead.)  You can apply it to
+your programs, too.
+
+  When we speak of free software, we are referring to freedom, not
+price.  Our General Public Licenses are designed to make sure that you
+have the freedom to distribute copies of free software (and charge for
+this service if you wish), that you receive source code or can get it
+if you want it, that you can change the software or use pieces of it
+in new free programs; and that you know you can do these things.
+
+  To protect your rights, we need to make restrictions that forbid
+anyone to deny you these rights or to ask you to surrender the rights.
+These restrictions translate to certain responsibilities for you if you
+distribute copies of the software, or if you modify it.
+
+  For example, if you distribute copies of such a program, whether
+gratis or for a fee, you must give the recipients all the rights that
+you have.  You must make sure that they, too, receive or can get the
+source code.  And you must show them these terms so they know their
+rights.
+
+  We protect your rights with two steps: (1) copyright the software, and
+(2) offer you this license which gives you legal permission to copy,
+distribute and/or modify the software.
+
+  Also, for each author's protection and ours, we want to make certain
+that everyone understands that there is no warranty for this free
+software.  If the software is modified by someone else and passed on, we
+want its recipients to know that what they have is not the original, so
+that any problems introduced by others will not reflect on the original
+authors' reputations.
+
+  Finally, any free program is threatened constantly by software
+patents.  We wish to avoid the danger that redistributors of a free
+program will individually obtain patent licenses, in effect making the
+program proprietary.  To prevent this, we have made it clear that any
+patent must be licensed for everyone's free use or not licensed at all.
+
+  The precise terms and conditions for copying, distribution and
+modification follow.
+
+		    GNU GENERAL PUBLIC LICENSE
+   TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
+
+  0. This License applies to any program or other work which contains
+a notice placed by the copyright holder saying it may be distributed
+under the terms of this General Public License.  The "Program", below,
+refers to any such program or work, and a "work based on the Program"
+means either the Program or any derivative work under copyright law:
+that is to say, a work containing the Program or a portion of it,
+either verbatim or with modifications and/or translated into another
+language.  (Hereinafter, translation is included without limitation in
+the term "modification".)  Each licensee is addressed as "you".
+
+Activities other than copying, distribution and modification are not
+covered by this License; they are outside its scope.  The act of
+running the Program is not restricted, and the output from the Program
+is covered only if its contents constitute a work based on the
+Program (independent of having been made by running the Program).
+Whether that is true depends on what the Program does.
+
+  1. You may copy and distribute verbatim copies of the Program's
+source code as you receive it, in any medium, provided that you
+conspicuously and appropriately publish on each copy an appropriate
+copyright notice and disclaimer of warranty; keep intact all the
+notices that refer to this License and to the absence of any warranty;
+and give any other recipients of the Program a copy of this License
+along with the Program.
+
+You may charge a fee for the physical act of transferring a copy, and
+you may at your option offer warranty protection in exchange for a fee.
+
+  2. You may modify your copy or copies of the Program or any portion
+of it, thus forming a work based on the Program, and copy and
+distribute such modifications or work under the terms of Section 1
+above, provided that you also meet all of these conditions:
+
+    a) You must cause the modified files to carry prominent notices
+    stating that you changed the files and the date of any change.
+
+    b) You must cause any work that you distribute or publish, that in
+    whole or in part contains or is derived from the Program or any
+    part thereof, to be licensed as a whole at no charge to all third
+    parties under the terms of this License.
+
+    c) If the modified program normally reads commands interactively
+    when run, you must cause it, when started running for such
+    interactive use in the most ordinary way, to print or display an
+    announcement including an appropriate copyright notice and a
+    notice that there is no warranty (or else, saying that you provide
+    a warranty) and that users may redistribute the program under
+    these conditions, and telling the user how to view a copy of this
+    License.  (Exception: if the Program itself is interactive but
+    does not normally print such an announcement, your work based on
+    the Program is not required to print an announcement.)
+
+These requirements apply to the modified work as a whole.  If
+identifiable sections of that work are not derived from the Program,
+and can be reasonably considered independent and separate works in
+themselves, then this License, and its terms, do not apply to those
+sections when you distribute them as separate works.  But when you
+distribute the same sections as part of a whole which is a work based
+on the Program, the distribution of the whole must be on the terms of
+this License, whose permissions for other licensees extend to the
+entire whole, and thus to each and every part regardless of who wrote it.
+
+Thus, it is not the intent of this section to claim rights or contest
+your rights to work written entirely by you; rather, the intent is to
+exercise the right to control the distribution of derivative or
+collective works based on the Program.
+
+In addition, mere aggregation of another work not based on the Program
+with the Program (or with a work based on the Program) on a volume of
+a storage or distribution medium does not bring the other work under
+the scope of this License.
+
+  3. You may copy and distribute the Program (or a work based on it,
+under Section 2) in object code or executable form under the terms of
+Sections 1 and 2 above provided that you also do one of the following:
+
+    a) Accompany it with the complete corresponding machine-readable
+    source code, which must be distributed under the terms of Sections
+    1 and 2 above on a medium customarily used for software interchange; or,
+
+    b) Accompany it with a written offer, valid for at least three
+    years, to give any third party, for a charge no more than your
+    cost of physically performing source distribution, a complete
+    machine-readable copy of the corresponding source code, to be
+    distributed under the terms of Sections 1 and 2 above on a medium
+    customarily used for software interchange; or,
+
+    c) Accompany it with the information you received as to the offer
+    to distribute corresponding source code.  (This alternative is
+    allowed only for noncommercial distribution and only if you
+    received the program in object code or executable form with such
+    an offer, in accord with Subsection b above.)
+
+The source code for a work means the preferred form of the work for
+making modifications to it.  For an executable work, complete source
+code means all the source code for all modules it contains, plus any
+associated interface definition files, plus the scripts used to
+control compilation and installation of the executable.  However, as a
+special exception, the source code distributed need not include
+anything that is normally distributed (in either source or binary
+form) with the major components (compiler, kernel, and so on) of the
+operating system on which the executable runs, unless that component
+itself accompanies the executable.
+
+If distribution of executable or object code is made by offering
+access to copy from a designated place, then offering equivalent
+access to copy the source code from the same place counts as
+distribution of the source code, even though third parties are not
+compelled to copy the source along with the object code.
+
+  4. You may not copy, modify, sublicense, or distribute the Program
+except as expressly provided under this License.  Any attempt
+otherwise to copy, modify, sublicense or distribute the Program is
+void, and will automatically terminate your rights under this License.
+However, parties who have received copies, or rights, from you under
+this License will not have their licenses terminated so long as such
+parties remain in full compliance.
+
+  5. You are not required to accept this License, since you have not
+signed it.  However, nothing else grants you permission to modify or
+distribute the Program or its derivative works.  These actions are
+prohibited by law if you do not accept this License.  Therefore, by
+modifying or distributing the Program (or any work based on the
+Program), you indicate your acceptance of this License to do so, and
+all its terms and conditions for copying, distributing or modifying
+the Program or works based on it.
+
+  6. Each time you redistribute the Program (or any work based on the
+Program), the recipient automatically receives a license from the
+original licensor to copy, distribute or modify the Program subject to
+these terms and conditions.  You may not impose any further
+restrictions on the recipients' exercise of the rights granted herein.
+You are not responsible for enforcing compliance by third parties to
+this License.
+
+  7. If, as a consequence of a court judgment or allegation of patent
+infringement or for any other reason (not limited to patent issues),
+conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License.  If you cannot
+distribute so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you
+may not distribute the Program at all.  For example, if a patent
+license would not permit royalty-free redistribution of the Program by
+all those who receive copies directly or indirectly through you, then
+the only way you could satisfy both it and this License would be to
+refrain entirely from distribution of the Program.
+
+If any portion of this section is held invalid or unenforceable under
+any particular circumstance, the balance of the section is intended to
+apply and the section as a whole is intended to apply in other
+circumstances.
+
+It is not the purpose of this section to induce you to infringe any
+patents or other property right claims or to contest validity of any
+such claims; this section has the sole purpose of protecting the
+integrity of the free software distribution system, which is
+implemented by public license practices.  Many people have made
+generous contributions to the wide range of software distributed
+through that system in reliance on consistent application of that
+system; it is up to the author/donor to decide if he or she is willing
+to distribute software through any other system and a licensee cannot
+impose that choice.
+
+This section is intended to make thoroughly clear what is believed to
+be a consequence of the rest of this License.
+
+  8. If the distribution and/or use of the Program is restricted in
+certain countries either by patents or by copyrighted interfaces, the
+original copyright holder who places the Program under this License
+may add an explicit geographical distribution limitation excluding
+those countries, so that distribution is permitted only in or among
+countries not thus excluded.  In such case, this License incorporates
+the limitation as if written in the body of this License.
+
+  9. The Free Software Foundation may publish revised and/or new versions
+of the General Public License from time to time.  Such new versions will
+be similar in spirit to the present version, but may differ in detail to
+address new problems or concerns.
+
+Each version is given a distinguishing version number.  If the Program
+specifies a version number of this License which applies to it and "any
+later version", you have the option of following the terms and conditions
+either of that version or of any later version published by the Free
+Software Foundation.  If the Program does not specify a version number of
+this License, you may choose any version ever published by the Free Software
+Foundation.
+
+  10. If you wish to incorporate parts of the Program into other free
+programs whose distribution conditions are different, write to the author
+to ask for permission.  For software which is copyrighted by the Free
+Software Foundation, write to the Free Software Foundation; we sometimes
+make exceptions for this.  Our decision will be guided by the two goals
+of preserving the free status of all derivatives of our free software and
+of promoting the sharing and reuse of software generally.
+
+			    NO WARRANTY
+
+  11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
+FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW.  EXCEPT WHEN
+OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
+PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
+OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
+MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.  THE ENTIRE RISK AS
+TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU.  SHOULD THE
+PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
+REPAIR OR CORRECTION.
+
+  12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
+WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
+REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
+INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING
+OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED
+TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY
+YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
+PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE
+POSSIBILITY OF SUCH DAMAGES.
+
+		     END OF TERMS AND CONDITIONS
+
+	    How to Apply These Terms to Your New Programs
+
+  If you develop a new program, and you want it to be of the greatest
+possible use to the public, the best way to achieve this is to make it
+free software which everyone can redistribute and change under these terms.
+
+  To do so, attach the following notices to the program.  It is safest
+to attach them to the start of each source file to most effectively
+convey the exclusion of warranty; and each file should have at least
+the "copyright" line and a pointer to where the full notice is found.
+
+    <one line to give the program's name and a brief idea of what it does.>
+    Copyright (C) <year>  <name of author>
+
+    This program is free software; you can redistribute it and/or modify
+    it under the terms of the GNU General Public License as published by
+    the Free Software Foundation; either version 2 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU General Public License for more details.
+
+    You should have received a copy of the GNU General Public License
+    along with this program; if not, write to the Free Software
+    Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+
+Also add information on how to contact you by electronic and paper mail.
+
+If the program is interactive, make it output a short notice like this
+when it starts in an interactive mode:
+
+    Gnomovision version 69, Copyright (C) year  name of author
+    Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
+    This is free software, and you are welcome to redistribute it
+    under certain conditions; type `show c' for details.
+
+The hypothetical commands `show w' and `show c' should show the appropriate
+parts of the General Public License.  Of course, the commands you use may
+be called something other than `show w' and `show c'; they could even be
+mouse-clicks or menu items--whatever suits your program.
+
+You should also get your employer (if you work as a programmer) or your
+school, if any, to sign a "copyright disclaimer" for the program, if
+necessary.  Here is a sample; alter the names:
+
+  Yoyodyne, Inc., hereby disclaims all copyright interest in the program
+  `Gnomovision' (which makes passes at compilers) written by James Hacker.
+
+  <signature of Ty Coon>, 1 April 1989
+  Ty Coon, President of Vice
+
+This General Public License does not permit incorporating your program into
+proprietary programs.  If your program is a subroutine library, you may
+consider it more useful to permit linking proprietary applications with the
+library.  If this is what you want to do, use the GNU Library General
+Public License instead of this License.

COPYING.LIB

@@ -0,0 +1,504 @@
+		  GNU LESSER GENERAL PUBLIC LICENSE
+		       Version 2.1, February 1999
+
+ Copyright (C) 1991, 1999 Free Software Foundation, Inc.
+ 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
+ Everyone is permitted to copy and distribute verbatim copies
+ of this license document, but changing it is not allowed.
+
+[This is the first released version of the Lesser GPL.  It also counts
+ as the successor of the GNU Library Public License, version 2, hence
+ the version number 2.1.]
+
+			    Preamble
+
+  The licenses for most software are designed to take away your
+freedom to share and change it.  By contrast, the GNU General Public
+Licenses are intended to guarantee your freedom to share and change
+free software--to make sure the software is free for all its users.
+
+  This license, the Lesser General Public License, applies to some
+specially designated software packages--typically libraries--of the
+Free Software Foundation and other authors who decide to use it.  You
+can use it too, but we suggest you first think carefully about whether
+this license or the ordinary General Public License is the better
+strategy to use in any particular case, based on the explanations below.
+
+  When we speak of free software, we are referring to freedom of use,
+not price.  Our General Public Licenses are designed to make sure that
+you have the freedom to distribute copies of free software (and charge
+for this service if you wish); that you receive source code or can get
+it if you want it; that you can change the software and use pieces of
+it in new free programs; and that you are informed that you can do
+these things.
+
+  To protect your rights, we need to make restrictions that forbid
+distributors to deny you these rights or to ask you to surrender these
+rights.  These restrictions translate to certain responsibilities for
+you if you distribute copies of the library or if you modify it.
+
+  For example, if you distribute copies of the library, whether gratis
+or for a fee, you must give the recipients all the rights that we gave
+you.  You must make sure that they, too, receive or can get the source
+code.  If you link other code with the library, you must provide
+complete object files to the recipients, so that they can relink them
+with the library after making changes to the library and recompiling
+it.  And you must show them these terms so they know their rights.
+
+  We protect your rights with a two-step method: (1) we copyright the
+library, and (2) we offer you this license, which gives you legal
+permission to copy, distribute and/or modify the library.
+
+  To protect each distributor, we want to make it very clear that
+there is no warranty for the free library.  Also, if the library is
+modified by someone else and passed on, the recipients should know
+that what they have is not the original version, so that the original
+author's reputation will not be affected by problems that might be
+introduced by others.
+
+  Finally, software patents pose a constant threat to the existence of
+any free program.  We wish to make sure that a company cannot
+effectively restrict the users of a free program by obtaining a
+restrictive license from a patent holder.  Therefore, we insist that
+any patent license obtained for a version of the library must be
+consistent with the full freedom of use specified in this license.
+
+  Most GNU software, including some libraries, is covered by the
+ordinary GNU General Public License.  This license, the GNU Lesser
+General Public License, applies to certain designated libraries, and
+is quite different from the ordinary General Public License.  We use
+this license for certain libraries in order to permit linking those
+libraries into non-free programs.
+
+  When a program is linked with a library, whether statically or using
+a shared library, the combination of the two is legally speaking a
+combined work, a derivative of the original library.  The ordinary
+General Public License therefore permits such linking only if the
+entire combination fits its criteria of freedom.  The Lesser General
+Public License permits more lax criteria for linking other code with
+the library.
+
+  We call this license the "Lesser" General Public License because it
+does Less to protect the user's freedom than the ordinary General
+Public License.  It also provides other free software developers Less
+of an advantage over competing non-free programs.  These disadvantages
+are the reason we use the ordinary General Public License for many
+libraries.  However, the Lesser license provides advantages in certain
+special circumstances.
+
+  For example, on rare occasions, there may be a special need to
+encourage the widest possible use of a certain library, so that it becomes
+a de-facto standard.  To achieve this, non-free programs must be
+allowed to use the library.  A more frequent case is that a free
+library does the same job as widely used non-free libraries.  In this
+case, there is little to gain by limiting the free library to free
+software only, so we use the Lesser General Public License.
+
+  In other cases, permission to use a particular library in non-free
+programs enables a greater number of people to use a large body of
+free software.  For example, permission to use the GNU C Library in
+non-free programs enables many more people to use the whole GNU
+operating system, as well as its variant, the GNU/Linux operating
+system.
+
+  Although the Lesser General Public License is Less protective of the
+users' freedom, it does ensure that the user of a program that is
+linked with the Library has the freedom and the wherewithal to run
+that program using a modified version of the Library.
+
+  The precise terms and conditions for copying, distribution and
+modification follow.  Pay close attention to the difference between a
+"work based on the library" and a "work that uses the library".  The
+former contains code derived from the library, whereas the latter must
+be combined with the library in order to run.
+
+		  GNU LESSER GENERAL PUBLIC LICENSE
+   TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
+
+  0. This License Agreement applies to any software library or other
+program which contains a notice placed by the copyright holder or
+other authorized party saying it may be distributed under the terms of
+this Lesser General Public License (also called "this License").
+Each licensee is addressed as "you".
+
+  A "library" means a collection of software functions and/or data
+prepared so as to be conveniently linked with application programs
+(which use some of those functions and data) to form executables.
+
+  The "Library", below, refers to any such software library or work
+which has been distributed under these terms.  A "work based on the
+Library" means either the Library or any derivative work under
+copyright law: that is to say, a work containing the Library or a
+portion of it, either verbatim or with modifications and/or translated
+straightforwardly into another language.  (Hereinafter, translation is
+included without limitation in the term "modification".)
+
+  "Source code" for a work means the preferred form of the work for
+making modifications to it.  For a library, complete source code means
+all the source code for all modules it contains, plus any associated
+interface definition files, plus the scripts used to control compilation
+and installation of the library.
+
+  Activities other than copying, distribution and modification are not
+covered by this License; they are outside its scope.  The act of
+running a program using the Library is not restricted, and output from
+such a program is covered only if its contents constitute a work based
+on the Library (independent of the use of the Library in a tool for
+writing it).  Whether that is true depends on what the Library does
+and what the program that uses the Library does.
+  
+  1. You may copy and distribute verbatim copies of the Library's
+complete source code as you receive it, in any medium, provided that
+you conspicuously and appropriately publish on each copy an
+appropriate copyright notice and disclaimer of warranty; keep intact
+all the notices that refer to this License and to the absence of any
+warranty; and distribute a copy of this License along with the
+Library.
+
+  You may charge a fee for the physical act of transferring a copy,
+and you may at your option offer warranty protection in exchange for a
+fee.
+
+  2. You may modify your copy or copies of the Library or any portion
+of it, thus forming a work based on the Library, and copy and
+distribute such modifications or work under the terms of Section 1
+above, provided that you also meet all of these conditions:
+
+    a) The modified work must itself be a software library.
+
+    b) You must cause the files modified to carry prominent notices
+    stating that you changed the files and the date of any change.
+
+    c) You must cause the whole of the work to be licensed at no
+    charge to all third parties under the terms of this License.
+
+    d) If a facility in the modified Library refers to a function or a
+    table of data to be supplied by an application program that uses
+    the facility, other than as an argument passed when the facility
+    is invoked, then you must make a good faith effort to ensure that,
+    in the event an application does not supply such function or
+    table, the facility still operates, and performs whatever part of
+    its purpose remains meaningful.
+
+    (For example, a function in a library to compute square roots has
+    a purpose that is entirely well-defined independent of the
+    application.  Therefore, Subsection 2d requires that any
+    application-supplied function or table used by this function must
+    be optional: if the application does not supply it, the square
+    root function must still compute square roots.)
+
+These requirements apply to the modified work as a whole.  If
+identifiable sections of that work are not derived from the Library,
+and can be reasonably considered independent and separate works in
+themselves, then this License, and its terms, do not apply to those
+sections when you distribute them as separate works.  But when you
+distribute the same sections as part of a whole which is a work based
+on the Library, the distribution of the whole must be on the terms of
+this License, whose permissions for other licensees extend to the
+entire whole, and thus to each and every part regardless of who wrote
+it.
+
+Thus, it is not the intent of this section to claim rights or contest
+your rights to work written entirely by you; rather, the intent is to
+exercise the right to control the distribution of derivative or
+collective works based on the Library.
+
+In addition, mere aggregation of another work not based on the Library
+with the Library (or with a work based on the Library) on a volume of
+a storage or distribution medium does not bring the other work under
+the scope of this License.
+
+  3. You may opt to apply the terms of the ordinary GNU General Public
+License instead of this License to a given copy of the Library.  To do
+this, you must alter all the notices that refer to this License, so
+that they refer to the ordinary GNU General Public License, version 2,
+instead of to this License.  (If a newer version than version 2 of the
+ordinary GNU General Public License has appeared, then you can specify
+that version instead if you wish.)  Do not make any other change in
+these notices.
+
+  Once this change is made in a given copy, it is irreversible for
+that copy, so the ordinary GNU General Public License applies to all
+subsequent copies and derivative works made from that copy.
+
+  This option is useful when you wish to copy part of the code of
+the Library into a program that is not a library.
+
+  4. You may copy and distribute the Library (or a portion or
+derivative of it, under Section 2) in object code or executable form
+under the terms of Sections 1 and 2 above provided that you accompany
+it with the complete corresponding machine-readable source code, which
+must be distributed under the terms of Sections 1 and 2 above on a
+medium customarily used for software interchange.
+
+  If distribution of object code is made by offering access to copy
+from a designated place, then offering equivalent access to copy the
+source code from the same place satisfies the requirement to
+distribute the source code, even though third parties are not
+compelled to copy the source along with the object code.
+
+  5. A program that contains no derivative of any portion of the
+Library, but is designed to work with the Library by being compiled or
+linked with it, is called a "work that uses the Library".  Such a
+work, in isolation, is not a derivative work of the Library, and
+therefore falls outside the scope of this License.
+
+  However, linking a "work that uses the Library" with the Library
+creates an executable that is a derivative of the Library (because it
+contains portions of the Library), rather than a "work that uses the
+library".  The executable is therefore covered by this License.
+Section 6 states terms for distribution of such executables.
+
+  When a "work that uses the Library" uses material from a header file
+that is part of the Library, the object code for the work may be a
+derivative work of the Library even though the source code is not.
+Whether this is true is especially significant if the work can be
+linked without the Library, or if the work is itself a library.  The
+threshold for this to be true is not precisely defined by law.
+
+  If such an object file uses only numerical parameters, data
+structure layouts and accessors, and small macros and small inline
+functions (ten lines or less in length), then the use of the object
+file is unrestricted, regardless of whether it is legally a derivative
+work.  (Executables containing this object code plus portions of the
+Library will still fall under Section 6.)
+
+  Otherwise, if the work is a derivative of the Library, you may
+distribute the object code for the work under the terms of Section 6.
+Any executables containing that work also fall under Section 6,
+whether or not they are linked directly with the Library itself.
+
+  6. As an exception to the Sections above, you may also combine or
+link a "work that uses the Library" with the Library to produce a
+work containing portions of the Library, and distribute that work
+under terms of your choice, provided that the terms permit
+modification of the work for the customer's own use and reverse
+engineering for debugging such modifications.
+
+  You must give prominent notice with each copy of the work that the
+Library is used in it and that the Library and its use are covered by
+this License.  You must supply a copy of this License.  If the work
+during execution displays copyright notices, you must include the
+copyright notice for the Library among them, as well as a reference
+directing the user to the copy of this License.  Also, you must do one
+of these things:
+
+    a) Accompany the work with the complete corresponding
+    machine-readable source code for the Library including whatever
+    changes were used in the work (which must be distributed under
+    Sections 1 and 2 above); and, if the work is an executable linked
+    with the Library, with the complete machine-readable "work that
+    uses the Library", as object code and/or source code, so that the
+    user can modify the Library and then relink to produce a modified
+    executable containing the modified Library.  (It is understood
+    that the user who changes the contents of definitions files in the
+    Library will not necessarily be able to recompile the application
+    to use the modified definitions.)
+
+    b) Use a suitable shared library mechanism for linking with the
+    Library.  A suitable mechanism is one that (1) uses at run time a
+    copy of the library already present on the user's computer system,
+    rather than copying library functions into the executable, and (2)
+    will operate properly with a modified version of the library, if
+    the user installs one, as long as the modified version is
+    interface-compatible with the version that the work was made with.
+
+    c) Accompany the work with a written offer, valid for at
+    least three years, to give the same user the materials
+    specified in Subsection 6a, above, for a charge no more
+    than the cost of performing this distribution.
+
+    d) If distribution of the work is made by offering access to copy
+    from a designated place, offer equivalent access to copy the above
+    specified materials from the same place.
+
+    e) Verify that the user has already received a copy of these
+    materials or that you have already sent this user a copy.
+
+  For an executable, the required form of the "work that uses the
+Library" must include any data and utility programs needed for
+reproducing the executable from it.  However, as a special exception,
+the materials to be distributed need not include anything that is
+normally distributed (in either source or binary form) with the major
+components (compiler, kernel, and so on) of the operating system on
+which the executable runs, unless that component itself accompanies
+the executable.
+
+  It may happen that this requirement contradicts the license
+restrictions of other proprietary libraries that do not normally
+accompany the operating system.  Such a contradiction means you cannot
+use both them and the Library together in an executable that you
+distribute.
+
+  7. You may place library facilities that are a work based on the
+Library side-by-side in a single library together with other library
+facilities not covered by this License, and distribute such a combined
+library, provided that the separate distribution of the work based on
+the Library and of the other library facilities is otherwise
+permitted, and provided that you do these two things:
+
+    a) Accompany the combined library with a copy of the same work
+    based on the Library, uncombined with any other library
+    facilities.  This must be distributed under the terms of the
+    Sections above.
+
+    b) Give prominent notice with the combined library of the fact
+    that part of it is a work based on the Library, and explaining
+    where to find the accompanying uncombined form of the same work.
+
+  8. You may not copy, modify, sublicense, link with, or distribute
+the Library except as expressly provided under this License.  Any
+attempt otherwise to copy, modify, sublicense, link with, or
+distribute the Library is void, and will automatically terminate your
+rights under this License.  However, parties who have received copies,
+or rights, from you under this License will not have their licenses
+terminated so long as such parties remain in full compliance.
+
+  9. You are not required to accept this License, since you have not
+signed it.  However, nothing else grants you permission to modify or
+distribute the Library or its derivative works.  These actions are
+prohibited by law if you do not accept this License.  Therefore, by
+modifying or distributing the Library (or any work based on the
+Library), you indicate your acceptance of this License to do so, and
+all its terms and conditions for copying, distributing or modifying
+the Library or works based on it.
+
+  10. Each time you redistribute the Library (or any work based on the
+Library), the recipient automatically receives a license from the
+original licensor to copy, distribute, link with or modify the Library
+subject to these terms and conditions.  You may not impose any further
+restrictions on the recipients' exercise of the rights granted herein.
+You are not responsible for enforcing compliance by third parties with
+this License.
+
+  11. If, as a consequence of a court judgment or allegation of patent
+infringement or for any other reason (not limited to patent issues),
+conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License.  If you cannot
+distribute so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you
+may not distribute the Library at all.  For example, if a patent
+license would not permit royalty-free redistribution of the Library by
+all those who receive copies directly or indirectly through you, then
+the only way you could satisfy both it and this License would be to
+refrain entirely from distribution of the Library.
+
+If any portion of this section is held invalid or unenforceable under any
+particular circumstance, the balance of the section is intended to apply,
+and the section as a whole is intended to apply in other circumstances.
+
+It is not the purpose of this section to induce you to infringe any
+patents or other property right claims or to contest validity of any
+such claims; this section has the sole purpose of protecting the
+integrity of the free software distribution system which is
+implemented by public license practices.  Many people have made
+generous contributions to the wide range of software distributed
+through that system in reliance on consistent application of that
+system; it is up to the author/donor to decide if he or she is willing
+to distribute software through any other system and a licensee cannot
+impose that choice.
+
+This section is intended to make thoroughly clear what is believed to
+be a consequence of the rest of this License.
+
+  12. If the distribution and/or use of the Library is restricted in
+certain countries either by patents or by copyrighted interfaces, the
+original copyright holder who places the Library under this License may add
+an explicit geographical distribution limitation excluding those countries,
+so that distribution is permitted only in or among countries not thus
+excluded.  In such case, this License incorporates the limitation as if
+written in the body of this License.
+
+  13. The Free Software Foundation may publish revised and/or new
+versions of the Lesser General Public License from time to time.
+Such new versions will be similar in spirit to the present version,
+but may differ in detail to address new problems or concerns.
+
+Each version is given a distinguishing version number.  If the Library
+specifies a version number of this License which applies to it and
+"any later version", you have the option of following the terms and
+conditions either of that version or of any later version published by
+the Free Software Foundation.  If the Library does not specify a
+license version number, you may choose any version ever published by
+the Free Software Foundation.
+
+  14. If you wish to incorporate parts of the Library into other free
+programs whose distribution conditions are incompatible with these,
+write to the author to ask for permission.  For software which is
+copyrighted by the Free Software Foundation, write to the Free
+Software Foundation; we sometimes make exceptions for this.  Our
+decision will be guided by the two goals of preserving the free status
+of all derivatives of our free software and of promoting the sharing
+and reuse of software generally.
+
+			    NO WARRANTY
+
+  15. BECAUSE THE LIBRARY IS LICENSED FREE OF CHARGE, THERE IS NO
+WARRANTY FOR THE LIBRARY, TO THE EXTENT PERMITTED BY APPLICABLE LAW.
+EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR
+OTHER PARTIES PROVIDE THE LIBRARY "AS IS" WITHOUT WARRANTY OF ANY
+KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+PURPOSE.  THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE
+LIBRARY IS WITH YOU.  SHOULD THE LIBRARY PROVE DEFECTIVE, YOU ASSUME
+THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
+
+  16. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN
+WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY
+AND/OR REDISTRIBUTE THE LIBRARY AS PERMITTED ABOVE, BE LIABLE TO YOU
+FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR
+CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE
+LIBRARY (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING
+RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A
+FAILURE OF THE LIBRARY TO OPERATE WITH ANY OTHER SOFTWARE), EVEN IF
+SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+DAMAGES.
+
+		     END OF TERMS AND CONDITIONS
+
+           How to Apply These Terms to Your New Libraries
+
+  If you develop a new library, and you want it to be of the greatest
+possible use to the public, we recommend making it free software that
+everyone can redistribute and change.  You can do so by permitting
+redistribution under these terms (or, alternatively, under the terms of the
+ordinary General Public License).
+
+  To apply these terms, attach the following notices to the library.  It is
+safest to attach them to the start of each source file to most effectively
+convey the exclusion of warranty; and each file should have at least the
+"copyright" line and a pointer to where the full notice is found.
+
+    <one line to give the library's name and a brief idea of what it does.>
+    Copyright (C) <year>  <name of author>
+
+    This library is free software; you can redistribute it and/or
+    modify it under the terms of the GNU Lesser General Public
+    License as published by the Free Software Foundation; either
+    version 2.1 of the License, or (at your option) any later version.
+
+    This library is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should have received a copy of the GNU Lesser General Public
+    License along with this library; if not, write to the Free Software
+    Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
+
+Also add information on how to contact you by electronic and paper mail.
+
+You should also get your employer (if you work as a programmer) or your
+school, if any, to sign a "copyright disclaimer" for the library, if
+necessary.  Here is a sample; alter the names:
+
+  Yoyodyne, Inc., hereby disclaims all copyright interest in the
+  library `Frob' (a library for tweaking knobs) written by James Random Hacker.
+
+  <signature of Ty Coon>, 1 April 1990
+  Ty Coon, President of Vice
+
+That's all there is to it!
+
+

INSTALL

@@ -0,0 +1,31 @@
+Installation
+============
+
+1) Generate custom makefiles.
+
+   Run the 'configure' script from the top directory.
+
+   If you don't want to include the LVM1 backwards-compatibility code use:
+     ./configure --with-lvm1=none 
+
+   To separate the LVM1 support into a shared library loaded by lvm.conf use:
+     ./configure --with-lvm1=shared
+
+   Use ./configure --help to see other options.
+
+2) Build and install.
+
+   Run 'make' from the top directory to build everything you configured.
+   Run 'make install' to build and install everything you configured.
+
+   If you only want the device-mapper libraries and tools use
+   'make device-mapper' or 'make install_device-mapper'.
+
+3) If using LVM2, create a configuration file.
+
+   The tools will work fine without a configuration file being
+   present, but you ought to review the example file in doc/example.conf.
+
+Please also refer to the WHATS_NEW file and the manual pages for the 
+individual commands.
+

Makefile.in

@@ -0,0 +1,184 @@
+#
+# Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
+# Copyright (C) 2004-2010 Red Hat, Inc. All rights reserved.
+#
+# This file is part of LVM2.
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions
+# of the GNU General Public License v.2.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+srcdir = @srcdir@
+top_srcdir = @top_srcdir@
+top_builddir = @top_builddir@
+
+SUBDIRS = doc include man
+
+ifeq ("@UDEV_RULES@", "yes")
+  SUBDIRS += udev
+endif
+
+ifeq ("@INTL@", "yes")
+  SUBDIRS += po
+endif
+
+SUBDIRS += lib tools daemons libdm libdaemon
+
+ifeq ("@APPLIB@", "yes")
+  SUBDIRS += liblvm
+endif
+
+ifeq ("@PYTHON_BINDINGS@", "yes")
+  SUBDIRS += python
+endif
+
+SUBDIRS += scripts
+
+# FIXME Should use intermediate Makefiles here!
+ifeq ($(MAKECMDGOALS),distclean)
+  SUBDIRS = doc include man test scripts \
+    libdaemon lib tools daemons libdm \
+    udev po liblvm python \
+    unit-tests/datastruct unit-tests/mm unit-tests/regex
+tools.distclean: test.distclean
+endif
+DISTCLEAN_DIRS += lcov_reports*
+DISTCLEAN_TARGETS += config.cache config.log config.status make.tmpl
+
+include make.tmpl
+
+libdm: include
+lib: libdm libdaemon
+liblvm: lib
+daemons: lib libdaemon tools
+tools: lib libdaemon device-mapper
+po: tools daemons
+scripts: liblvm libdm
+
+lib.device-mapper: include.device-mapper
+libdm.device-mapper: include.device-mapper
+liblvm.device-mapper: include.device-mapper
+daemons.device-mapper: libdm.device-mapper
+tools.device-mapper: libdm.device-mapper
+device-mapper: tools.device-mapper daemons.device-mapper man.device-mapper
+
+ifeq ("@INTL@", "yes")
+lib.pofile: include.pofile
+tools.pofile: lib.pofile
+daemons.pofile: lib.pofile
+po.pofile: tools.pofile daemons.pofile
+pofile: po.pofile
+endif
+
+ifeq ("@PYTHON_BINDINGS@", "yes")
+python: liblvm
+endif
+
+ifneq ("$(CFLOW_CMD)", "")
+tools.cflow: libdm.cflow lib.cflow
+daemons.cflow: tools.cflow
+cflow: include.cflow
+endif
+
+ifneq ("@CSCOPE_CMD@", "")
+cscope.out:
+	@CSCOPE_CMD@ -b -R -s$(top_srcdir)
+all: cscope.out
+endif
+DISTCLEAN_TARGETS += cscope.out
+
+check check_cluster check_local check_lvmetad unit: all
+	$(MAKE) -C test $(@)
+
+install_system_dirs:
+	$(INSTALL_DIR) $(DESTDIR)$(DEFAULT_SYS_DIR)
+	$(INSTALL_ROOT_DIR) $(DESTDIR)$(DEFAULT_ARCHIVE_DIR)
+	$(INSTALL_ROOT_DIR) $(DESTDIR)$(DEFAULT_BACKUP_DIR)
+	$(INSTALL_ROOT_DIR) $(DESTDIR)$(DEFAULT_CACHE_DIR)
+	$(INSTALL_ROOT_DIR) $(DESTDIR)$(DEFAULT_LOCK_DIR)
+	$(INSTALL_ROOT_DIR) $(DESTDIR)$(DEFAULT_RUN_DIR)
+	$(INSTALL_ROOT_DATA) /dev/null $(DESTDIR)$(DEFAULT_CACHE_DIR)/.cache
+
+install_initscripts: 
+	$(MAKE) -C scripts install_initscripts
+
+install_systemd_generators:
+	$(MAKE) -C scripts install_systemd_generators
+
+install_systemd_units:
+	$(MAKE) -C scripts install_systemd_units
+
+ifeq ("@PYTHON_BINDINGS@", "yes")
+install_python_bindings:
+	$(MAKE) -C liblvm/python install_python_bindings
+endif
+
+install_tmpfiles_configuration:
+	$(MAKE) -C scripts install_tmpfiles_configuration
+
+LCOV_TRACES = libdm.info lib.info tools.info \
+	daemons/dmeventd.info daemons/clvmd.info
+CLEAN_TARGETS += $(LCOV_TRACES)
+
+ifneq ("$(LCOV)", "")
+.PHONY: lcov-reset lcov lcov-dated $(LCOV_TRACES)
+
+ifeq ($(MAKECMDGOALS),lcov-dated)
+LCOV_REPORTS_DIR := lcov_reports-$(shell date +%Y%m%d%k%M%S)
+lcov-dated: lcov
+else
+LCOV_REPORTS_DIR := lcov_reports
+endif
+
+lcov-reset:
+	$(LCOV) --zerocounters $(addprefix -d , $(basename $(LCOV_TRACES)))
+
+# maybe use subdirs processing to create tracefiles...
+$(LCOV_TRACES):
+	$(LCOV) -b $(basename $@) -d $(basename $@) \
+		--ignore-errors source -c -o - | $(SED) \
+		-e "s/\(dmeventd_lvm.[ch]\)/plugins\/lvm2\/\1/" \
+		-e "s/dmeventd_\(mirror\|snapshot\|thin\|raid\)\.c/plugins\/\1\/dmeventd_\1\.c/" \
+		>$@
+
+ifneq ("$(GENHTML)", "")
+lcov: $(LCOV_TRACES)
+	$(RM) -r $(LCOV_REPORTS_DIR)
+	$(MKDIR_P) $(LCOV_REPORTS_DIR)
+	for i in $(LCOV_TRACES); do \
+		test -s $$i && lc="$$lc $$i"; \
+	done; \
+	test -z "$$lc" || $(GENHTML) -p @abs_top_builddir@ \
+		-o $(LCOV_REPORTS_DIR) $$lc
+endif
+
+endif
+
+ifeq ("$(TESTING)", "yes")
+# testing and report generation
+RUBY=ruby1.9 -Ireport-generators/lib -Ireport-generators/test
+
+.PHONY: unit-test ruby-test test-programs
+
+# FIXME: put dependencies on libdm and liblvm
+# FIXME: Should be handled by Makefiles in subdirs, not here at top level.
+test-programs:
+	cd unit-tests/regex && $(MAKE)
+	cd unit-tests/datastruct && $(MAKE)
+	cd unit-tests/mm && $(MAKE)
+
+unit-test: test-programs
+	$(RUBY) report-generators/unit_test.rb $(shell find . -name TESTS)
+	$(RUBY) report-generators/title_page.rb
+
+memcheck: test-programs
+	$(RUBY) report-generators/memcheck.rb $(shell find . -name TESTS)
+	$(RUBY) report-generators/title_page.rb
+
+ruby-test:
+	$(RUBY) report-generators/test/ts.rb
+endif

README

@@ -1,2 +1,33 @@
-This is pretty much empty so far...if you can't see subdirectories,
-try 'cvs -f update'
+This tree contains the LVM2 and device-mapper tools and libraries.
+
+For more information about LVM2 read the changelog in the WHATS_NEW file.
+Installation instructions are in INSTALL.
+
+There is no warranty - see COPYING and COPYING.LIB.
+
+Tarballs are available from:
+  ftp://sources.redhat.com/pub/lvm2/
+
+The source code is stored in git:
+  http://git.fedorahosted.org/git/lvm2.git
+  git clone git://git.fedorahosted.org/git/lvm2.git
+
+Mailing list for general discussion related to LVM2:
+  linux-lvm@redhat.com
+  Subscribe from https://www.redhat.com/mailman/listinfo/linux-lvm
+
+Mailing lists for LVM2 development, patches and commits:
+  lvm-devel@redhat.com
+  Subscribe from https://www.redhat.com/mailman/listinfo/linux-lvm
+
+  lvm2-commits@lists.fedorahosted.org (Read-only archive of commits)
+  Subscribe from https://fedorahosted.org/mailman/listinfo/lvm2-commits
+
+Mailing list for device-mapper development, including kernel patches
+and multipath-tools:
+  dm-devel@redhat.com
+  Subscribe from https://www.redhat.com/mailman/listinfo/dm-devel
+
+The source code repository used until 7th June 2012 is accessible here:
+  http://sources.redhat.com/cgi-bin/cvsweb.cgi/LVM2/?cvsroot=lvm2.
+

VERSION

@@ -0,0 +1 @@
+2.02.99(2)-git (2012-10-22)

VERSION_DM

@@ -0,0 +1 @@
+1.02.78-git (2012-10-22)

WHATS_NEW_DM

@@ -0,0 +1,854 @@
+Version 1.02.78 - 
+===================================
+  Add DM_TO_STRING public macro.
+  Always return success on dmeventd -V command call.
+  Fix parsing of 64bit snapshot status in dmeventd snapshot plugin.
+  Add dm_get_status_snapshot() for parsing snapshot status.
+  Detecte mounted fs also via reading /proc/self/mountinfo.
+  Add dm_mountinfo_read() for parsing /proc/self/mountinfo.
+  Report error for nonexisting devices in dmeventd communication.
+  Prevent double free error after dmeventd call of _fill_device_data().
+  Update dmevent structure message_data to simplify/fix error path handling.
+  Validate passed params to dm_get_status_raid/thin/thin_pool(). 
+  Fix 'dmsetup splitname -o' to not fail if used without '-c' switch (1.02.68).
+  Add dm_config_write_{node_out/one_node_out} for enhanced config output.
+  Add dm_config_value_is_bool to check for boolean value in supported formats.
+  Fix config node lookup inside empty sections to not return the section itself.
+  Extend support for status info of thin pool target.
+  Fix segfault for truncated string token in config file after the first '"'.
+  Close open dmeventd FIFO file descriptors on exec (FD_CLOEXEC).
+  Fix resource leak in error path of dmeventd's umount of thin volume.
+  Automatically deactivate failed preloaded dm tree node.
+  Add DM_DISABLE_UDEV environment variable to manage dev nodes by libdm only.
+  Fix dm_task_set_cookie to properly process udev flags if udev_sync disabled.
+
+Version 1.02.77 - 15th October 2012
+===================================
+  Support unmount of thin volumes from pool above thin pool threshold.
+  Update man page to reflect that dm UUIDs are being mangled as well.
+  Apply 'dmsetup mangle' for dm UUIDs besides dm names.
+  Add 'mangled_uuid' and 'unmangled_uuid' fields to dmsetup info -c -o.
+  Mangle device UUID on dm_task_set_uuid/newuuid call if necessary.
+  Add dm_task_get_uuid_mangled/unmangled to libdevmapper.
+  Always reset delay_resume_if_new flag when stacking thin pool over anything.
+  Don't create value for dm_config_node and require dm_config_create_value call.
+  Check for existing new_name for dmsetup rename.
+  Fix memory leak in dmsetup _get_split_name() error path.
+
+Version 1.02.76 - 7th August 2012
+=================================
+  Add dm_vasprintf to libdevmapper.
+  Allow --noflush with dmsetup status and wait (for thin target).
+  Add dm_config_write_one_node to libdevmapper.
+  Support thin pool message release/reserve_metadata_snap in libdevmapper.
+  Support thin pool discards and external origin features in libdevmapper.
+  Add configure --enable-udev-rule-exec-detection to detect exec path in rules.
+  Use sbindir in udev rules by default and remove executable path detection.
+  Remove hard-coded paths for dmeventd fifos and use default-dm-run-dir.
+  Add configure --with-lvmetad-pidfile to remove hard-coded value.
+  Add configure --with-default-pid-dir for common directory with pid files.
+  Add configure --with-default-dm-run-dir to set run directory for dm tools.
+  Detect kernel_send() errors in cmirrord.
+  Add __attribute__ instrumentation to libdevmapper.h.
+  Print clean_bits instead of sync_bits in pull_state in cmirrord.
+  Add tests for errors from closedir(), close() in cmirrord.
+  Add documentation references in systemd units.
+  Remove veritysetup.  Now maintained with cryptsetup.
+
+Version 1.02.75 - 8th June 2012
+===============================
+  Upstream source repo now fedorahosted.org git not sources.redhat.com CVS.
+  Remove unsupported udev_get_dev_path libudev call used for checking udev dir.
+  Set delay_resume_if_new on deptree snapshot origin.
+  Log value chosen in _find_config_bool like other variable types do.
+  Wait for dmeventd to exit after sending it DM_EVENT_CMD_DIE when restarting.
+  Append 'Used' to {Blk}DevNames/DevNos dmsetup report headers for clarity.
+  Add configure --with-veritysetup for independent veritysetup tool.
+  Properly support supplied dmevent path in dm_event_register_handler().
+  Remove dmeventd fifos on exit if they are not managed by systemd.
+  Use SD_ACTIVATION environment variable in systemd units to detect systemd.
+  Only start a new dmeventd instance on restart if one was already running.
+  Extend the time waited for input from dmeventd fifo to 5 secs. (1.02.73)
+
+Version 1.02.74 - 6th March 2012
+================================
+  Check for multiply-mangled names in auto mangling mode.
+  Fix dm_task_get_name_unmangled to not unmangle already unmangled name.
+  Check whether device names are properly mangled on ioctl return.
+  Deactivation of failed thin check on thin pool returns success.
+
+Version 1.02.73 - 3rd March 2012
+================================
+  Test _thread_registry list with holding mutex in dmeventd.
+  Add dm_tree_node_set_callback() for preload and deactivation hooks.
+  Drop unsupported TRIM message for thin pool.
+  Improve logging for fifo startup in dmeventd.
+  Better detection of missing dmeventd fifo connection (1.02.71).
+  Add a few pointer validations in dmsetup.
+  Support dm_task_get_driver_version() query without version string.
+  Log failure of pthread_join when cleaning unused threads in dmeventd.
+  Fix empty string warning logic in _find_config_str. (1.02.68)
+  Fix dm_task_set_name to properly resolve path to dm name (1.02.71).
+  Add dm_strncpy() function as a faster strncpy() replacement.
+
+Version 1.02.72 - 23rd February 2012
+====================================
+  Avoid memory reallocation for dm_asprintf.
+
+Version 1.02.71 - 20th February 2012
+====================================
+  Switch to using built-in blkid in 13-dm-disk.rules.
+  Add "watch" rule to 13-dm-disk.rules.
+  Detect failing fifo and skip 20s retry communication period.
+  Add DM_DEFAULT_NAME_MANGLING_MODE environment variable as an override.
+  Add dm_lib_init to automatically initialise device-mapper library on load.
+  Replace any '\' char with '\\' in dm table specification on input.
+  Add mangle command to dmsetup to provide renaming to correct mangled form.
+  Add 'mangled_name' and 'unmangled_name' fields to dmsetup info -c -o.
+  Add --manglename option to dmsetup to select the name mangling mode.
+  Add dm_task_get_name_mangled/unmangled to libdevmapper.
+  Mangle device name on dm_task_set_name/newname call if necessary.
+  Add dm_set/get_name_mangling_mode to set/get name mangling in libdevmapper.
+  Add configure --with-default-name-mangling for udev-friendly dev name charset.
+  Test for parsed words in _umount() dmeventd snapshot plugin.
+  Fix memory leak in fail path of parse_loop_device_name() in dmsetup.
+  Check for missing reply_uuid in dm_event_get_registered_device().
+  Check for allocation failure in dmeventd restart().
+  Add few missing allocation failures tests in dmsetup.
+  Fix potential risk of writing in front of buffer in _sysfs_get_dm_name().
+
+Version 1.02.70 - 12th February 2012
+====================================
+  Fix dm_event_get_version() check.
+  Add pointer test for dependency check in _add_dev().
+  Validate name and uuid params of dm_tree_add_new_dev_with_udev_flags().
+  Do not crash for dm_report_init() sort_key == NULL and behave like "".
+  Return error for failing allocation in dm_asprintf().
+  Add missing test for failing allocation in dm_realloc() code.
+  Add test for memory allocation failures in regex matcher code.
+  Simplify dm_task_set_geometry() and use dm_asprintf().
+  Set all parameters to 0 for dm_get_next_target() for NULL return.
+  Fix fd resource leak in error path for _udev_notify_sem_create().
+  Leave space for '\0' for readline() call in _sysfs_get_kernel_name().
+
+Version 1.02.69 - 1st February 2012
+===================================
+  Clean up dmeventd systemd unit ordering and requirements.
+
+Version 1.02.68 - 26th January 2012
+===================================
+  Reset all members of info struct in dm_tree_add_new_dev_with_udev_flags.
+  Add dmsetup wipe_table to replace table with one that uses error target.
+  Add 'blkdevname' and 'blkdevs_used' fields to dmsetup info -c -o.
+  Add 'blkdevname' option to dmsetup ls --tree to see block device names.
+  Add -o devno/blkdevname/devname to dmsetup deps and ls.
+  Add dm_device_get_name to get map name or block device name for given devno.
+  Remove empty devices when clearing left-over inactive tables in deptree.
+  Add dm_uuid_prefix/dm_set_uuid_prefix to override hard-coded LVM- prefix.
+  Improve dmsetup man page description of readahead parameter.
+  Use sysfs to set/get readahead if possible.
+  Fix lvm2-monitor init script to use normalized output when using vgs.
+  Add test for max length (DM_MAX_TYPE_NAME) of target type name.
+  Include a copy of kernel DM documentation in doc/kernel.
+  Improve man page style for dmsetup and mention more targets.
+  Fix _get_proc_number to be tolerant of malformed /proc/misc entries.
+  Fix missing thread list manipulation protection in dmeventd.
+  Add ExecReload to dm-event.service for systemd to reload dmeventd properly.
+  Add dm_config_tree_find_str_allow_empty and dm_config_find_str_allow_empty.
+  Fix compile-time pool memory locking with DEBUG_MEM.
+  Fix valgrind error reports in free of pool chunks with DEBUG_MEM.
+  Align size of structure chunk for fast pool allocator to 8 bytes.
+  Simplify some pointer operations in dm_free_aux() debug code.
+  Remove unused dbg_malloc.h file from source tree.
+  Cleanup backtraces for _create_and_load_v4().
+  Fix alignment warning in bitcount calculation for raid segment.
+  Allocate dm_tree structure from dm_tree pool.
+  Update debug logging for _resume_node.
+  Add functions to support thin provisioning target.
+  Improve libdm-config error path reporting.
+  Update dmsetup resume man with --addnodeonresume/create options.
+  Add dependency for dm man pages to man subdirectory make all target.
+  Add dm_tree_retry_remove to use retry logic for device removal in a dm_tree.
+  Add dm_device_has_mounted_fs fn to check mounted filesystem on a device.
+  Add dm_device_has_holders fn to to check use of the device by another device.
+  Add dm_sysfs_dir to libdevmapper to retrieve sysfs location set.
+  Add dm_set_sysfs_dir to libdevmapper to set sysfs location.
+  Add --retry option for dmsetup remove to retry removal if not successful.
+  Add dm_task_retry_remove fn to use retry logic for device removal.
+  Remove unused passed parameters for _mirror_emit_segment_line().
+  Add dm_config and string character escaping functions to libdevmapper.
+  Mark unreleased memory pools as internal error.
+
+Version 1.02.67 - 19th August 2011
+==================================
+  Add dm_tree_node_add_null_area for temporarily-missing raid devs tracked.
+
+Version 1.02.66 - 12th August 2011
+==================================
+  Release geometry buffer in dm_task_destroy.
+  Update udev rules to skip DM flags decoding for removed devices.
+  Add compile-time pool memory locking options (to debug shared VG structs).
+  Remove device name prefix from dmsetup line output if -j & -m or -u supplied.
+  Remove support for the original version 1 dm ioctls.
+  Add missing check for allocation failure _create_dir_recursive().
+  Add support for systemd file descriptor handover in dmeventd.
+  Fix memory leak in dmsetup _message() memory allocation error path.
+  Use new oom killer adjustment interface (oom_score_adj) when available.
+  Add systemd unit files for dmeventd.
+  Fix read-only identical table reload supression.
+
+Version 1.02.65 - 8th July 2011
+===============================
+  Remove dev name prefix from dmsetup line output if exactly one dev requested.
+  Report internal error if suspending a device using an already-suspended dev.
+  Report error if a table load requiring target parameters has none supplied.
+  Add dmsetup --checks and dm_task_enable_checks framework to validate ioctls.
+  Add age_in_minutes parameter to dmsetup udevcomplete_all.
+  Return immediately from dm_lib_exit() if called more than once.
+  Disable udev fallback by default and add --verifyudev option to dmsetup.
+  Report internal error if any table is loaded while any dev is known suspended.
+  Add dm_get_suspended_counter() for number of devs in suspended state by lib.
+  Fix "all" report field prefix matching to include label fields with pv_all.
+  Delay resuming new preloaded mirror devices with core logs in deptree code.
+  Accept new kernel version 3 uname formats in initialisation.
+
+Version 1.02.64 - 29th April 2011
+==================================
+  Require libudev >= 143 when compiling with udev support.
+  Use word alignment for dm_pool_strdup() and dm_pool_strndup().
+  Use dm_snprintf() to fix signedness warning in dm_set_dev_dir().
+  Use unsigned loop counter to fix signedness warning in _other_node_ops().
+  Fix const cast in dmsetup calls of dm_report_field_string().
+  Streamline /dev/mapper/control node code for common cases.
+  Use hard-coded dm control node device number for 2.6.36 kernels and above.
+  Improve stack debug reporting in dm_task_create().
+  Fallback to control node creation only if node doesn't exist yet.
+  Change dm_hash binary functions to take void *key instead of char *.
+  Fix uninitialised memory use with empty params in _reload_with_suppression_v4.
+  Lower severity of selabel_lookup and matchpathcon failure to log_debug.
+  Add test for failed allocation from dm_task_set_uuid() in dmeventd.
+  Add dm_event_get_version to dmeventd for use with -R.
+  Avoid dmeventd core dumps when handling request with unknown command ID.
+  Have dmeventd -R start up even when no existing copy is running.
+  Accept multiple mapped device names on many dmsetup command lines.
+  Fix dm_udev_wait calls in dmsetup to occur before readahead display not after.
+  Include an implicit dm_task_update_nodes() within dm_udev_wait().
+  Fix _create_and_load_v4 not to lose the --addnodeoncreate setting (1.02.62).
+  Add inactive table query support for kernel driver >= 4.11.6 (RHEL 5.7).
+  Log debug open_count in _node_has_closed_parents().
+  Add a const to dm_report_field_string() data parameter.
+
+Version 1.02.63 - 9th February 2011
+===================================
+  Reinstate DEBUG_MEM as it's part of the API. (1.02.62)
+
+Version 1.02.62 - 4th February 2011
+===================================
+  Add configure --with-device-nodes-on=create for previous behaviour.
+  Move creation of device nodes from 'create' to 'resume'.
+  Add --addnodeonresume and --addnodeoncreate options to dmsetup.
+  Add dm_task_set_add_node to libdevmapper to control dev node creation time.
+  Add dm_task_secure_data to libdevmapper to wipe ioctl buffers in kernel.
+  Log debug message when expected uevent is not generated.
+  Only compile memory debugging code when DEBUG_MEM is set.
+  Set DM_UDEV_DISABLE_OTHER_RULES_FLAG for suspended DM devices in udev rules.
+  Begin a new pool object for each row in _output_as_rows() correctly.
+
+Version 1.02.61 - 10th January 2011
+===================================
+  Add DM_COOKIE_AUTO_CREATE to libdevmapper.h.
+  Export DM_CONTROL_NODE_UMASK and use it while creating /dev/mapper/control.
+
+Version 1.02.60 - 20th December 2010
+====================================
+  Check for unlink failure in remove_lockfile() in dmeventd.
+  Use dm_free for dm_malloc-ed areas in _clog_ctr/_clog_dtr in cmirrord.
+  Use char* arithmetic in _process_all() & _targets() in dmsetup.
+  Change dm_regex_create() API to accept const char * const *patterns.
+  Add new dm_prepare_selinux_context fn to libdevmapper and use it throughout.
+  Detect existence of new SELinux selabel interface during configure.
+
+Version 1.02.59 - 6th December 2010
+===================================
+  Add backtraces to _process_mapper_dir and  _create_and_load_v4 error paths.
+  Remove superfluous checks for NULL before calling dm_free.
+
+Version 1.02.58 - 22nd November 2010
+====================================
+  Fix _output_field crash from field_id free with DEBUG_MEM. (1.02.57)
+
+Version 1.02.57 - 8th November 2010
+===================================
+  Fix regex optimiser not to ignore RHS of OR nodes in _find_leftmost_common.
+  Add dmeventd -R to restart dmeventd without losing monitoring state. (1.02.56)
+  Fix memory leak of field_id in _output_field function.
+  Allocate buffer for reporting functions dynamically to support long outputs.
+
+Version 1.02.56 - 25th October 2010
+===================================
+  Return const pointer from dm_basename() in libdevmapper.
+  Implement dmeventd -R to restart without state loss.
+  Add dm_zalloc and use it and dm_pool_zalloc throughout.
+  Add --setuuid to dmsetup rename.
+  Add dm_task_set_newuuid to set uuid of mapped device post-creation.
+
+Version 1.02.55 - 24th September 2010
+=====================================
+  Fix the way regions are marked complete to avoid slow --nosync cmirror I/O.
+  Add DM_REPORT_FIELD_TYPE_ID_LEN to libdevmapper.h.
+
+Version 1.02.54 - 18th August 2010
+==================================
+  Fix dm-mod autoloading logic to not assume control node is set correctly.
+  Add dmeventd/executable to lvm.conf to test alternative dmeventd.
+  Export dm_event_handler_set_dmeventd_path to override built-in dmeventd path.
+  Generate libdevmapper-event exported symbols.
+  Remove superfluous NULL pointer tests before dm_free from dmeventd.
+  Assume dm-mod autoloading support is in kernel 2.6.36 and higher, not 2.6.35.
+  Fix udev rules to support udev database content generated by older rules.
+  Reinstate detection of inappropriate uevent with DISK_RO set and suppress it.
+  Fix regex ttree off-by-one error.
+  Add --enable-valgrind-pool to configure.
+  Fix segfault in regex matcher with characters of ordinal value > 127.
+  Fix 'void*' arithmetic warnings in dbg_malloc.c and libdm-iface.c.
+  Wait for node creation before displaying debug info in dmsetup.
+  Fix return status 0 for "dmsetup info -c -o help".
+  Add check for kernel semaphore support and disable udev_sync if not available.
+
+Version 1.02.53 - 28th July 2010
+================================
+  Revert failed table load preparation after "create, load and resume".
+  Switch dmeventd to use dm_create_lockfile and drop duplicate code.
+  Add dm_create_lockfile to libdm to handle pidfiles for all daemons.
+  Replace lookup with next in struct dfa_state & calculate states on demand.
+  Improve the regex matcher, reducing the number of charset nodes used.
+  Add dm_regex_fingerprint to facilitate regex testing.
+  Skip ffs(0) in _test_word in bitset functions.
+  Use "nowatch" udev rule for inappropriate devices.
+
+Version 1.02.52 - 6th July 2010
+===============================
+  Fix dmlosetup snprintf %llu compiler warning.
+  Add parentheses to some libdevmapper.h macro arguments.
+  Add printf format attributes to dm_{sn,as}printf and fix a caller.
+  Move dmeventd man page from install_lvm2 to install_device-mapper. (1.02.50)
+
+Version 1.02.51 - 30th June 2010
+================================
+  Generate libdevmapper exported symbols from header file.
+
+Version 1.02.50 - 23rd June 2010
+================================
+  Fix INTERNAL_ERROR typo in ioctl iface unknown task message.
+  Fix udev rules to handle spurious events properly.
+  Use C99 [] not [0] in dm_ulog_request struct to avoid abort when fortified.
+  Allow use of devmapper header file in C++ mode (extern "C" and __typeof__).
+  Add dmeventd man page.
+
+Version 1.02.49 - 4th June 2010
+===============================
+  Support autoloading of dm-mod module for kernels from 2.6.35.
+  Document 'clear' in dmsetup man page.
+  Fix semctl parameter (union) to avoid misaligned parameter on some arches.
+  Add dm_tree_node_set_presuspend_node() to presuspend child when deactivating.
+  Initial support for replicator target.
+
+Version 1.02.48 - 17th May 2010
+================================
+  Use -d to control level of messages sent to syslog by dmeventd.
+  Change -d to -f to run dmeventd in foreground.
+  Do not print encryption key in message debug output (cryptsetup luksResume).
+  Fix dmeventd static build library dependencies.
+  Fix udev flags on remove in create_and_load error path.
+
+Version 1.02.47 - 30th April 2010
+=================================
+  Add support for new IMPORT{db} udev rule.
+  Add DM_UDEV_PRIMARY_SOURCE_FLAG udev flag to recognize proper DM events.
+  Also include udev libs in libdevmapper.pc when udev_sync is enabled.
+  Cache bitset locations to speed up _calc_states.
+  Add a regex optimisation pass for shared prefixes and suffixes.
+  Add dm_bit_and and dm_bitset_equal to libdevmapper.
+  Simplify dm_bitset_create.
+  Speed up dm_bit_get_next with ffs().
+
+Version 1.02.46 - 14th April 2010
+=================================
+  Change dm_tree_deactivate_children to fail if device is open.
+  Wipe memory buffers for dm-ioctl parameters before releasing.
+  Strictly require libudev if udev_sync is used.
+  Add support for ioctl's DM_UEVENT_GENERATED_FLAG.
+
+Version 1.02.45 - 9th March 2010
+================================
+  Add --showkeys parameter description to dmsetup man page.
+  Add --help option as synonym for help command.
+
+Version 1.02.44 - 15th February 2010
+====================================
+  Add DM_UDEV_DISABLE_LIBRARY_FALLBACK udev flag to rely on udev only.
+  Export dm_udev_create_cookie function to create new cookies on demand.
+  Add --udevcookie, udevcreatecookie and udevreleasecookie to dmsetup.
+  Set udev state automatically instead of using DM_UDEV_DISABLE_CHECKING.
+
+Version 1.02.43 - 21st January 2010
+===================================
+  Remove bitset, hash and pool headers superceded by libdevmapper.h.
+  Fix off-by-one error causing bad cluster mirror table construction.
+
+Version 1.02.42 - 14th January 2010
+===================================
+  Add support for the "snapshot-merge" kernel target (2.6.33-rc1).
+  Introduce a third activation_priority level in dm_tree_activate_children.
+
+Version 1.02.41 - 12th January 2010
+===================================
+  If DM_UDEV_DISABLE_CHECKING is set in environment, disable udev warnings.
+  Add dm_tree_add_dev_with_udev_flags to provide wider support for udev flags.
+  Add --noudevrules option for dmsetup to disable /dev node management by udev.
+  Fix 'dmsetup info -c -o all' to show all fields.
+  Return errors if dm_tree_*_children functions fail.
+  Fix coredump and memory leak for 'dmsetup help -c'.
+  Disable udev rules for change events with DISK_RO set.
+
+Version 1.02.40 - 19th November 2009
+====================================
+  Fix install_device-mapper Makefile target to not build dmeventd plugins.
+  Support udev flags even when udev_sync is disabled or not compiled in.
+  Remove 'last_rule' from udev rules: honour DM_UDEV_DISABLE_OTHER_RULES_FLAG.
+  Add dmsetup --inactive support.
+  Add dm_task_query_inactive_table to libdevmapper for kernel driver >= 4.16.
+  Fix hash lookup segfault when keys compared are different lengths.
+
+Version 1.02.39 - 26th October 2009
+===================================
+  Remove strict default permissions for DM devices from 95-dm-notify.rules.
+  Add dmsetup udevflags command to decode udev flags in given cookie value.
+  Support udev flags in libdevmapper incl. dm_tree_add_new_dev_with_udev_flags.
+  Make libdm ABI consistent when built with/without selinux support.
+
+Version 1.02.38 - 25th September 2009
+=====================================
+  Export DM_DEV_DIR_UMASK, the default umask for /dev directories created.
+  Handle any path supplied to dm_task_set_name by looking up in /dev/mapper.
+  Add several examples to 12-dm-permissions.rules.
+  Add splitname and --yes to dmsetup man page.
+  Fix _mirror_emit_segment_line return code.
+  Fix dmeventd _temporary_log_fn parameters. (2.02.50)
+
+Version 1.02.37 - 15th September 2009
+=====================================
+  Add dmsetup manpage entries for udevcomplete_all and udevcookies.
+  Check udev is running when processing cookies and retain state internally.
+  Add y|--yes option to dmsetup for default 'yes' answer to prompts.
+  Fix tools Makefile to process dmsetup sources separately.
+  Restore umask when device node creation fails.
+  Check kernel vsn to use 'block_on_error' or 'handle_errors' in mirror table.
+  Add dm-log-userspace.h to tree for cmirrord builds.
+
+Version 1.02.36 - 6th August 2009
+=================================
+  Add udevcookies, udevcomplete, udevcomplete_all and --noudevwait to dmsetup.
+  Add libdevmapper functions to support synchronisation with udev.
+
+Version 1.02.35 - 28th July 2009
+================================
+  Add LOG_LINE_WITH_ERRNO macro.
+  Use log_error macro consistently throughout in place of log_err.
+
+Version 1.02.34 - 15th July 2009
+================================
+  Use _exit() not exit() after forking to avoid flushing libc buffers twice.
+  Rename plog macro to LOG_LINE & add LOG_MESG variant for dm_dump_memory_debug.
+  Change plog to use dm_log_with_errno unless deprecated dm_log_init was used.
+  Add dm_log_with_errno and dm_log_with_errno_init, deprecating the old fns.
+  Fix whitespace in linear target line to fix identical table line detection.
+  Add device number to more log messages during activation.
+
+Version 1.02.33 - 30th June 2009
+================================
+  Don't fallback to default major number: use dm_task_set_major_minor. (1.02.31)
+  Do not fork daemon when dmeventd cannot be found.
+  Add crypt target handling to libdevmapper tree nodes.
+  Add splitname command to dmsetup.
+  Add subsystem, vg_name, lv_name, lv_layer fields to dmsetup reports.
+  Make mempool optional in dm_split_lvm_name().
+
+Version 1.02.32 - 21st May 2009
+===============================
+  Only generate libdevmapper.a when configured to link statically.
+  Export dm_tree_node_size_changed() from libdevmapper.
+  Propagate the table size_changed property up the dm device tree.
+  Detect failure to free memory pools when releasing the library.
+  Fix segfault when getopt processes dmsetup -U, -G and -M options.
+
+Version 1.02.31 - 3rd March 2009
+================================
+  If kernel supports only one dm major number, use in place of any supplied.
+
+Version 1.02.30 - 26th January 2009
+====================================
+  Add "all" field to reports expanding to all fields of report type.
+  Enforce device name length and character limitations in libdm.
+  Replace _dm_snprintf with EMIT_PARAMS macro for creating target lines.
+
+Version 1.02.29 - 10th November 2008
+====================================
+  Merge device-mapper into the LVM2 tree.
+  Split out dm-logging.h from log.h.
+  Use lvm-types.h.
+  Add usrsbindir to configure.
+
+Version 1.02.28 - 18th September 2008
+=====================================
+  Only resume devices in dm_tree_preload_children if size changes.
+  Extend deptree buffers so the largest possible device numbers fit.
+  Generate versioned libdevmapper-event.so.
+  Underline longer report help text headings.
+
+Version 1.02.27 - 25th June 2008
+================================
+  Align struct memblock in dbg_malloc for sparc.
+  Add --unquoted and --rows to dmsetup.
+  Avoid compiler warning about cast in dmsetup.c's OFFSET_OF macro.
+  Fix inverted no_flush debug message.
+  Remove --enable-jobs from configure. (Set at runtime instead.)
+  Bring configure.in and list.h into line with the lvm2 versions.
+
+Version 1.02.26 - 6th June 2008
+===============================
+  Initialise params buffer to empty string in _emit_segment.
+  Skip add_dev_node when ioctls disabled.
+  Make dm_hash_iter safe against deletion.
+  Accept a NULL pointer to dm_free silently.
+  Add tables_loaded, readonly and suspended columns to reports.
+  Add --nameprefixes to dmsetup.
+  Add field name prefix option to reporting functions.
+  Calculate string size within dm_pool_grow_object.
+
+Version 1.02.25 - 10th April 2008
+=================================
+  Remove redundant if-before-free tests.
+  Use log_warn for reporting field help text instead of log_print.
+  Change cluster mirror log type name (s/clustered_/clustered-/)
+
+Version 1.02.24 - 20th December 2007
+====================================
+  Fix deptree to pass new name to _resume_node after a rename.
+  Suppress other node operations if node is deleted.
+  Add node operation stack debug messages.
+  Report error when empty device name passed to readahead functions.
+  Fix minimum readahead debug message.
+
+Version 1.02.23 - 5th December 2007
+===================================
+  Update dm-ioctl.h after removal of compat code.
+  Add readahead support to libdevmapper and dmsetup.
+  Fix double free in a libdevmapper-event error path.
+  Fix configure --with-dmeventd-path substitution.
+  Allow a DM_DEV_DIR environment variable to override /dev in dmsetup.
+  Create a libdevmapper.so.$LIB_VERSION symlink within the build tree.
+  Avoid static link failure with some SELinux libraries that require libpthread.
+  Remove obsolete dmfs code from tree and update INSTALL.
+
+Version 1.02.22 - 21st August 2007
+==================================
+  Fix inconsistent licence notices: executables are GPLv2; libraries LGPLv2.1.
+  Update to use autoconf 2.61, while still supporting 2.57.
+  Avoid repeated dm_task free on some dm_event_get_registered_device errors.
+  Introduce log_sys_* macros from LVM2.
+  Export dm_fclose and dm_create_dir; remove libdm-file.h.
+  Don't log EROFS mkdir failures in _create_dir_recursive (for LVM2).
+  Add fclose wrapper dm_fclose that catches write failures (using ferror).
+
+Version 1.02.21 - 13th July 2007
+================================
+  Introduce _LOG_STDERR to send log_warn() messages to stderr not stdout.
+  Fix dmsetup -o devno string termination. (1.02.20)
+
+Version 1.02.20 - 15th June 2007
+================================
+  Fix default dmsetup report buffering and add --unbuffered.
+  Add tree-based and dependency fields to dmsetup reports.
+
+Version 1.02.19 - 27th April 2007
+=================================
+  Standardise protective include file #defines.
+  Add regex functions to library.
+  Avoid trailing separator in reports when there are hidden sort fields.
+  Fix segfault in 'dmsetup status' without --showkeys against crypt target.
+  Deal with some more compiler warnings.
+  Introduce _add_field() and _is_same_field() to libdm-report.c.
+  Fix some libdevmapper-event and dmeventd memory leaks.
+  Remove unnecessary memset() return value checks.
+  Fix a few leaks in reporting error paths. [1.02.15+]
+
+Version 1.02.18 - 13th February 2007
+====================================
+  Improve dmeventd messaging protocol: drain pipe and tag messages.
+
+Version 1.02.17 - 29th January 2007
+===================================
+  Add recent reporting options to dmsetup man page.
+  Revise some report fields names.
+  Add dmsetup 'help' command and update usage text.
+  Use fixed-size fields in report interface and reorder.
+
+Version 1.02.16 - 25th January 2007
+===================================
+  Add some missing close() and fclose() return value checks.
+  Migrate dmsetup column-based output over to new libdevmapper report framework.
+  Add descriptions to reporting field definitions.
+  Add a dso-private variable to dmeventd dso interface.
+  Add dm_event_handler_[gs]et_timeout functions.
+  Streamline dm_report_field_* interface.
+  Add cmdline debug & version options to dmeventd.
+  Add DM_LIB_VERSION definition to configure.h.
+  Suppress 'Unrecognised field' error if report field is 'help'.
+  Add --separator and --sort to dmsetup (unused).
+  Make alignment flag optional when specifying report fields.
+
+Version 1.02.15 - 17th January 2007
+===================================
+  Add basic reporting functions to libdevmapper.
+  Fix a malloc error path in dmsetup message.
+  More libdevmapper-event interface changes and fixes.
+  Rename dm_saprintf() to dm_asprintf().
+  Report error if NULL pointer is supplied to dm_strdup_aux().
+  Reinstate dm_event_get_registered_device.
+
+Version 1.02.14 - 11th January 2007
+===================================
+  Add dm_saprintf().
+  Use CFLAGS when linking so mixed sparc builds can supply -m64.
+  Add dm_tree_use_no_flush_suspend().
+  Lots of dmevent changes including revised interface.
+  Export dm_basename().
+  Cope with a trailing space when comparing tables prior to possible reload.
+  Fix dmeventd to cope if monitored device disappears.
+
+Version 1.02.13 - 28 Nov 2006
+=============================
+  Update dmsetup man page (setgeometry & message).
+  Fix dmsetup free after getline with debug.
+  Suppress encryption key in 'dmsetup table' output unless --showkeys supplied.
+
+Version 1.02.12 - 13 Oct 2006
+=============================
+  Avoid deptree attempting to suspend a device that's already suspended.
+
+Version 1.02.11 -  12 Oct 2006
+==============================
+  Add suspend noflush support.
+  Add basic dmsetup loop support.
+  Switch dmsetup to use dm_malloc and dm_free.
+
+Version 1.02.10 - 19 Sep 2006
+=============================
+  Add dm_snprintf(), dm_split_words() and dm_split_lvm_name() to libdevmapper.
+  Reorder mm bounds_check code to reduce window for a dmeventd race.
+
+Version 1.02.09 - 15 Aug 2006
+=============================
+  Add --table argument to dmsetup for a one-line table.
+  Abort if errors are found during cmdline option processing.
+  Add lockfs indicator to debug output.
+
+Version 1.02.08 - 17 July 2006
+==============================
+  Append full patch to check in emails.
+  Avoid duplicate dmeventd subdir with 'make distclean'.
+  Update dmsetup man page.
+  Add --force to dmsetup remove* to load error target.
+  dmsetup remove_all also performs mknodes.
+  Don't suppress identical table reloads if permission changes.
+  Fix corelog segment line.
+  Suppress some compiler warnings.
+
+Version 1.02.07 - 11 May 2006
+=============================
+  Add DM_CORELOG flag to dm_tree_node_add_mirror_target().
+  Avoid a dmeventd compiler warning.
+
+Version 1.02.06 - 10 May 2006
+=============================
+  Move DEFS into configure.h.
+  Fix leaks in error paths found by coverity.
+  Remove dmsetup line buffer limitation.
+
+Version 1.02.05 - 19 Apr 2006
+=============================
+  Separate install_include target in makefiles.
+  Separate out DEFS from CFLAGS.
+  Support pkg-config.
+  Check for libsepol.
+
+Version 1.02.04 - 14 Apr 2006
+=============================
+  Bring dmsetup man page up-to-date.
+  Use name-based device refs if kernel doesn't support device number refs.
+  Fix memory leak (struct dm_ioctl) when struct dm_task is reused.
+  If _create_and_load_v4 fails part way through, revert the creation.
+  dmeventd thread/fifo fixes.
+  Add file & line to dm_strdup_aux().
+  Add setgeometry.
+
+Version 1.02.03 - 7 Feb 2006
+============================
+  Add exported functions to set uid, gid and mode.
+  Rename _log to dm_log and export.
+  Add dm_tree_skip_lockfs.
+  Fix dm_strdup debug definition.
+  Fix hash function to avoid using a negative array offset.
+  Don't inline _find in hash.c and tidy signed/unsigned etc.
+  Fix libdevmapper.h #endif.
+  Fix dmsetup version driver version.
+  Add sync, nosync and block_on_error mirror log parameters.
+  Add hweight32.
+  Fix dmeventd build.
+
+Version 1.02.02 - 2 Dec 2005
+============================
+  dmeventd added.
+  Export dm_task_update_nodes.
+  Use names instead of numbers in messages when ioctls fail.
+
+Version 1.02.01 - 23 Nov 2005
+=============================
+  Resume snapshot-origins last.
+  Drop leading zeros from dm_format_dev.
+  Suppress attempt to reload identical table.
+  Additional LVM- prefix matching for transitional period.
+
+Version 1.02.00 - 10 Nov 2005
+=============================
+  Added activation functions to library.
+  Added return macros.
+  Also suppress error if device doesn't exist with DM_DEVICE_STATUS.
+  Export dm_set_selinux_context().
+  Add dm_driver_version().
+  Added dependency tree functions to library.
+  Added hash, bitset, pool, dbg_malloc to library.
+  Added ls --tree to dmsetup.
+  Added dmsetup --nolockfs support for suspend/reload.
+
+Version 1.01.05 - 26 Sep 2005
+=============================
+  Resync list.h with LVM2.
+  Remember increased buffer size and use for subsequent calls.
+  On 'buffer full' condition, double buffer size and repeat ioctl.
+  Fix termination of getopt_long() option array.
+  Report 'buffer full' condition with v4 ioctl as well as with v1.
+
+Version 1.01.04 - 2 Aug 2005
+============================
+  Fix dmsetup ls -j and status --target with empty table.
+
+Version 1.01.03 - 13 Jun 2005
+=============================
+  Use matchpathcon mode parameter.
+  Fix configure script to re-enable selinux.
+
+Version 1.01.02 - 17 May 2005
+=============================
+  Call dm_lib_exit() and dm_lib_release() automatically now.
+  Add --target <target_type> filter to dmsetup table/status/ls.
+  Add --exec <command> to dmsetup ls.
+  Fix dmsetup getopt_long usage.
+
+Version 1.01.01 - 29 Mar 2005
+=============================
+  Update dmsetup man page.
+  Drop-in devmap_name replacement.
+  Add option to compile without ioctl for testing.
+  Fix DM_LIB_VERSION sed.
+
+Version 1.01.00 - 17 Jan 2005
+=============================
+  Add dm_task_no_open_count() to skip getting open_count.
+
+Version 1.00.21 - 7 Jan 2005
+============================
+  Fix /proc/devices parsing.
+
+Version 1.00.20 - 6 Jan 2005
+============================
+  Attempt to fix /dev/mapper/control transparently if it's wrong.
+  Configuration-time option for setting uid/gid/mode for /dev/mapper nodes.
+  Update kernel patches for 2.4.27/2.4.28-pre-4 (includes minor fixes).
+  Add --noheadings columns option for colon-separated dmsetup output.
+  Support device referencing by uuid or major/minor.
+  Warn if kernel data didn't fit in buffer.
+  Fix a printf.
+
+Version 1.00.19 - 3 July 2004
+=============================
+  More autoconf fixes.
+  Fix a dmsetup newline.
+  Fix device number handling for 2.6 kernels.
+
+Version 1.00.18 - 20 Jun 2004
+=============================
+  Fix a uuid free in libdm-iface.
+  Fix a targets string size calc in driver.
+  Add -c to dmsetup for column-based output.
+  Add target message-passing ioctl.
+
+Version 1.00.17 - 17 Apr 2004
+=============================
+  configure --with-owner= --with-group= to avoid -o and -g args to 'install'
+  Fix library selinux linking.
+
+Version 1.00.16 - 16 Apr 2004
+=============================
+  Ignore error setting selinux file context if fs doesn't support it.
+
+Version 1.00.15 - 7 Apr 2004
+============================
+  Fix status overflow check in kernel patches.
+
+Version 1.00.14 - 6 Apr 2004
+============================
+  Fix static selinux build.
+
+Version 1.00.13 - 6 Apr 2004
+============================
+  Add some basic selinux support.
+
+Version 1.00.12 - 6 Apr 2004
+============================
+  Fix dmsetup.static install.
+
+Version 1.00.11 - 5 Apr 2004
+============================
+  configure --enable-static_link does static build in addition to dynamic.
+  Moved Makefile library targets definition into template.
+
+Version 1.00.10 - 2 Apr 2004
+============================
+  Fix DESTDIR handling.
+  Static build installs to dmsetup.static.
+  Basic support for internationalisation.
+  Minor Makefile tidy-ups/fixes.
+
+Version 1.00.09 - 31 Mar 2004
+=============================
+  Update copyright notices to Red Hat.
+  Move full mknodes functionality from dmsetup into libdevmapper.
+  Avoid sscanf %as for uClibc compatibility.
+  Cope if DM_LIST_VERSIONS is not defined.
+  Add DM_LIST_VERSIONS functionality to kernel patches.
+  Generate new kernel patches for 2.4.26-rc1.
+
+Version 1.00.08 - 27 Feb 2004
+=============================
+  Added 'dmsetup targets'.
+  Added event_nr support to 'dmsetup wait'.
+  Updated dmsetup man page.
+  Allow logging function to be reset to use internal one.
+  Bring log macros in line with LVM2 ones.
+  Added 'make install_static_lib' which installs libdevmapper.a.
+  Made configure/makefiles closer to LVM2 versions.
+  Fixed DESTDIR for make install/install_static_lib.
+  Updated README/INSTALL to reflect move to sources.redhat.com.
+  Updated autoconf files to 2003-06-17.

aclocal.m4

@@ -0,0 +1,171 @@
+# generated automatically by aclocal 1.11.1 -*- Autoconf -*-
+
+# Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
+# 2005, 2006, 2007, 2008, 2009  Free Software Foundation, Inc.
+# This file is free software; the Free Software Foundation
+# gives unlimited permission to copy and/or distribute it,
+# with or without modifications, as long as this notice is preserved.
+
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY, to the extent permitted by law; without
+# even the implied warranty of MERCHANTABILITY or FITNESS FOR A
+# PARTICULAR PURPOSE.
+
+# pkg.m4 - Macros to locate and utilise pkg-config.            -*- Autoconf -*-
+# serial 1 (pkg-config-0.24)
+# 
+# Copyright © 2004 Scott James Remnant <scott@netsplit.com>.
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+#
+# As a special exception to the GNU General Public License, if you
+# distribute this file as part of a program that contains a
+# configuration script generated by Autoconf, you may include it under
+# the same distribution terms that you use for the rest of that program.
+
+# PKG_PROG_PKG_CONFIG([MIN-VERSION])
+# ----------------------------------
+AC_DEFUN([PKG_PROG_PKG_CONFIG],
+[m4_pattern_forbid([^_?PKG_[A-Z_]+$])
+m4_pattern_allow([^PKG_CONFIG(_PATH)?$])
+AC_ARG_VAR([PKG_CONFIG], [path to pkg-config utility])
+AC_ARG_VAR([PKG_CONFIG_PATH], [directories to add to pkg-config's search path])
+AC_ARG_VAR([PKG_CONFIG_LIBDIR], [path overriding pkg-config's built-in search path])
+
+if test "x$ac_cv_env_PKG_CONFIG_set" != "xset"; then
+	AC_PATH_TOOL([PKG_CONFIG], [pkg-config])
+fi
+if test -n "$PKG_CONFIG"; then
+	_pkg_min_version=m4_default([$1], [0.9.0])
+	AC_MSG_CHECKING([pkg-config is at least version $_pkg_min_version])
+	if $PKG_CONFIG --atleast-pkgconfig-version $_pkg_min_version; then
+		AC_MSG_RESULT([yes])
+	else
+		AC_MSG_RESULT([no])
+		PKG_CONFIG=""
+	fi
+fi[]dnl
+])# PKG_PROG_PKG_CONFIG
+
+# PKG_CHECK_EXISTS(MODULES, [ACTION-IF-FOUND], [ACTION-IF-NOT-FOUND])
+#
+# Check to see whether a particular set of modules exists.  Similar
+# to PKG_CHECK_MODULES(), but does not set variables or print errors.
+#
+# Please remember that m4 expands AC_REQUIRE([PKG_PROG_PKG_CONFIG])
+# only at the first occurence in configure.ac, so if the first place
+# it's called might be skipped (such as if it is within an "if", you
+# have to call PKG_CHECK_EXISTS manually
+# --------------------------------------------------------------
+AC_DEFUN([PKG_CHECK_EXISTS],
+[AC_REQUIRE([PKG_PROG_PKG_CONFIG])dnl
+if test -n "$PKG_CONFIG" && \
+    AC_RUN_LOG([$PKG_CONFIG --exists --print-errors "$1"]); then
+  m4_default([$2], [:])
+m4_ifvaln([$3], [else
+  $3])dnl
+fi])
+
+# _PKG_CONFIG([VARIABLE], [COMMAND], [MODULES])
+# ---------------------------------------------
+m4_define([_PKG_CONFIG],
+[if test -n "$$1"; then
+    pkg_cv_[]$1="$$1"
+ elif test -n "$PKG_CONFIG"; then
+    PKG_CHECK_EXISTS([$3],
+                     [pkg_cv_[]$1=`$PKG_CONFIG --[]$2 "$3" 2>/dev/null`],
+		     [pkg_failed=yes])
+ else
+    pkg_failed=untried
+fi[]dnl
+])# _PKG_CONFIG
+
+# _PKG_SHORT_ERRORS_SUPPORTED
+# -----------------------------
+AC_DEFUN([_PKG_SHORT_ERRORS_SUPPORTED],
+[AC_REQUIRE([PKG_PROG_PKG_CONFIG])
+if $PKG_CONFIG --atleast-pkgconfig-version 0.20; then
+        _pkg_short_errors_supported=yes
+else
+        _pkg_short_errors_supported=no
+fi[]dnl
+])# _PKG_SHORT_ERRORS_SUPPORTED
+
+
+# PKG_CHECK_MODULES(VARIABLE-PREFIX, MODULES, [ACTION-IF-FOUND],
+# [ACTION-IF-NOT-FOUND])
+#
+#
+# Note that if there is a possibility the first call to
+# PKG_CHECK_MODULES might not happen, you should be sure to include an
+# explicit call to PKG_PROG_PKG_CONFIG in your configure.ac
+#
+#
+# --------------------------------------------------------------
+AC_DEFUN([PKG_CHECK_MODULES],
+[AC_REQUIRE([PKG_PROG_PKG_CONFIG])dnl
+AC_ARG_VAR([$1][_CFLAGS], [C compiler flags for $1, overriding pkg-config])dnl
+AC_ARG_VAR([$1][_LIBS], [linker flags for $1, overriding pkg-config])dnl
+
+pkg_failed=no
+AC_MSG_CHECKING([for $1])
+
+_PKG_CONFIG([$1][_CFLAGS], [cflags], [$2])
+_PKG_CONFIG([$1][_LIBS], [libs], [$2])
+
+m4_define([_PKG_TEXT], [Alternatively, you may set the environment variables $1[]_CFLAGS
+and $1[]_LIBS to avoid the need to call pkg-config.
+See the pkg-config man page for more details.])
+
+if test $pkg_failed = yes; then
+   	AC_MSG_RESULT([no])
+        _PKG_SHORT_ERRORS_SUPPORTED
+        if test $_pkg_short_errors_supported = yes; then
+	        $1[]_PKG_ERRORS=`$PKG_CONFIG --short-errors --print-errors "$2" 2>&1`
+        else 
+	        $1[]_PKG_ERRORS=`$PKG_CONFIG --print-errors "$2" 2>&1`
+        fi
+	# Put the nasty error message in config.log where it belongs
+	echo "$$1[]_PKG_ERRORS" >&AS_MESSAGE_LOG_FD
+
+	m4_default([$4], [AC_MSG_ERROR(
+[Package requirements ($2) were not met:
+
+$$1_PKG_ERRORS
+
+Consider adjusting the PKG_CONFIG_PATH environment variable if you
+installed software in a non-standard prefix.
+
+_PKG_TEXT])
+        ])
+elif test $pkg_failed = untried; then
+     	AC_MSG_RESULT([no])
+	m4_default([$4], [AC_MSG_FAILURE(
+[The pkg-config script could not be found or is too old.  Make sure it
+is in your PATH or set the PKG_CONFIG environment variable to the full
+path to pkg-config.
+
+_PKG_TEXT
+
+To get pkg-config, see <http://pkg-config.freedesktop.org/>.])
+        ])
+else
+	$1[]_CFLAGS=$pkg_cv_[]$1[]_CFLAGS
+	$1[]_LIBS=$pkg_cv_[]$1[]_LIBS
+        AC_MSG_RESULT([yes])
+	$3
+fi[]dnl
+])# PKG_CHECK_MODULES
+

autoconf/install-sh

@@ -1,36 +1,62 @@
 #!/bin/sh
-#
 # install - install a program, script, or datafile
-# This comes from X11R5 (mit/util/scripts/install.sh).
+
+scriptversion=2006-10-14.15
+
+# This originates from X11R5 (mit/util/scripts/install.sh), which was
+# later released in X11R6 (xc/config/util/install.sh) with the
+# following copyright and license.
 #
-# Copyright 1991 by the Massachusetts Institute of Technology
+# Copyright (C) 1994 X Consortium
 #
-# Permission to use, copy, modify, distribute, and sell this software and its
-# documentation for any purpose is hereby granted without fee, provided that
-# the above copyright notice appear in all copies and that both that
-# copyright notice and this permission notice appear in supporting
-# documentation, and that the name of M.I.T. not be used in advertising or
-# publicity pertaining to distribution of the software without specific,
-# written prior permission.  M.I.T. makes no representations about the
-# suitability of this software for any purpose.  It is provided "as is"
-# without express or implied warranty.
+# Permission is hereby granted, free of charge, to any person obtaining a copy
+# of this software and associated documentation files (the "Software"), to
+# deal in the Software without restriction, including without limitation the
+# rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+# sell copies of the Software, and to permit persons to whom the Software is
+# furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in
+# all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL THE
+# X CONSORTIUM BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+# AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNEC-
+# TION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+#
+# Except as contained in this notice, the name of the X Consortium shall not
+# be used in advertising or otherwise to promote the sale, use or other deal-
+# ings in this Software without prior written authorization from the X Consor-
+# tium.
+#
+#
+# FSF changes to this file are in the public domain.
 #
 # Calling this script install-sh is preferred over install.sh, to prevent
 # `make' implicit rules from creating a file called install from it
 # when there is no Makefile.
 #
 # This script is compatible with the BSD install script, but was written
-# from scratch.  It can only install one file at a time, a restriction
-# shared with many OS's install programs.
+# from scratch.
 
+nl='
+'
+IFS=" ""	$nl"
 
 # set DOITPROG to echo to test this script
 
 # Don't use :- since 4.3BSD and earlier shells don't like it.
 doit="${DOITPROG-}"
+if test -z "$doit"; then
+  doit_exec=exec
+else
+  doit_exec=$doit
+fi
 
-
-# put in absolute paths if you don't have them in your path; or use env. vars.
+# Put in absolute file names if you don't have them in your path;
+# or use environment vars.
 
 mvprog="${MVPROG-mv}"
 cpprog="${CPPROG-cp}"
@@ -41,211 +67,441 @@ stripprog="${STRIPPROG-strip}"
 rmprog="${RMPROG-rm}"
 mkdirprog="${MKDIRPROG-mkdir}"
 
-transformbasename=""
-transform_arg=""
-instcmd="$mvprog"
-chmodcmd="$chmodprog 0755"
-chowncmd=""
-chgrpcmd=""
-stripcmd=""
+posix_glob=
+posix_mkdir=
+
+# Desired mode of installed file.
+mode=0755
+
+chmodcmd=$chmodprog
+chowncmd=
+chgrpcmd=
+stripcmd=
 rmcmd="$rmprog -f"
 mvcmd="$mvprog"
-src=""
-dst=""
-dir_arg=""
+src=
+dst=
+dir_arg=
+dstarg=
+no_target_directory=
 
-while [ x"$1" != x ]; do
+usage="Usage: $0 [OPTION]... [-T] SRCFILE DSTFILE
+   or: $0 [OPTION]... SRCFILES... DIRECTORY
+   or: $0 [OPTION]... -t DIRECTORY SRCFILES...
+   or: $0 [OPTION]... -d DIRECTORIES...
+
+In the 1st form, copy SRCFILE to DSTFILE.
+In the 2nd and 3rd, copy all SRCFILES to DIRECTORY.
+In the 4th, create DIRECTORIES.
+
+Options:
+-c         (ignored)
+-d         create directories instead of installing files.
+-g GROUP   $chgrpprog installed files to GROUP.
+-m MODE    $chmodprog installed files to MODE.
+-o USER    $chownprog installed files to USER.
+-s         $stripprog installed files.
+-t DIRECTORY  install into DIRECTORY.
+-T         report an error if DSTFILE is a directory.
+--help     display this help and exit.
+--version  display version info and exit.
+
+Environment variables override the default commands:
+  CHGRPPROG CHMODPROG CHOWNPROG CPPROG MKDIRPROG MVPROG RMPROG STRIPPROG
+"
+
+while test $# -ne 0; do
   case $1 in
-	-c) instcmd="$cpprog"
-	    shift
+    -c) shift
         continue;;
 
     -d) dir_arg=true
         shift
         continue;;
 
-	-m) chmodcmd="$chmodprog $2"
-	    shift
-	    shift
-	    continue;;
-
-	-o) chowncmd="$chownprog $2"
-	    shift
-	    shift
-	    continue;;
-
     -g) chgrpcmd="$chgrpprog $2"
         shift
         shift
         continue;;
 
-	-s) stripcmd="$stripprog"
+    --help) echo "$usage"; exit $?;;
+
+    -m) mode=$2
+        shift
+        shift
+	case $mode in
+	  *' '* | *'	'* | *'
+'*	  | *'*'* | *'?'* | *'['*)
+	    echo "$0: invalid mode: $mode" >&2
+	    exit 1;;
+	esac
+        continue;;
+
+    -o) chowncmd="$chownprog $2"
+        shift
         shift
         continue;;
 
-	-t=*) transformarg=`echo $1 | sed 's/-t=//'`
+    -s) stripcmd=$stripprog
         shift
         continue;;
 
-	-b=*) transformbasename=`echo $1 | sed 's/-b=//'`
+    -t) dstarg=$2
+	shift
 	shift
 	continue;;
 
-	*)  if [ x"$src" = x ]
-	    then
-		src=$1
-	    else
-		# this colon is to work around a 386BSD /bin/sh bug
-		:
-		dst=$1
-	    fi
+    -T) no_target_directory=true
 	shift
 	continue;;
+
+    --version) echo "$0 $scriptversion"; exit $?;;
+
+    --)	shift
+	break;;
+
+    -*)	echo "$0: invalid option: $1" >&2
+	exit 1;;
+
+    *)  break;;
   esac
 done
 
-if [ x"$src" = x ]
-then
-	echo "install:	no input file specified"
-	exit 1
-else
-	true
+if test $# -ne 0 && test -z "$dir_arg$dstarg"; then
+  # When -d is used, all remaining arguments are directories to create.
+  # When -t is used, the destination is already specified.
+  # Otherwise, the last argument is the destination.  Remove it from $@.
+  for arg
+  do
+    if test -n "$dstarg"; then
+      # $@ is not empty: it contains at least $arg.
+      set fnord "$@" "$dstarg"
+      shift # fnord
+    fi
+    shift # arg
+    dstarg=$arg
+  done
 fi
 
-if [ x"$dir_arg" != x ]; then
+if test $# -eq 0; then
+  if test -z "$dir_arg"; then
+    echo "$0: no input file specified." >&2
+    exit 1
+  fi
+  # It's OK to call `install-sh -d' without argument.
+  # This can happen when creating conditional directories.
+  exit 0
+fi
+
+if test -z "$dir_arg"; then
+  trap '(exit $?); exit' 1 2 13 15
+
+  # Set umask so as not to create temps with too-generous modes.
+  # However, 'strip' requires both read and write access to temps.
+  case $mode in
+    # Optimize common cases.
+    *644) cp_umask=133;;
+    *755) cp_umask=22;;
+
+    *[0-7])
+      if test -z "$stripcmd"; then
+	u_plus_rw=
+      else
+	u_plus_rw='% 200'
+      fi
+      cp_umask=`expr '(' 777 - $mode % 1000 ')' $u_plus_rw`;;
+    *)
+      if test -z "$stripcmd"; then
+	u_plus_rw=
+      else
+	u_plus_rw=,u+rw
+      fi
+      cp_umask=$mode$u_plus_rw;;
+  esac
+fi
+
+for src
+do
+  # Protect names starting with `-'.
+  case $src in
+    -*) src=./$src ;;
+  esac
+
+  if test -n "$dir_arg"; then
     dst=$src
-	src=""
-	
-	if [ -d $dst ]; then
-		instcmd=:
-		chmodcmd=""
+    dstdir=$dst
+    test -d "$dstdir"
+    dstdir_status=$?
   else
-		instcmd=mkdir
-	fi
-else
 
-# Waiting for this to be detected by the "$instcmd $src $dsttmp" command
-# might cause directories to be created, which would be especially bad 
-# if $src (and thus $dsttmp) contains '*'.
-
-	if [ -f $src -o -d $src ]
-	then
-		true
-	else
-		echo "install:  $src does not exist"
+    # Waiting for this to be detected by the "$cpprog $src $dsttmp" command
+    # might cause directories to be created, which would be especially bad
+    # if $src (and thus $dsttmp) contains '*'.
+    if test ! -f "$src" && test ! -d "$src"; then
+      echo "$0: $src does not exist." >&2
       exit 1
     fi
 
-	if [ x"$dst" = x ]
-	then
-		echo "install:	no destination specified"
+    if test -z "$dstarg"; then
+      echo "$0: no destination specified." >&2
       exit 1
-	else
-		true
     fi
 
-# If destination is a directory, append the input filename; if your system
-# does not like double slashes in filenames, you may need to add some logic
+    dst=$dstarg
+    # Protect names starting with `-'.
+    case $dst in
+      -*) dst=./$dst ;;
+    esac
 
-	if [ -d $dst ]
+    # If destination is a directory, append the input filename; won't work
+    # if double slashes aren't ignored.
+    if test -d "$dst"; then
+      if test -n "$no_target_directory"; then
+	echo "$0: $dstarg: Is a directory" >&2
+	exit 1
+      fi
+      dstdir=$dst
+      dst=$dstdir/`basename "$src"`
+      dstdir_status=0
+    else
+      # Prefer dirname, but fall back on a substitute if dirname fails.
+      dstdir=`
+	(dirname "$dst") 2>/dev/null ||
+	expr X"$dst" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \
+	     X"$dst" : 'X\(//\)[^/]' \| \
+	     X"$dst" : 'X\(//\)$' \| \
+	     X"$dst" : 'X\(/\)' \| . 2>/dev/null ||
+	echo X"$dst" |
+	    sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{
+		   s//\1/
+		   q
+		 }
+		 /^X\(\/\/\)[^/].*/{
+		   s//\1/
+		   q
+		 }
+		 /^X\(\/\/\)$/{
+		   s//\1/
+		   q
+		 }
+		 /^X\(\/\).*/{
+		   s//\1/
+		   q
+		 }
+		 s/.*/./; q'
+      `
+
+      test -d "$dstdir"
+      dstdir_status=$?
+    fi
+  fi
+
+  obsolete_mkdir_used=false
+
+  if test $dstdir_status != 0; then
+    case $posix_mkdir in
+      '')
+	# Create intermediate dirs using mode 755 as modified by the umask.
+	# This is like FreeBSD 'install' as of 1997-10-28.
+	umask=`umask`
+	case $stripcmd.$umask in
+	  # Optimize common cases.
+	  *[2367][2367]) mkdir_umask=$umask;;
+	  .*0[02][02] | .[02][02] | .[02]) mkdir_umask=22;;
+
+	  *[0-7])
+	    mkdir_umask=`expr $umask + 22 \
+	      - $umask % 100 % 40 + $umask % 20 \
+	      - $umask % 10 % 4 + $umask % 2
+	    `;;
+	  *) mkdir_umask=$umask,go-w;;
+	esac
+
+	# With -d, create the new directory with the user-specified mode.
+	# Otherwise, rely on $mkdir_umask.
+	if test -n "$dir_arg"; then
+	  mkdir_mode=-m$mode
+	else
+	  mkdir_mode=
+	fi
+
+	posix_mkdir=false
+	case $umask in
+	  *[123567][0-7][0-7])
+	    # POSIX mkdir -p sets u+wx bits regardless of umask, which
+	    # is incompatible with FreeBSD 'install' when (umask & 300) != 0.
+	    ;;
+	  *)
+	    tmpdir=${TMPDIR-/tmp}/ins$RANDOM-$$
+	    trap 'ret=$?; rmdir "$tmpdir/d" "$tmpdir" 2>/dev/null; exit $ret' 0
+
+	    if (umask $mkdir_umask &&
+		exec $mkdirprog $mkdir_mode -p -- "$tmpdir/d") >/dev/null 2>&1
 	    then
-		dst="$dst"/`basename $src`
-	else
-		true
+	      if test -z "$dir_arg" || {
+		   # Check for POSIX incompatibilities with -m.
+		   # HP-UX 11.23 and IRIX 6.5 mkdir -m -p sets group- or
+		   # other-writeable bit of parent directory when it shouldn't.
+		   # FreeBSD 6.1 mkdir -m -p sets mode of existing directory.
+		   ls_ld_tmpdir=`ls -ld "$tmpdir"`
+		   case $ls_ld_tmpdir in
+		     d????-?r-*) different_mode=700;;
+		     d????-?--*) different_mode=755;;
+		     *) false;;
+		   esac &&
+		   $mkdirprog -m$different_mode -p -- "$tmpdir" && {
+		     ls_ld_tmpdir_1=`ls -ld "$tmpdir"`
+		     test "$ls_ld_tmpdir" = "$ls_ld_tmpdir_1"
+		   }
+		 }
+	      then posix_mkdir=:
 	      fi
-fi
+	      rmdir "$tmpdir/d" "$tmpdir"
+	    else
+	      # Remove any dirs left behind by ancient mkdir implementations.
+	      rmdir ./$mkdir_mode ./-p ./-- 2>/dev/null
+	    fi
+	    trap '' 0;;
+	esac;;
+    esac
 
-## this sed command emulates the dirname command
-dstdir=`echo $dst | sed -e 's,[^/]*$,,;s,/$,,;s,^$,.,'`
+    if
+      $posix_mkdir && (
+	umask $mkdir_umask &&
+	$doit_exec $mkdirprog $mkdir_mode -p -- "$dstdir"
+      )
+    then :
+    else
 
-# Make sure that the destination directory exists.
-#  this part is taken from Noah Friedman's mkinstalldirs script
+      # The umask is ridiculous, or mkdir does not conform to POSIX,
+      # or it failed possibly due to a race condition.  Create the
+      # directory the slow way, step by step, checking for races as we go.
 
-# Skip lots of stat calls in the usual case.
-if [ ! -d "$dstdir" ]; then
-defaultIFS='	
-'
-IFS="${IFS-${defaultIFS}}"
+      case $dstdir in
+	/*) prefix=/ ;;
+	-*) prefix=./ ;;
+	*)  prefix= ;;
+      esac
 
-oIFS="${IFS}"
-# Some sh's can't handle IFS=/ for some reason.
-IFS='%'
-set - `echo ${dstdir} | sed -e 's@/@%@g' -e 's@^%@/@'`
-IFS="${oIFS}"
+      case $posix_glob in
+        '')
+	  if (set -f) 2>/dev/null; then
+	    posix_glob=true
+	  else
+	    posix_glob=false
+	  fi ;;
+      esac
 
-pathcomp=''
-
-while [ $# -ne 0 ] ; do
-	pathcomp="${pathcomp}${1}"
+      oIFS=$IFS
+      IFS=/
+      $posix_glob && set -f
+      set fnord $dstdir
       shift
+      $posix_glob && set +f
+      IFS=$oIFS
 
-	if [ ! -d "${pathcomp}" ] ;
-        then
-		$mkdirprog "${pathcomp}"
+      prefixes=
+
+      for d
+      do
+	test -z "$d" && continue
+
+	prefix=$prefix$d
+	if test -d "$prefix"; then
+	  prefixes=
 	else
-		true
+	  if $posix_mkdir; then
+	    (umask=$mkdir_umask &&
+	     $doit_exec $mkdirprog $mkdir_mode -p -- "$dstdir") && break
+	    # Don't fail if two instances are running concurrently.
+	    test -d "$prefix" || exit 1
+	  else
+	    case $prefix in
+	      *\'*) qprefix=`echo "$prefix" | sed "s/'/'\\\\\\\\''/g"`;;
+	      *) qprefix=$prefix;;
+	    esac
+	    prefixes="$prefixes '$qprefix'"
+	  fi
+	fi
+	prefix=$prefix/
+      done
+
+      if test -n "$prefixes"; then
+	# Don't fail if two instances are running concurrently.
+	(umask $mkdir_umask &&
+	 eval "\$doit_exec \$mkdirprog $prefixes") ||
+	  test -d "$dstdir" || exit 1
+	obsolete_mkdir_used=true
+      fi
+    fi
   fi
 
-	pathcomp="${pathcomp}/"
+  if test -n "$dir_arg"; then
+    { test -z "$chowncmd" || $doit $chowncmd "$dst"; } &&
+    { test -z "$chgrpcmd" || $doit $chgrpcmd "$dst"; } &&
+    { test "$obsolete_mkdir_used$chowncmd$chgrpcmd" = false ||
+      test -z "$chmodcmd" || $doit $chmodcmd $mode "$dst"; } || exit 1
+  else
+
+    # Make a couple of temp file names in the proper directory.
+    dsttmp=$dstdir/_inst.$$_
+    rmtmp=$dstdir/_rm.$$_
+
+    # Trap to clean up those temp files at exit.
+    trap 'ret=$?; rm -f "$dsttmp" "$rmtmp" && exit $ret' 0
+
+    # Copy the file name to the temp name.
+    (umask $cp_umask && $doit_exec $cpprog "$src" "$dsttmp") &&
+
+    # and set any options; do chmod last to preserve setuid bits.
+    #
+    # If any of these fail, we abort the whole thing.  If we want to
+    # ignore errors from any of these, just make sure not to ignore
+    # errors from the above "$doit $cpprog $src $dsttmp" command.
+    #
+    { test -z "$chowncmd" || $doit $chowncmd "$dsttmp"; } \
+      && { test -z "$chgrpcmd" || $doit $chgrpcmd "$dsttmp"; } \
+      && { test -z "$stripcmd" || $doit $stripcmd "$dsttmp"; } \
+      && { test -z "$chmodcmd" || $doit $chmodcmd $mode "$dsttmp"; } &&
+
+    # Now rename the file to the real destination.
+    { $doit $mvcmd -f "$dsttmp" "$dst" 2>/dev/null \
+      || {
+	   # The rename failed, perhaps because mv can't rename something else
+	   # to itself, or perhaps because mv is so ancient that it does not
+	   # support -f.
+
+	   # Now remove or move aside any old file at destination location.
+	   # We try this two ways since rm can't unlink itself on some
+	   # systems and the destination file might be busy for other
+	   # reasons.  In this case, the final cleanup might fail but the new
+	   # file should still install successfully.
+	   {
+	     if test -f "$dst"; then
+	       $doit $rmcmd -f "$dst" 2>/dev/null \
+	       || { $doit $mvcmd -f "$dst" "$rmtmp" 2>/dev/null \
+		     && { $doit $rmcmd -f "$rmtmp" 2>/dev/null; :; }; }\
+	       || {
+		 echo "$0: cannot unlink or rename $dst" >&2
+		 (exit 1); exit 1
+	       }
+	     else
+	       :
+	     fi
+	   } &&
+
+	   # Now rename the file to the real destination.
+	   $doit $mvcmd "$dsttmp" "$dst"
+	 }
+    } || exit 1
+
+    trap '' 0
+  fi
 done
-fi
 
-if [ x"$dir_arg" != x ]
-then
-	$doit $instcmd $dst &&
-
-	if [ x"$chowncmd" != x ]; then $doit $chowncmd $dst; else true ; fi &&
-	if [ x"$chgrpcmd" != x ]; then $doit $chgrpcmd $dst; else true ; fi &&
-	if [ x"$stripcmd" != x ]; then $doit $stripcmd $dst; else true ; fi &&
-	if [ x"$chmodcmd" != x ]; then $doit $chmodcmd $dst; else true ; fi
-else
-
-# If we're going to rename the final executable, determine the name now.
-
-	if [ x"$transformarg" = x ] 
-	then
-		dstfile=`basename $dst`
-	else
-		dstfile=`basename $dst $transformbasename | 
-			sed $transformarg`$transformbasename
-	fi
-
-# don't allow the sed command to completely eliminate the filename
-
-	if [ x"$dstfile" = x ] 
-	then
-		dstfile=`basename $dst`
-	else
-		true
-	fi
-
-# Make a temp file name in the proper directory.
-
-	dsttmp=$dstdir/#inst.$$#
-
-# Move or copy the file name to the temp name
-
-	$doit $instcmd $src $dsttmp &&
-
-	trap "rm -f ${dsttmp}" 0 &&
-
-# and set any options; do chmod last to preserve setuid bits
-
-# If any of these fail, we abort the whole thing.  If we want to
-# ignore errors from any of these, just make sure not to ignore
-# errors from the above "$doit $instcmd $src $dsttmp" command.
-
-	if [ x"$chowncmd" != x ]; then $doit $chowncmd $dsttmp; else true;fi &&
-	if [ x"$chgrpcmd" != x ]; then $doit $chgrpcmd $dsttmp; else true;fi &&
-	if [ x"$stripcmd" != x ]; then $doit $stripcmd $dsttmp; else true;fi &&
-	if [ x"$chmodcmd" != x ]; then $doit $chmodcmd $dsttmp; else true;fi &&
-
-# Now rename the file to the real destination.
-
-	$doit $rmcmd -f $dstdir/$dstfile &&
-	$doit $mvcmd $dsttmp $dstdir/$dstfile 
-
-fi &&
-
-
-exit 0
+# Local variables:
+# eval: (add-hook 'write-file-hooks 'time-stamp)
+# time-stamp-start: "scriptversion="
+# time-stamp-format: "%:y-%02m-%02d.%02H"
+# time-stamp-end: "$"
+# End:

daemons/Makefile.in

@@ -0,0 +1,47 @@
+#
+# Copyright (C) 2004-2011 Red Hat, Inc. All rights reserved.
+#
+# This file is part of LVM2.
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions
+# of the GNU General Public License v.2.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+srcdir = @srcdir@
+top_srcdir = @top_srcdir@
+top_builddir = @top_builddir@
+
+ifeq ("@BUILD_LVMETAD@", "yes")
+  SUBDIRS += lvmetad
+endif
+
+.PHONY: dmeventd clvmd cmirrord lvmetad
+
+ifneq ("@CLVMD@", "none")
+  SUBDIRS += clvmd
+endif
+
+ifeq ("@BUILD_CMIRRORD@", "yes")
+  SUBDIRS += cmirrord
+endif
+
+ifeq ("@BUILD_DMEVENTD@", "yes")
+  SUBDIRS += dmeventd
+ifneq ("$(CFLOW_CMD)", "")
+daemons.cflow: dmeventd.cflow
+endif
+endif
+
+ifeq ($(MAKECMDGOALS),distclean)
+  SUBDIRS = clvmd cmirrord dmeventd lvmetad
+endif
+
+include $(top_builddir)/make.tmpl
+
+ifeq ("@BUILD_DMEVENTD@", "yes")
+device-mapper: dmeventd.device-mapper
+endif

daemons/clvmd/Makefile.in

@@ -0,0 +1,107 @@
+#
+# Copyright (C) 2004 Red Hat, Inc. All rights reserved.
+#
+# This file is part of LVM2.
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions
+# of the GNU General Public License v.2.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+srcdir = @srcdir@
+top_srcdir = @top_srcdir@
+top_builddir = @top_builddir@
+
+CMAN_LIBS = @CMAN_LIBS@
+CMAN_CFLAGS = @CMAN_CFLAGS@
+CMAP_LIBS = @CMAP_LIBS@
+CMAP_CFLAGS = @CMAP_CFLAGS@
+CONFDB_LIBS = @CONFDB_LIBS@
+CONFDB_CFLAGS = @CONFDB_CFLAGS@
+CPG_LIBS = @CPG_LIBS@
+CPG_CFLAGS = @CPG_CFLAGS@
+DLM_LIBS = @DLM_LIBS@
+DLM_CFLAGS = @DLM_CFLAGS@
+QUORUM_LIBS = @QUORUM_LIBS@
+QUORUM_CFLAGS = @QUORUM_CFLAGS@
+SALCK_LIBS = @SALCK_LIBS@
+SALCK_CFLAGS = @SALCK_CFLAGS@
+
+SOURCES = \
+	clvmd-command.c  \
+	clvmd.c          \
+	lvm-functions.c  \
+	refresh_clvmd.c
+
+ifeq ("@DEBUG@", "yes")
+	DEFS += -DDEBUG
+endif
+
+ifneq (,$(findstring cman,, "@CLVMD@,"))
+	SOURCES += clvmd-cman.c
+	LMLIBS += $(CMAN_LIBS) $(CONFDB_LIBS) $(DLM_LIBS)
+	CFLAGS += $(CMAN_CFLAGS) $(CONFDB_CFLAGS) $(DLM_CFLAGS)
+	DEFS += -DUSE_CMAN
+endif
+
+ifneq (,$(findstring openais,, "@CLVMD@,"))
+	SOURCES += clvmd-openais.c
+	LMLIBS += $(CONFDB_LIBS) $(CPG_LIBS) $(SALCK_LIBS)
+	CFLAGS += $(CONFDB_CFLAGS) $(CPG_CFLAGS) $(SALCK_CFLAGS)
+	DEFS += -DUSE_OPENAIS
+endif
+
+ifneq (,$(findstring corosync,, "@CLVMD@,"))
+	SOURCES += clvmd-corosync.c
+	LMLIBS += $(CMAP_LIBS) $(CONFDB_LIBS) $(CPG_LIBS) $(DLM_LIBS) $(QUORUM_LIBS)
+	CFLAGS += $(CMAP_CFLAGS) $(CONFDB_CFLAGS) $(CPG_CFLAGS) $(DLM_CFLAGS) $(QUORUM_CFLAGS)
+	DEFS += -DUSE_COROSYNC
+endif
+
+ifneq (,$(findstring singlenode,, "@CLVMD@,"))
+	SOURCES += clvmd-singlenode.c
+	DEFS += -DUSE_SINGLENODE
+endif
+
+ifeq ($(MAKECMDGOALS),distclean)
+	SOURCES += clvmd-cman.c
+	SOURCES += clvmd-openais.c
+	SOURCES += clvmd-corosync.c
+	SOURCES += clvmd-singlenode.c
+endif
+
+TARGETS = \
+	clvmd
+
+LVMLIBS = $(LVMINTERNAL_LIBS)
+
+ifeq ("@DMEVENTD@", "yes")
+	LVMLIBS += -ldevmapper-event
+endif
+ 
+include $(top_builddir)/make.tmpl
+
+LVMLIBS += -ldevmapper
+LIBS += $(PTHREAD_LIBS)
+
+DEFS += -D_REENTRANT
+CFLAGS += -fno-strict-aliasing
+
+INSTALL_TARGETS = \
+	install_clvmd
+
+clvmd: $(OBJECTS) $(top_builddir)/lib/liblvm-internal.a
+	$(CC) $(CFLAGS) $(LDFLAGS) -o clvmd $(OBJECTS) \
+		$(LVMLIBS) $(LMLIBS) $(LIBS)
+
+.PHONY: install_clvmd
+
+install_clvmd: $(TARGETS)
+	$(INSTALL_PROGRAM) -D clvmd $(usrsbindir)/clvmd
+
+install: $(INSTALL_TARGETS)
+
+install_cluster: $(INSTALL_TARGETS)

daemons/clvmd/clvm.h

@@ -0,0 +1,82 @@
+/*
+ * Copyright (C) 2002-2004 Sistina Software, Inc. All rights reserved.
+ * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU General Public License v.2.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+/* Definitions for CLVMD server and clients */
+
+/*
+ * The protocol spoken over the cluster and across the local socket.
+ */
+
+#ifndef _CLVM_H
+#define _CLVM_H
+
+#include "configure.h"
+
+struct clvm_header {
+	uint8_t  cmd;	        /* See below */
+	uint8_t  flags;	        /* See below */
+	uint16_t xid;	        /* Transaction ID */
+	uint32_t clientid;	/* Only used in Daemon->Daemon comms */
+	int32_t  status;	/* For replies, whether request succeeded */
+	uint32_t arglen;	/* Length of argument below.
+				   If >1500 then it will be passed
+				   around the cluster in the system LV */
+	char node[1];		/* Actually a NUL-terminated string, node name.
+				   If this is empty then the command is
+				   forwarded to all cluster nodes unless
+				   FLAG_LOCAL or FLAG_REMOTE is also set. */
+	char args[1];		/* Arguments for the command follow the
+				   node name, This member is only
+				   valid if the node name is empty */
+} __attribute__ ((packed));
+
+/* Flags */
+#define CLVMD_FLAG_LOCAL	1	/* Only do this on the local node */
+#define CLVMD_FLAG_SYSTEMLV	2	/* Data in system LV under my node name */
+#define CLVMD_FLAG_NODEERRS	4	/* Reply has errors in node-specific portion */
+#define CLVMD_FLAG_REMOTE	8	/* Do this on all nodes except for the local node */
+
+/* Name of the local socket to communicate between lvm and clvmd */
+static const char CLVMD_SOCKNAME[]= DEFAULT_RUN_DIR "/clvmd.sock";
+
+/* Internal commands & replies */
+#define CLVMD_CMD_REPLY    1
+#define CLVMD_CMD_VERSION  2	/* Send version around cluster when we start */
+#define CLVMD_CMD_GOAWAY   3	/* Die if received this - we are running
+				   an incompatible version */
+#define CLVMD_CMD_TEST     4	/* Just for mucking about */
+
+#define CLVMD_CMD_LOCK              30
+#define CLVMD_CMD_UNLOCK            31
+
+/* Lock/Unlock commands */
+#define CLVMD_CMD_LOCK_LV           50
+#define CLVMD_CMD_LOCK_VG           51
+#define CLVMD_CMD_LOCK_QUERY	    52
+
+/* Misc functions */
+#define CLVMD_CMD_REFRESH	    40
+#define CLVMD_CMD_GET_CLUSTERNAME   41
+#define CLVMD_CMD_SET_DEBUG	    42
+#define CLVMD_CMD_VG_BACKUP	    43
+#define CLVMD_CMD_RESTART	    44
+#define CLVMD_CMD_SYNC_NAMES	    45
+
+/* Used internally by some callers, but not part of the protocol.*/
+#define NODE_ALL	"*"
+#define NODE_LOCAL	"."
+#define NODE_REMOTE	"^"
+
+#endif

daemons/clvmd/clvmd-cman.c

@@ -0,0 +1,506 @@
+/*
+ * Copyright (C) 2002-2004 Sistina Software, Inc. All rights reserved.
+ * Copyright (C) 2004 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU General Public License v.2.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+/*
+ * CMAN communication layer for clvmd.
+ */
+
+#include "clvmd-common.h"
+
+#include <pthread.h>
+
+#include "clvmd-comms.h"
+#include "clvm.h"
+#include "clvmd.h"
+#include "lvm-functions.h"
+
+#include <libdlm.h>
+
+#include <syslog.h>
+
+#define LOCKSPACE_NAME "clvmd"
+
+struct clvmd_node
+{
+	struct cman_node *node;
+	int clvmd_up;
+};
+
+static int num_nodes;
+static struct cman_node *nodes = NULL;
+static struct cman_node this_node;
+static int count_nodes; /* size of allocated nodes array */
+static struct dm_hash_table *node_updown_hash;
+static dlm_lshandle_t *lockspace;
+static cman_handle_t c_handle;
+
+static void count_clvmds_running(void);
+static void get_members(void);
+static int nodeid_from_csid(const char *csid);
+static int name_from_nodeid(int nodeid, char *name);
+static void event_callback(cman_handle_t handle, void *private, int reason, int arg);
+static void data_callback(cman_handle_t handle, void *private,
+			  char *buf, int len, uint8_t port, int nodeid);
+
+struct lock_wait {
+	pthread_cond_t cond;
+	pthread_mutex_t mutex;
+	struct dlm_lksb lksb;
+};
+
+static int _init_cluster(void)
+{
+	node_updown_hash = dm_hash_create(100);
+
+	/* Open the cluster communication socket */
+	c_handle = cman_init(NULL);
+	if (!c_handle) {
+		syslog(LOG_ERR, "Can't open cluster manager socket: %m");
+		return -1;
+	}
+	DEBUGLOG("Connected to CMAN\n");
+
+	if (cman_start_recv_data(c_handle, data_callback, CLUSTER_PORT_CLVMD)) {
+		syslog(LOG_ERR, "Can't bind cluster socket: %m");
+		return -1;
+	}
+
+	if (cman_start_notification(c_handle, event_callback)) {
+		syslog(LOG_ERR, "Can't start cluster event listening");
+		return -1;
+	}
+
+	/* Get the cluster members list */
+	get_members();
+	count_clvmds_running();
+
+	DEBUGLOG("CMAN initialisation complete\n");
+
+	/* Create a lockspace for LV & VG locks to live in */
+	lockspace = dlm_open_lockspace(LOCKSPACE_NAME);
+	if (!lockspace) {
+		lockspace = dlm_create_lockspace(LOCKSPACE_NAME, 0600);
+		if (!lockspace) {
+			syslog(LOG_ERR, "Unable to create DLM lockspace for CLVM: %m");
+			return -1;
+		}
+		DEBUGLOG("Created DLM lockspace for CLVMD.\n");
+	} else
+		DEBUGLOG("Opened existing DLM lockspace for CLVMD.\n");
+
+	dlm_ls_pthread_init(lockspace);
+	DEBUGLOG("DLM initialisation complete\n");
+	return 0;
+}
+
+static void _cluster_init_completed(void)
+{
+	clvmd_cluster_init_completed();
+}
+
+static int _get_main_cluster_fd()
+{
+	return cman_get_fd(c_handle);
+}
+
+static int _get_num_nodes()
+{
+	int i;
+	int nnodes = 0;
+
+	/* return number of ACTIVE nodes */
+	for (i=0; i<num_nodes; i++) {
+		if (nodes[i].cn_member && nodes[i].cn_nodeid)
+			nnodes++;
+	}
+	return nnodes;
+}
+
+/* send_message with the fd check removed */
+static int _cluster_send_message(const void *buf, int msglen, const char *csid,
+				 const char *errtext)
+{
+	int nodeid = 0;
+
+	if (csid)
+		memcpy(&nodeid, csid, CMAN_MAX_CSID_LEN);
+
+	if (cman_send_data(c_handle, buf, msglen, 0, CLUSTER_PORT_CLVMD, nodeid) <= 0)
+	{
+		log_error("%s", errtext);
+	}
+	return msglen;
+}
+
+static void _get_our_csid(char *csid)
+{
+	if (this_node.cn_nodeid == 0) {
+		cman_get_node(c_handle, 0, &this_node);
+	}
+	memcpy(csid, &this_node.cn_nodeid, CMAN_MAX_CSID_LEN);
+}
+
+/* Call a callback routine for each node is that known (down means not running a clvmd) */
+static int _cluster_do_node_callback(struct local_client *client,
+				     void (*callback) (struct local_client *,
+						       const char *,
+						       int))
+{
+	int i;
+	int somedown = 0;
+
+	for (i = 0; i < _get_num_nodes(); i++) {
+		if (nodes[i].cn_member && nodes[i].cn_nodeid) {
+			int up = (int)(long)dm_hash_lookup_binary(node_updown_hash, (char *)&nodes[i].cn_nodeid, sizeof(int));
+
+			callback(client, (char *)&nodes[i].cn_nodeid, up);
+			if (!up)
+				somedown = -1;
+		}
+	}
+	return somedown;
+}
+
+/* Process OOB messages from the cluster socket */
+static void event_callback(cman_handle_t handle, void *private, int reason, int arg)
+{
+	char namebuf[MAX_CLUSTER_MEMBER_NAME_LEN];
+
+	switch (reason) {
+        case CMAN_REASON_PORTCLOSED:
+		name_from_nodeid(arg, namebuf);
+		log_notice("clvmd on node %s has died\n", namebuf);
+		DEBUGLOG("Got port closed message, removing node %s\n", namebuf);
+
+		dm_hash_insert_binary(node_updown_hash, (char *)&arg, sizeof(int), (void *)0);
+		break;
+
+	case CMAN_REASON_STATECHANGE:
+		DEBUGLOG("Got state change message, re-reading members list\n");
+		get_members();
+		break;
+
+#if defined(LIBCMAN_VERSION) && LIBCMAN_VERSION >= 2
+	case CMAN_REASON_PORTOPENED:
+		/* Ignore this, wait for startup message from clvmd itself */
+		break;
+
+	case CMAN_REASON_TRY_SHUTDOWN:
+		DEBUGLOG("Got try shutdown, sending OK\n");
+		cman_replyto_shutdown(c_handle, 1);
+		break;
+#endif
+	default:
+		/* ERROR */
+		DEBUGLOG("Got unknown event callback message: %d\n", reason);
+		break;
+	}
+}
+
+static struct local_client *cman_client;
+static int _cluster_fd_callback(struct local_client *fd, char *buf, int len,
+				const char *csid,
+				struct local_client **new_client)
+{
+
+	/* Save this for data_callback */
+	cman_client = fd;
+
+	/* We never return a new client */
+	*new_client = NULL;
+
+	return cman_dispatch(c_handle, 0);
+}
+
+
+static void data_callback(cman_handle_t handle, void *private,
+			  char *buf, int len, uint8_t port, int nodeid)
+{
+	/* Ignore looped back messages */
+	if (nodeid == this_node.cn_nodeid)
+		return;
+	process_message(cman_client, buf, len, (char *)&nodeid);
+}
+
+static void _add_up_node(const char *csid)
+{
+	/* It's up ! */
+	int nodeid = nodeid_from_csid(csid);
+
+	dm_hash_insert_binary(node_updown_hash, (char *)&nodeid, sizeof(int), (void *)1);
+	DEBUGLOG("Added new node %d to updown list\n", nodeid);
+}
+
+static void _cluster_closedown()
+{
+	destroy_lvhash();
+	dlm_release_lockspace(LOCKSPACE_NAME, lockspace, 1);
+	cman_finish(c_handle);
+}
+
+static int is_listening(int nodeid)
+{
+	int status;
+
+	do {
+		status = cman_is_listening(c_handle, nodeid, CLUSTER_PORT_CLVMD);
+		if (status < 0 && errno == EBUSY) {	/* Don't busywait */
+			sleep(1);
+			errno = EBUSY;	/* In case sleep trashes it */
+		}
+	}
+	while (status < 0 && errno == EBUSY);
+
+	return status;
+}
+
+/* Populate the list of CLVMDs running.
+   called only at startup time */
+static void count_clvmds_running(void)
+{
+	int i;
+
+	for (i = 0; i < num_nodes; i++) {
+		int nodeid = nodes[i].cn_nodeid;
+
+		if (is_listening(nodeid) == 1)
+			dm_hash_insert_binary(node_updown_hash, (void *)&nodeid, sizeof(int), (void*)1);
+		else
+			dm_hash_insert_binary(node_updown_hash, (void *)&nodeid, sizeof(int), (void*)0);
+	}
+}
+
+/* Get a list of active cluster members */
+static void get_members()
+{
+	int retnodes;
+	int status;
+	int i;
+	int high_nodeid = 0;
+
+	num_nodes = cman_get_node_count(c_handle);
+	if (num_nodes == -1) {
+		log_error("Unable to get node count");
+		return;
+	}
+
+	/* Not enough room for new nodes list ? */
+	if (num_nodes > count_nodes && nodes) {
+		free(nodes);
+		nodes = NULL;
+	}
+
+	if (nodes == NULL) {
+		count_nodes = num_nodes + 10; /* Overallocate a little */
+		nodes = malloc(count_nodes * sizeof(struct cman_node));
+		if (!nodes) {
+			log_error("Unable to allocate nodes array\n");
+			exit(5);
+		}
+	}
+
+	status = cman_get_nodes(c_handle, count_nodes, &retnodes, nodes);
+	if (status < 0) {
+		log_error("Unable to get node details");
+		exit(6);
+	}
+
+	/* Get the highest nodeid */
+	for (i=0; i<retnodes; i++) {
+		if (nodes[i].cn_nodeid > high_nodeid)
+			high_nodeid = nodes[i].cn_nodeid;
+	}
+}
+
+
+/* Convert a node name to a CSID */
+static int _csid_from_name(char *csid, const char *name)
+{
+	int i;
+
+	for (i = 0; i < num_nodes; i++) {
+		if (strcmp(name, nodes[i].cn_name) == 0) {
+			memcpy(csid, &nodes[i].cn_nodeid, CMAN_MAX_CSID_LEN);
+			return 0;
+		}
+	}
+	return -1;
+}
+
+/* Convert a CSID to a node name */
+static int _name_from_csid(const char *csid, char *name)
+{
+	int i;
+
+	for (i = 0; i < num_nodes; i++) {
+		if (memcmp(csid, &nodes[i].cn_nodeid, CMAN_MAX_CSID_LEN) == 0) {
+			strcpy(name, nodes[i].cn_name);
+			return 0;
+		}
+	}
+	/* Who?? */
+	strcpy(name, "Unknown");
+	return -1;
+}
+
+/* Convert a node ID to a node name */
+static int name_from_nodeid(int nodeid, char *name)
+{
+	int i;
+
+	for (i = 0; i < num_nodes; i++) {
+		if (nodeid == nodes[i].cn_nodeid) {
+			strcpy(name, nodes[i].cn_name);
+			return 0;
+		}
+	}
+	/* Who?? */
+	strcpy(name, "Unknown");
+	return -1;
+}
+
+/* Convert a CSID to a node ID */
+static int nodeid_from_csid(const char *csid)
+{
+        int nodeid;
+
+	memcpy(&nodeid, csid, CMAN_MAX_CSID_LEN);
+
+	return nodeid;
+}
+
+static int _is_quorate()
+{
+	return cman_is_quorate(c_handle);
+}
+
+static void sync_ast_routine(void *arg)
+{
+	struct lock_wait *lwait = arg;
+
+	pthread_mutex_lock(&lwait->mutex);
+	pthread_cond_signal(&lwait->cond);
+	pthread_mutex_unlock(&lwait->mutex);
+}
+
+static int _sync_lock(const char *resource, int mode, int flags, int *lockid)
+{
+	int status;
+	struct lock_wait lwait;
+
+	if (!lockid) {
+		errno = EINVAL;
+		return -1;
+	}
+
+	DEBUGLOG("sync_lock: '%s' mode:%d flags=%d\n", resource,mode,flags);
+	/* Conversions need the lockid in the LKSB */
+	if (flags & LKF_CONVERT)
+		lwait.lksb.sb_lkid = *lockid;
+
+	pthread_cond_init(&lwait.cond, NULL);
+	pthread_mutex_init(&lwait.mutex, NULL);
+	pthread_mutex_lock(&lwait.mutex);
+
+	status = dlm_ls_lock(lockspace,
+			     mode,
+			     &lwait.lksb,
+			     flags,
+			     resource,
+			     strlen(resource),
+			     0, sync_ast_routine, &lwait, NULL, NULL);
+	if (status)
+		return status;
+
+	/* Wait for it to complete */
+	pthread_cond_wait(&lwait.cond, &lwait.mutex);
+	pthread_mutex_unlock(&lwait.mutex);
+
+	*lockid = lwait.lksb.sb_lkid;
+
+	errno = lwait.lksb.sb_status;
+	DEBUGLOG("sync_lock: returning lkid %x\n", *lockid);
+	if (lwait.lksb.sb_status)
+		return -1;
+	else
+		return 0;
+}
+
+static int _sync_unlock(const char *resource /* UNUSED */, int lockid)
+{
+	int status;
+	struct lock_wait lwait;
+
+	DEBUGLOG("sync_unlock: '%s' lkid:%x\n", resource, lockid);
+
+	pthread_cond_init(&lwait.cond, NULL);
+	pthread_mutex_init(&lwait.mutex, NULL);
+	pthread_mutex_lock(&lwait.mutex);
+
+	status = dlm_ls_unlock(lockspace, lockid, 0, &lwait.lksb, &lwait);
+
+	if (status)
+		return status;
+
+	/* Wait for it to complete */
+	pthread_cond_wait(&lwait.cond, &lwait.mutex);
+	pthread_mutex_unlock(&lwait.mutex);
+
+	errno = lwait.lksb.sb_status;
+	if (lwait.lksb.sb_status != EUNLOCK)
+		return -1;
+	else
+		return 0;
+
+}
+
+static int _get_cluster_name(char *buf, int buflen)
+{
+	cman_cluster_t cluster_info;
+	int status;
+
+	status = cman_get_cluster(c_handle, &cluster_info);
+	if (!status) {
+		strncpy(buf, cluster_info.ci_name, buflen);
+	}
+	return status;
+}
+
+static struct cluster_ops _cluster_cman_ops = {
+	.name                     = "cman",
+	.cluster_init_completed   = _cluster_init_completed,
+	.cluster_send_message     = _cluster_send_message,
+	.name_from_csid           = _name_from_csid,
+	.csid_from_name           = _csid_from_name,
+	.get_num_nodes            = _get_num_nodes,
+	.cluster_fd_callback      = _cluster_fd_callback,
+	.get_main_cluster_fd      = _get_main_cluster_fd,
+	.cluster_do_node_callback = _cluster_do_node_callback,
+	.is_quorate               = _is_quorate,
+	.get_our_csid             = _get_our_csid,
+	.add_up_node              = _add_up_node,
+	.cluster_closedown        = _cluster_closedown,
+	.get_cluster_name         = _get_cluster_name,
+	.sync_lock                = _sync_lock,
+	.sync_unlock              = _sync_unlock,
+};
+
+struct cluster_ops *init_cman_cluster(void)
+{
+	if (!_init_cluster())
+		return &_cluster_cman_ops;
+	else
+		return NULL;
+}

daemons/clvmd/clvmd-command.c

@@ -0,0 +1,435 @@
+/*
+ * Copyright (C) 2002-2004 Sistina Software, Inc. All rights reserved.
+ * Copyright (C) 2004-2011 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU General Public License v.2.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+/*
+
+  CLVMD Cluster LVM daemon command processor.
+
+  To add commands to the daemon simply add a processor in do_command and return
+  and messages back in buf and the length in *retlen. The initial value of
+  buflen is the maximum size of the buffer. if buf is not large enough then it
+  may be reallocated by the functions in here to a suitable size bearing in
+  mind that anything larger than the passed-in size will have to be returned
+  using the system LV and so performance will suffer.
+
+  The status return will be negated and passed back to the originating node.
+
+  pre- and post- command routines are called only on the local node. The
+  purpose is primarily to get and release locks, though the pre- routine should
+  also do any other local setups required by the command (if any) and can
+  return a failure code that prevents the command from being distributed around
+  the cluster
+
+  The pre- and post- routines are run in their own thread so can block as long
+  they like, do_command is run in the main clvmd thread so should not block for
+  too long. If the pre-command returns an error code (!=0) then the command
+  will not be propogated around the cluster but the post-command WILL be called
+
+  Also note that the pre and post routine are *always* called on the local
+  node, even if the command to be executed was only requested to run on a
+  remote node. It may peek inside the client structure to check the status of
+  the command.
+
+  The clients of the daemon must, naturally, understand the return messages and
+  codes.
+
+  Routines in here may only READ the values in the client structure passed in
+  apart from client->private which they are free to do what they like with.
+
+*/
+
+#include "clvmd-common.h"
+#include "clvmd-comms.h"
+#include "clvm.h"
+#include "clvmd.h"
+#include "lvm-globals.h"
+#include "lvm-functions.h"
+
+#include "locking.h"
+
+#include <sys/utsname.h>
+
+extern struct cluster_ops *clops;
+static int restart_clvmd(void);
+
+/* This is where all the real work happens:
+   NOTE: client will be NULL when this is executed on a remote node */
+int do_command(struct local_client *client, struct clvm_header *msg, int msglen,
+	       char **buf, int buflen, int *retlen)
+{
+	char *args = msg->node + strlen(msg->node) + 1;
+	int arglen = msglen - sizeof(struct clvm_header) - strlen(msg->node);
+	int status = 0;
+	char *lockname;
+	const char *locktype;
+	struct utsname nodeinfo;
+	unsigned char lock_cmd;
+	unsigned char lock_flags;
+
+	/* Reset test mode before we start */
+	init_test(0);
+
+	/* Do the command */
+	switch (msg->cmd) {
+		/* Just a test message */
+	case CLVMD_CMD_TEST:
+		if (arglen > buflen) {
+			char *new_buf;
+			buflen = arglen + 200;
+			new_buf = realloc(*buf, buflen);
+			if (new_buf == NULL) {
+				status = errno;
+				free (*buf);
+			}
+			*buf = new_buf;
+		}
+		if (*buf) {
+			if (uname(&nodeinfo))
+				memset(&nodeinfo, 0, sizeof(nodeinfo));
+
+			*retlen = 1 + dm_snprintf(*buf, buflen,
+						  "TEST from %s: %s v%s",
+						  nodeinfo.nodename, args,
+						  nodeinfo.release);
+		}
+		break;
+
+	case CLVMD_CMD_LOCK_VG:
+		lock_cmd = args[0];
+		lock_flags = args[1];
+		lockname = &args[2];
+		/* Check to see if the VG is in use by LVM1 */
+		status = do_check_lvm1(lockname);
+		if (lock_flags & LCK_TEST_MODE)
+			init_test(1);
+		do_lock_vg(lock_cmd, lock_flags, lockname);
+		break;
+
+	case CLVMD_CMD_LOCK_LV:
+		/* This is the biggie */
+		lock_cmd = args[0];
+		lock_flags = args[1];
+		lockname = &args[2];
+		if (lock_flags & LCK_TEST_MODE)
+			init_test(1);
+		status = do_lock_lv(lock_cmd, lock_flags, lockname);
+		/* Replace EIO with something less scary */
+		if (status == EIO) {
+			*retlen = 1 + dm_snprintf(*buf, buflen, "%s",
+						  get_last_lvm_error());
+			return EIO;
+		}
+		break;
+
+	case CLVMD_CMD_LOCK_QUERY:
+		lockname = &args[2];
+		if (buflen < 3)
+			return EIO;
+		if ((locktype = do_lock_query(lockname)))
+			*retlen = 1 + dm_snprintf(*buf, buflen, "%s", locktype);
+		break;
+
+	case CLVMD_CMD_REFRESH:
+		do_refresh_cache();
+		break;
+
+	case CLVMD_CMD_SYNC_NAMES:
+		lvm_do_fs_unlock();
+		break;
+
+	case CLVMD_CMD_SET_DEBUG:
+		clvmd_set_debug((debug_t) args[0]);
+		break;
+
+	case CLVMD_CMD_RESTART:
+		status = restart_clvmd();
+		break;
+
+	case CLVMD_CMD_GET_CLUSTERNAME:
+		status = clops->get_cluster_name(*buf, buflen);
+		if (!status)
+			*retlen = strlen(*buf)+1;
+		break;
+
+	case CLVMD_CMD_VG_BACKUP:
+		/*
+		 * Do not run backup on local node, caller should do that.
+		 */
+		if (!client)
+			lvm_do_backup(&args[2]);
+		break;
+
+	default:
+		/* Won't get here because command is validated in pre_command */
+		break;
+	}
+
+	/* Check the status of the command and return the error text */
+	if (status) {
+		*retlen = 1 + ((*buf) ? dm_snprintf(*buf, buflen, "%s",
+						    strerror(status)) : -1);
+	}
+
+	return status;
+
+}
+
+static int lock_vg(struct local_client *client)
+{
+    struct dm_hash_table *lock_hash;
+    struct clvm_header *header =
+	(struct clvm_header *) client->bits.localsock.cmd;
+    unsigned char lock_cmd;
+    int lock_mode;
+    char *args = header->node + strlen(header->node) + 1;
+    int lkid;
+    int status = 0;
+    char *lockname;
+
+    /* Keep a track of VG locks in our own hash table. In current
+       practice there should only ever be more than two VGs locked
+       if a user tries to merge lots of them at once */
+    if (client->bits.localsock.private) {
+	lock_hash = (struct dm_hash_table *)client->bits.localsock.private;
+    }
+    else {
+	lock_hash = dm_hash_create(3);
+	if (!lock_hash)
+	    return ENOMEM;
+	client->bits.localsock.private = (void *)lock_hash;
+    }
+
+    lock_cmd = args[0] & (LCK_NONBLOCK | LCK_HOLD | LCK_SCOPE_MASK | LCK_TYPE_MASK);
+    lock_mode = ((int)lock_cmd & LCK_TYPE_MASK);
+    /* lock_flags = args[1]; */
+    lockname = &args[2];
+    DEBUGLOG("doing PRE command LOCK_VG '%s' at %x (client=%p)\n", lockname, lock_cmd, client);
+
+    if (lock_mode == LCK_UNLOCK) {
+
+	lkid = (int)(long)dm_hash_lookup(lock_hash, lockname);
+	if (lkid == 0)
+	    return EINVAL;
+
+	status = sync_unlock(lockname, lkid);
+	if (status)
+	    status = errno;
+	else
+	    dm_hash_remove(lock_hash, lockname);
+    }
+    else {
+	/* Read locks need to be PR; other modes get passed through */
+	if (lock_mode == LCK_READ)
+	    lock_mode = LCK_PREAD;
+	status = sync_lock(lockname, lock_mode, (lock_cmd & LCK_NONBLOCK) ? LCKF_NOQUEUE : 0, &lkid);
+	if (status)
+	    status = errno;
+	else
+	    if (!dm_hash_insert(lock_hash, lockname, (void *)(long)lkid))
+                    return ENOMEM;
+    }
+
+    return status;
+}
+
+
+/* Pre-command is a good place to get locks that are needed only for the duration
+   of the commands around the cluster (don't forget to free them in post-command),
+   and to sanity check the command arguments */
+int do_pre_command(struct local_client *client)
+{
+	struct clvm_header *header =
+	    (struct clvm_header *) client->bits.localsock.cmd;
+	unsigned char lock_cmd;
+	unsigned char lock_flags;
+	char *args = header->node + strlen(header->node) + 1;
+	int lockid = 0;
+	int status = 0;
+	char *lockname;
+
+	init_test(0);
+	switch (header->cmd) {
+	case CLVMD_CMD_TEST:
+		status = sync_lock("CLVMD_TEST", LCK_EXCL, 0, &lockid);
+		client->bits.localsock.private = (void *)(long)lockid;
+		break;
+
+	case CLVMD_CMD_LOCK_VG:
+		lockname = &args[2];
+		/* We take out a real lock unless LCK_CACHE was set */
+		if (!strncmp(lockname, "V_", 2) ||
+		    !strncmp(lockname, "P_#", 3))
+			status = lock_vg(client);
+		break;
+
+	case CLVMD_CMD_LOCK_LV:
+		lock_cmd = args[0];
+		lock_flags = args[1];
+		lockname = &args[2];
+		if (lock_flags & LCK_TEST_MODE)
+			init_test(1);
+		status = pre_lock_lv(lock_cmd, lock_flags, lockname);
+		break;
+
+	case CLVMD_CMD_REFRESH:
+	case CLVMD_CMD_GET_CLUSTERNAME:
+	case CLVMD_CMD_SET_DEBUG:
+	case CLVMD_CMD_VG_BACKUP:
+	case CLVMD_CMD_SYNC_NAMES:
+	case CLVMD_CMD_LOCK_QUERY:
+	case CLVMD_CMD_RESTART:
+		break;
+
+	default:
+		log_error("Unknown command %d received\n", header->cmd);
+		status = EINVAL;
+	}
+	return status;
+}
+
+/* Note that the post-command routine is called even if the pre-command or the real command
+   failed */
+int do_post_command(struct local_client *client)
+{
+	struct clvm_header *header =
+	    (struct clvm_header *) client->bits.localsock.cmd;
+	int status = 0;
+	unsigned char lock_cmd;
+	unsigned char lock_flags;
+	char *args = header->node + strlen(header->node) + 1;
+	char *lockname;
+
+	init_test(0);
+	switch (header->cmd) {
+	case CLVMD_CMD_TEST:
+		status =
+		    sync_unlock("CLVMD_TEST", (int) (long) client->bits.localsock.private);
+		client->bits.localsock.private = 0;
+		break;
+
+	case CLVMD_CMD_LOCK_LV:
+		lock_cmd = args[0];
+		lock_flags = args[1];
+		lockname = &args[2];
+		if (lock_flags & LCK_TEST_MODE)
+			init_test(1);
+		status = post_lock_lv(lock_cmd, lock_flags, lockname);
+		break;
+
+	default:
+		/* Nothing to do here */
+		break;
+	}
+	return status;
+}
+
+
+/* Called when the client is about to be deleted */
+void cmd_client_cleanup(struct local_client *client)
+{
+	struct dm_hash_node *v;
+	struct dm_hash_table *lock_hash;
+	int lkid;
+	char *lockname;
+
+	if (!client->bits.localsock.private)
+		return;
+
+	lock_hash = (struct dm_hash_table *)client->bits.localsock.private;
+
+	dm_hash_iterate(v, lock_hash) {
+		lkid = (int)(long)dm_hash_get_data(lock_hash, v);
+		lockname = dm_hash_get_key(lock_hash, v);
+		DEBUGLOG("cleanup: Unlocking lock %s %x\n", lockname, lkid);
+		(void) sync_unlock(lockname, lkid);
+	}
+
+	dm_hash_destroy(lock_hash);
+	client->bits.localsock.private = 0;
+}
+
+
+static int restart_clvmd(void)
+{
+	const char **argv;
+	char *lv_name;
+	int argc = 0, max_locks = 0;
+	struct dm_hash_node *hn = NULL;
+	char debug_arg[16];
+	const char *clvmd = getenv("LVM_CLVMD_BINARY") ? : CLVMD_PATH;
+
+	DEBUGLOG("clvmd restart requested\n");
+
+	/* Count exclusively-open LVs */
+	do {
+		hn = get_next_excl_lock(hn, &lv_name);
+		if (lv_name) {
+			max_locks++;
+			if (!*lv_name)
+				break; /* FIXME: Is this error ? */
+		}
+	} while (hn);
+
+	/* clvmd + locks (-E uuid) + debug (-d X) + NULL */
+	if (!(argv = malloc((max_locks * 2 + 6) * sizeof(*argv))))
+		goto_out;
+
+	/*
+	 * Build the command-line
+	 */
+	argv[argc++] = "clvmd";
+
+	/* Propagate debug options */
+	if (clvmd_get_debug()) {
+		if (dm_snprintf(debug_arg, sizeof(debug_arg), "-d%u", clvmd_get_debug()) < 0)
+			goto_out;
+		argv[argc++] = debug_arg;
+	}
+
+	/* Propagate foreground options */
+	if (clvmd_get_foreground())
+		argv[argc++] = "-f";
+
+	argv[argc++] = "-I";
+	argv[argc++] = clops->name;
+
+	/* Now add the exclusively-open LVs */
+	hn = NULL;
+	do {
+		hn = get_next_excl_lock(hn, &lv_name);
+		if (lv_name) {
+			if (!*lv_name)
+				break; /* FIXME: Is this error ? */
+			argv[argc++] = "-E";
+			argv[argc++] = lv_name;
+			DEBUGLOG("excl lock: %s\n", lv_name);
+		}
+	} while (hn);
+	argv[argc] = NULL;
+
+	/* Exec new clvmd */
+	DEBUGLOG("--- Restarting %s ---\n", clvmd);
+	for (argc = 1; argv[argc]; argc++) DEBUGLOG("--- %d: %s\n", argc, argv[argc]);
+
+	/* NOTE: This will fail when downgrading! */
+	execvp(clvmd, (char **)argv);
+out:
+	/* We failed */
+	DEBUGLOG("Restart of clvmd failed.\n");
+
+	free(argv);
+
+	return EIO;
+}

daemons/clvmd/clvmd-common.h

@@ -0,0 +1,33 @@
+/*
+ * Copyright (C) 2010 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+/*
+ * This file must be included first by every clvmd source file.
+ */
+#ifndef _LVM_CLVMD_COMMON_H
+#define _LVM_CLVMD_COMMON_H
+
+#include "configure.h"
+
+#define _GNU_SOURCE
+#define _FILE_OFFSET_BITS 64
+
+#include "libdevmapper.h"
+
+#include "lvm-logging.h"
+
+#include <unistd.h>
+#include <sys/stat.h>
+
+#endif

daemons/clvmd/clvmd-comms.h

@@ -0,0 +1,119 @@
+/*
+ * Copyright (C) 2002-2004 Sistina Software, Inc. All rights reserved.
+ * Copyright (C) 2004-2011 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU General Public License v.2.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+/*
+ * Abstraction layer for clvmd cluster communications
+ */
+
+#ifndef _CLVMD_COMMS_H
+#define _CLVMD_COMMS_H
+
+struct local_client;
+
+struct cluster_ops {
+	const char *name;
+	void (*cluster_init_completed) (void);
+
+	int (*cluster_send_message) (const void *buf, int msglen,
+				     const char *csid,
+				     const char *errtext);
+	int (*name_from_csid) (const char *csid, char *name);
+	int (*csid_from_name) (char *csid, const char *name);
+	int (*get_num_nodes) (void);
+	int (*cluster_fd_callback) (struct local_client *fd, char *buf, int len,
+				    const char *csid,
+				    struct local_client **new_client);
+	int (*get_main_cluster_fd) (void);	/* gets accept FD or cman cluster socket */
+	int (*cluster_do_node_callback) (struct local_client *client,
+					 void (*callback) (struct local_client *,
+							   const char *csid,
+							   int node_up));
+	int (*is_quorate) (void);
+
+	void (*get_our_csid) (char *csid);
+	void (*add_up_node) (const char *csid);
+	void (*reread_config) (void);
+	void (*cluster_closedown) (void);
+
+	int (*get_cluster_name)(char *buf, int buflen);
+
+	int (*sync_lock) (const char *resource, int mode,
+			  int flags, int *lockid);
+	int (*sync_unlock) (const char *resource, int lockid);
+
+};
+
+#ifdef USE_CMAN
+#  include <netinet/in.h>
+#  include "libcman.h"
+#  define CMAN_MAX_CSID_LEN 4
+#  ifndef MAX_CSID_LEN
+#    define MAX_CSID_LEN CMAN_MAX_CSID_LEN
+#  endif
+#  undef MAX_CLUSTER_MEMBER_NAME_LEN
+#  define MAX_CLUSTER_MEMBER_NAME_LEN   CMAN_MAX_NODENAME_LEN
+#  define CMAN_MAX_CLUSTER_MESSAGE 1500
+#  define CLUSTER_PORT_CLVMD 11
+struct cluster_ops *init_cman_cluster(void);
+#endif
+
+#ifdef USE_OPENAIS
+#  include <openais/saAis.h>
+#  include <corosync/totem/totem.h>
+#  define OPENAIS_CSID_LEN (sizeof(int))
+#  define OPENAIS_MAX_CLUSTER_MESSAGE         MESSAGE_SIZE_MAX
+#  define OPENAIS_MAX_CLUSTER_MEMBER_NAME_LEN SA_MAX_NAME_LENGTH
+#  ifndef MAX_CLUSTER_MEMBER_NAME_LEN
+#    define MAX_CLUSTER_MEMBER_NAME_LEN       SA_MAX_NAME_LENGTH
+#  endif
+#  ifndef CMAN_MAX_CLUSTER_MESSAGE
+#    define CMAN_MAX_CLUSTER_MESSAGE          MESSAGE_SIZE_MAX
+#  endif
+#  ifndef MAX_CSID_LEN
+#    define MAX_CSID_LEN sizeof(int)
+#  endif
+struct cluster_ops *init_openais_cluster(void);
+#endif
+
+#ifdef USE_COROSYNC
+#  include <corosync/corotypes.h>
+#  define COROSYNC_CSID_LEN (sizeof(int))
+#  define COROSYNC_MAX_CLUSTER_MESSAGE         65535
+#  define COROSYNC_MAX_CLUSTER_MEMBER_NAME_LEN CS_MAX_NAME_LENGTH
+#  ifndef MAX_CLUSTER_MEMBER_NAME_LEN
+#    define MAX_CLUSTER_MEMBER_NAME_LEN       CS_MAX_NAME_LENGTH
+#  endif
+#  ifndef CMAN_MAX_CLUSTER_MESSAGE
+#    define CMAN_MAX_CLUSTER_MESSAGE          65535
+#  endif
+#  ifndef MAX_CSID_LEN
+#    define MAX_CSID_LEN sizeof(int)
+#  endif
+struct cluster_ops *init_corosync_cluster(void);
+#endif
+
+#ifdef USE_SINGLENODE
+#  define SINGLENODE_CSID_LEN (sizeof(int))
+#  ifndef MAX_CLUSTER_MEMBER_NAME_LEN
+#    define MAX_CLUSTER_MEMBER_NAME_LEN       64
+#  endif
+#  define SINGLENODE_MAX_CLUSTER_MESSAGE          65535
+#  ifndef MAX_CSID_LEN
+#    define MAX_CSID_LEN sizeof(int)
+#  endif
+struct cluster_ops *init_singlenode_cluster(void);
+#endif
+
+#endif

daemons/clvmd/clvmd-corosync.c

@@ -0,0 +1,684 @@
+/*
+ * Copyright (C) 2009-2012 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+/*
+ * This provides the interface between clvmd and corosync/DLM as the cluster
+ * and lock manager.
+ */
+
+#include "clvmd-common.h"
+
+#include <pthread.h>
+
+#include "clvm.h"
+#include "clvmd-comms.h"
+#include "clvmd.h"
+#include "lvm-functions.h"
+
+#include "locking.h"
+
+#include <corosync/cpg.h>
+#include <corosync/quorum.h>
+
+#ifdef HAVE_COROSYNC_CONFDB_H
+#  include <corosync/confdb.h>
+#elif defined HAVE_COROSYNC_CMAP_H
+#  include <corosync/cmap.h>
+#else
+#  error "Either HAVE_COROSYNC_CONFDB_H or HAVE_COROSYNC_CMAP_H must be defined."
+#endif
+
+#include <libdlm.h>
+
+#include <syslog.h>
+
+/* Timeout value for several corosync calls */
+#define LOCKSPACE_NAME "clvmd"
+
+static void corosync_cpg_deliver_callback (cpg_handle_t handle,
+				  const struct cpg_name *groupName,
+				  uint32_t nodeid,
+				  uint32_t pid,
+				  void *msg,
+				  size_t msg_len);
+static void corosync_cpg_confchg_callback(cpg_handle_t handle,
+				 const struct cpg_name *groupName,
+				 const struct cpg_address *member_list, size_t member_list_entries,
+				 const struct cpg_address *left_list, size_t left_list_entries,
+				 const struct cpg_address *joined_list, size_t joined_list_entries);
+static void _cluster_closedown(void);
+
+/* Hash list of nodes in the cluster */
+static struct dm_hash_table *node_hash;
+
+/* Number of active nodes */
+static int num_nodes;
+static unsigned int our_nodeid;
+
+static struct local_client *cluster_client;
+
+/* Corosync handles */
+static cpg_handle_t cpg_handle;
+static quorum_handle_t quorum_handle;
+
+/* DLM Handle */
+static dlm_lshandle_t *lockspace;
+
+static struct cpg_name cpg_group_name;
+
+/* Corosync callback structs */
+cpg_callbacks_t corosync_cpg_callbacks = {
+	.cpg_deliver_fn =            corosync_cpg_deliver_callback,
+	.cpg_confchg_fn =            corosync_cpg_confchg_callback,
+};
+
+quorum_callbacks_t quorum_callbacks = {
+	.quorum_notify_fn = NULL,
+};
+
+struct node_info
+{
+	enum {NODE_UNKNOWN, NODE_DOWN, NODE_UP, NODE_CLVMD} state;
+	int nodeid;
+};
+
+
+/* Set errno to something approximating the right value and return 0 or -1 */
+static int cs_to_errno(cs_error_t err)
+{
+	switch(err)
+	{
+	case CS_OK:
+		return 0;
+        case CS_ERR_LIBRARY:
+		errno = EINVAL;
+		break;
+        case CS_ERR_VERSION:
+		errno = EINVAL;
+		break;
+        case CS_ERR_INIT:
+		errno = EINVAL;
+		break;
+        case CS_ERR_TIMEOUT:
+		errno = ETIME;
+		break;
+        case CS_ERR_TRY_AGAIN:
+		errno = EAGAIN;
+		break;
+        case CS_ERR_INVALID_PARAM:
+		errno = EINVAL;
+		break;
+        case CS_ERR_NO_MEMORY:
+		errno = ENOMEM;
+		break;
+        case CS_ERR_BAD_HANDLE:
+		errno = EINVAL;
+		break;
+        case CS_ERR_BUSY:
+		errno = EBUSY;
+		break;
+        case CS_ERR_ACCESS:
+		errno = EPERM;
+		break;
+        case CS_ERR_NOT_EXIST:
+		errno = ENOENT;
+		break;
+        case CS_ERR_NAME_TOO_LONG:
+		errno = ENAMETOOLONG;
+		break;
+        case CS_ERR_EXIST:
+		errno = EEXIST;
+		break;
+        case CS_ERR_NO_SPACE:
+		errno = ENOSPC;
+		break;
+        case CS_ERR_INTERRUPT:
+		errno = EINTR;
+		break;
+	case CS_ERR_NAME_NOT_FOUND:
+		errno = ENOENT;
+		break;
+        case CS_ERR_NO_RESOURCES:
+		errno = ENOMEM;
+		break;
+        case CS_ERR_NOT_SUPPORTED:
+		errno = EOPNOTSUPP;
+		break;
+        case CS_ERR_BAD_OPERATION:
+		errno = EINVAL;
+		break;
+        case CS_ERR_FAILED_OPERATION:
+		errno = EIO;
+		break;
+        case CS_ERR_MESSAGE_ERROR:
+		errno = EIO;
+		break;
+        case CS_ERR_QUEUE_FULL:
+		errno = EXFULL;
+		break;
+        case CS_ERR_QUEUE_NOT_AVAILABLE:
+		errno = EINVAL;
+		break;
+        case CS_ERR_BAD_FLAGS:
+		errno = EINVAL;
+		break;
+        case CS_ERR_TOO_BIG:
+		errno = E2BIG;
+		break;
+        case CS_ERR_NO_SECTIONS:
+		errno = ENOMEM;
+		break;
+	default:
+		errno = EINVAL;
+		break;
+	}
+	return -1;
+}
+
+static char *print_corosync_csid(const char *csid)
+{
+	static char buf[128];
+	int id;
+
+	memcpy(&id, csid, sizeof(int));
+	sprintf(buf, "%d", id);
+	return buf;
+}
+
+static void corosync_cpg_deliver_callback (cpg_handle_t handle,
+				  const struct cpg_name *groupName,
+				  uint32_t nodeid,
+				  uint32_t pid,
+				  void *msg,
+				  size_t msg_len)
+{
+	int target_nodeid;
+
+	memcpy(&target_nodeid, msg, COROSYNC_CSID_LEN);
+
+	DEBUGLOG("%u got message from nodeid %d for %d. len %zd\n",
+		 our_nodeid, nodeid, target_nodeid, msg_len-4);
+
+	if (nodeid != our_nodeid)
+		if (target_nodeid == our_nodeid || target_nodeid == 0)
+			process_message(cluster_client, (char *)msg+COROSYNC_CSID_LEN,
+					msg_len-COROSYNC_CSID_LEN, (char*)&nodeid);
+}
+
+static void corosync_cpg_confchg_callback(cpg_handle_t handle,
+				 const struct cpg_name *groupName,
+				 const struct cpg_address *member_list, size_t member_list_entries,
+				 const struct cpg_address *left_list, size_t left_list_entries,
+				 const struct cpg_address *joined_list, size_t joined_list_entries)
+{
+	int i;
+	struct node_info *ninfo;
+
+	DEBUGLOG("confchg callback. %zd joined, %zd left, %zd members\n",
+		 joined_list_entries, left_list_entries, member_list_entries);
+
+	for (i=0; i<joined_list_entries; i++) {
+		ninfo = dm_hash_lookup_binary(node_hash,
+					      (char *)&joined_list[i].nodeid,
+					      COROSYNC_CSID_LEN);
+		if (!ninfo) {
+			ninfo = malloc(sizeof(struct node_info));
+			if (!ninfo) {
+				break;
+			}
+			else {
+				ninfo->nodeid = joined_list[i].nodeid;
+				dm_hash_insert_binary(node_hash,
+						      (char *)&ninfo->nodeid,
+						      COROSYNC_CSID_LEN, ninfo);
+			}
+		}
+		ninfo->state = NODE_CLVMD;
+	}
+
+	for (i=0; i<left_list_entries; i++) {
+		ninfo = dm_hash_lookup_binary(node_hash,
+					      (char *)&left_list[i].nodeid,
+					      COROSYNC_CSID_LEN);
+		if (ninfo)
+			ninfo->state = NODE_DOWN;
+	}
+
+	for (i=0; i<member_list_entries; i++) {
+		if (member_list[i].nodeid == 0) continue;
+		ninfo = dm_hash_lookup_binary(node_hash,
+				(char *)&member_list[i].nodeid,
+				COROSYNC_CSID_LEN);
+		if (!ninfo) {
+			ninfo = malloc(sizeof(struct node_info));
+			if (!ninfo) {
+				break;
+			}
+			else {
+				ninfo->nodeid = member_list[i].nodeid;
+				dm_hash_insert_binary(node_hash,
+						(char *)&ninfo->nodeid,
+						COROSYNC_CSID_LEN, ninfo);
+			}
+		}
+		ninfo->state = NODE_CLVMD;
+	}
+
+	num_nodes = member_list_entries;
+}
+
+static int _init_cluster(void)
+{
+	cs_error_t err;
+
+#ifdef QUORUM_SET	/* corosync/quorum.h */
+	uint32_t quorum_type;
+#endif
+
+	node_hash = dm_hash_create(100);
+
+	err = cpg_initialize(&cpg_handle,
+			     &corosync_cpg_callbacks);
+	if (err != CS_OK) {
+		syslog(LOG_ERR, "Cannot initialise Corosync CPG service: %d",
+		       err);
+		DEBUGLOG("Cannot initialise Corosync CPG service: %d", err);
+		return cs_to_errno(err);
+	}
+
+#ifdef QUORUM_SET
+	err = quorum_initialize(&quorum_handle,
+				&quorum_callbacks,
+				&quorum_type);
+
+	if (quorum_type != QUORUM_SET) {
+		syslog(LOG_ERR, "Corosync quorum service is not configured");
+		DEBUGLOG("Corosync quorum service is not configured");
+		return EINVAL;
+	}
+#else
+	err = quorum_initialize(&quorum_handle,
+				&quorum_callbacks);
+#endif
+
+	if (err != CS_OK) {
+		syslog(LOG_ERR, "Cannot initialise Corosync quorum service: %d",
+		       err);
+		DEBUGLOG("Cannot initialise Corosync quorum service: %d", err);
+		return cs_to_errno(err);
+	}
+
+	/* Create a lockspace for LV & VG locks to live in */
+	lockspace = dlm_open_lockspace(LOCKSPACE_NAME);
+	if (!lockspace) {
+		lockspace = dlm_create_lockspace(LOCKSPACE_NAME, 0600);
+		if (!lockspace) {
+			syslog(LOG_ERR, "Unable to create DLM lockspace for CLVM: %m");
+			return -1;
+		}
+		DEBUGLOG("Created DLM lockspace for CLVMD.\n");
+	} else
+		DEBUGLOG("Opened existing DLM lockspace for CLVMD.\n");
+
+	dlm_ls_pthread_init(lockspace);
+	DEBUGLOG("DLM initialisation complete\n");
+
+	/* Connect to the clvmd group */
+	strcpy((char *)cpg_group_name.value, "clvmd");
+	cpg_group_name.length = strlen((char *)cpg_group_name.value);
+	err = cpg_join(cpg_handle, &cpg_group_name);
+	if (err != CS_OK) {
+		cpg_finalize(cpg_handle);
+		quorum_finalize(quorum_handle);
+		dlm_release_lockspace(LOCKSPACE_NAME, lockspace, 1);
+		syslog(LOG_ERR, "Cannot join clvmd process group");
+		DEBUGLOG("Cannot join clvmd process group: %d\n", err);
+		return cs_to_errno(err);
+	}
+
+	err = cpg_local_get(cpg_handle,
+			    &our_nodeid);
+	if (err != CS_OK) {
+		cpg_finalize(cpg_handle);
+		quorum_finalize(quorum_handle);
+		dlm_release_lockspace(LOCKSPACE_NAME, lockspace, 1);
+		syslog(LOG_ERR, "Cannot get local node id\n");
+		return cs_to_errno(err);
+	}
+	DEBUGLOG("Our local node id is %d\n", our_nodeid);
+
+	DEBUGLOG("Connected to Corosync\n");
+
+	return 0;
+}
+
+static void _cluster_closedown(void)
+{
+	DEBUGLOG("cluster_closedown\n");
+	destroy_lvhash();
+
+	dlm_release_lockspace(LOCKSPACE_NAME, lockspace, 1);
+	cpg_finalize(cpg_handle);
+	quorum_finalize(quorum_handle);
+}
+
+static void _get_our_csid(char *csid)
+{
+	memcpy(csid, &our_nodeid, sizeof(int));
+}
+
+/* Corosync doesn't really have nmode names so we
+   just use the node ID in hex instead */
+static int _csid_from_name(char *csid, const char *name)
+{
+	int nodeid;
+	struct node_info *ninfo;
+
+	if (sscanf(name, "%x", &nodeid) == 1) {
+		ninfo = dm_hash_lookup_binary(node_hash, csid, COROSYNC_CSID_LEN);
+		if (ninfo)
+			return nodeid;
+	}
+	return -1;
+}
+
+static int _name_from_csid(const char *csid, char *name)
+{
+	struct node_info *ninfo;
+
+	ninfo = dm_hash_lookup_binary(node_hash, csid, COROSYNC_CSID_LEN);
+	if (!ninfo)
+	{
+		sprintf(name, "UNKNOWN %s", print_corosync_csid(csid));
+		return -1;
+	}
+
+	sprintf(name, "%x", ninfo->nodeid);
+	return 0;
+}
+
+static int _get_num_nodes()
+{
+	DEBUGLOG("num_nodes = %d\n", num_nodes);
+	return num_nodes;
+}
+
+/* Node is now known to be running a clvmd */
+static void _add_up_node(const char *csid)
+{
+	struct node_info *ninfo;
+
+	ninfo = dm_hash_lookup_binary(node_hash, csid, COROSYNC_CSID_LEN);
+	if (!ninfo) {
+		DEBUGLOG("corosync_add_up_node no node_hash entry for csid %s\n",
+			 print_corosync_csid(csid));
+		return;
+	}
+
+	DEBUGLOG("corosync_add_up_node %d\n", ninfo->nodeid);
+
+	ninfo->state = NODE_CLVMD;
+
+	return;
+}
+
+/* Call a callback for each node, so the caller knows whether it's up or down */
+static int _cluster_do_node_callback(struct local_client *master_client,
+				     void (*callback)(struct local_client *,
+						      const char *csid, int node_up))
+{
+	struct dm_hash_node *hn;
+	struct node_info *ninfo;
+	int somedown = 0;
+
+	dm_hash_iterate(hn, node_hash)
+	{
+		char csid[COROSYNC_CSID_LEN];
+
+		ninfo = dm_hash_get_data(node_hash, hn);
+		memcpy(csid, dm_hash_get_key(node_hash, hn), COROSYNC_CSID_LEN);
+
+		DEBUGLOG("down_callback. node %d, state = %d\n", ninfo->nodeid,
+			 ninfo->state);
+
+		if (ninfo->state != NODE_DOWN)
+			callback(master_client, csid, ninfo->state == NODE_CLVMD);
+		if (ninfo->state != NODE_CLVMD)
+			somedown = -1;
+	}
+	return somedown;
+}
+
+/* Real locking */
+static int _lock_resource(const char *resource, int mode, int flags, int *lockid)
+{
+	struct dlm_lksb lksb;
+	int err;
+
+	DEBUGLOG("lock_resource '%s', flags=%d, mode=%d\n", resource, flags, mode);
+
+	if (flags & LKF_CONVERT)
+		lksb.sb_lkid = *lockid;
+
+	err = dlm_ls_lock_wait(lockspace,
+			       mode,
+			       &lksb,
+			       flags,
+			       resource,
+			       strlen(resource),
+			       0,
+			       NULL, NULL, NULL);
+
+	if (err != 0)
+	{
+		DEBUGLOG("dlm_ls_lock returned %d\n", errno);
+		return err;
+	}
+	if (lksb.sb_status != 0)
+	{
+		DEBUGLOG("dlm_ls_lock returns lksb.sb_status %d\n", lksb.sb_status);
+		errno = lksb.sb_status;
+		return -1;
+	}
+
+	DEBUGLOG("lock_resource returning %d, lock_id=%x\n", err, lksb.sb_lkid);
+
+	*lockid = lksb.sb_lkid;
+
+	return 0;
+}
+
+
+static int _unlock_resource(const char *resource, int lockid)
+{
+	struct dlm_lksb lksb;
+	int err;
+
+	DEBUGLOG("unlock_resource: %s lockid: %x\n", resource, lockid);
+	lksb.sb_lkid = lockid;
+
+	err = dlm_ls_unlock_wait(lockspace,
+				 lockid,
+				 0,
+				 &lksb);
+	if (err != 0)
+	{
+		DEBUGLOG("Unlock returned %d\n", err);
+		return err;
+	}
+	if (lksb.sb_status != EUNLOCK)
+	{
+		DEBUGLOG("dlm_ls_unlock_wait returns lksb.sb_status: %d\n", lksb.sb_status);
+		errno = lksb.sb_status;
+		return -1;
+	}   
+
+
+	return 0;
+}
+
+static int _is_quorate()
+{
+	int quorate;
+	if (quorum_getquorate(quorum_handle, &quorate) == CS_OK)
+		return quorate;
+	else
+		return 0;
+}
+
+static int _get_main_cluster_fd(void)
+{
+	int select_fd;
+
+	cpg_fd_get(cpg_handle, &select_fd);
+	return select_fd;
+}
+
+static int _cluster_fd_callback(struct local_client *fd, char *buf, int len,
+				const char *csid,
+				struct local_client **new_client)
+{
+	cluster_client = fd;
+	*new_client = NULL;
+	cpg_dispatch(cpg_handle, CS_DISPATCH_ONE);
+	return 1;
+}
+
+static int _cluster_send_message(const void *buf, int msglen, const char *csid,
+				 const char *errtext)
+{
+	struct iovec iov[2];
+	cs_error_t err;
+	int target_node;
+
+	if (csid)
+		memcpy(&target_node, csid, COROSYNC_CSID_LEN);
+	else
+		target_node = 0;
+
+	iov[0].iov_base = &target_node;
+	iov[0].iov_len = sizeof(int);
+	iov[1].iov_base = (char *)buf;
+	iov[1].iov_len = msglen;
+
+	err = cpg_mcast_joined(cpg_handle, CPG_TYPE_AGREED, iov, 2);
+	return cs_to_errno(err);
+}
+
+#ifdef HAVE_COROSYNC_CONFDB_H
+/*
+ * We are not necessarily connected to a Red Hat Cluster system,
+ * but if we are, this returns the cluster name from cluster.conf.
+ * I've used confdb rather than ccs to reduce the inter-package
+ * dependancies as well as to allow people to set a cluster name
+ * for themselves even if they are not running on RH cluster.
+ */
+static int _get_cluster_name(char *buf, int buflen)
+{
+	confdb_handle_t handle;
+	int result;
+	size_t namelen = buflen;
+	hdb_handle_t cluster_handle;
+	confdb_callbacks_t callbacks = {
+		.confdb_key_change_notify_fn = NULL,
+		.confdb_object_create_change_notify_fn = NULL,
+		.confdb_object_delete_change_notify_fn = NULL
+	};
+
+	/* This is a default in case everything else fails */
+	strncpy(buf, "Corosync", buflen);
+
+	/* Look for a cluster name in confdb */
+	result = confdb_initialize (&handle, &callbacks);
+        if (result != CS_OK)
+		return 0;
+
+        result = confdb_object_find_start(handle, OBJECT_PARENT_HANDLE);
+	if (result != CS_OK)
+		goto out;
+
+        result = confdb_object_find(handle, OBJECT_PARENT_HANDLE, (void *)"cluster", strlen("cluster"), &cluster_handle);
+        if (result != CS_OK)
+		goto out;
+
+        result = confdb_key_get(handle, cluster_handle, (void *)"name", strlen("name"), buf, &namelen);
+        if (result != CS_OK)
+		goto out;
+
+	buf[namelen] = '\0';
+
+out:
+	confdb_finalize(handle);
+	return 0;
+}
+
+#elif defined HAVE_COROSYNC_CMAP_H
+
+static int _get_cluster_name(char *buf, int buflen)
+{
+	cmap_handle_t cmap_handle = 0;
+	int result;
+	char *name = NULL;
+
+	/* This is a default in case everything else fails */
+	strncpy(buf, "Corosync", buflen);
+
+	/* Look for a cluster name in cmap */
+	result = cmap_initialize(&cmap_handle);
+	if (result != CS_OK)
+		return 0;
+
+	result = cmap_get_string(cmap_handle, "totem.cluster_name", &name);
+	if (result != CS_OK)
+		goto out;
+
+	memset(buf, 0, buflen);
+	strncpy(buf, name, buflen - 1);
+
+out:
+	if (name)
+		free(name);
+	cmap_finalize(cmap_handle);
+	return 0;
+}
+
+#endif
+
+static struct cluster_ops _cluster_corosync_ops = {
+	.name                     = "corosync",
+	.cluster_init_completed   = NULL,
+	.cluster_send_message     = _cluster_send_message,
+	.name_from_csid           = _name_from_csid,
+	.csid_from_name           = _csid_from_name,
+	.get_num_nodes            = _get_num_nodes,
+	.cluster_fd_callback      = _cluster_fd_callback,
+	.get_main_cluster_fd      = _get_main_cluster_fd,
+	.cluster_do_node_callback = _cluster_do_node_callback,
+	.is_quorate               = _is_quorate,
+	.get_our_csid             = _get_our_csid,
+	.add_up_node              = _add_up_node,
+	.reread_config            = NULL,
+	.cluster_closedown        = _cluster_closedown,
+	.get_cluster_name         = _get_cluster_name,
+	.sync_lock                = _lock_resource,
+	.sync_unlock              = _unlock_resource,
+};
+
+struct cluster_ops *init_corosync_cluster(void)
+{
+	if (!_init_cluster())
+		return &_cluster_corosync_ops;
+	else
+		return NULL;
+}

daemons/clvmd/clvmd-openais.c

@@ -0,0 +1,694 @@
+/*
+ * Copyright (C) 2007-2009 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+/*
+ * This provides the interface between clvmd and OpenAIS as the cluster
+ * and lock manager.
+ */
+
+#include "clvmd-common.h"
+
+#include <pthread.h>
+#include <fcntl.h>
+#include <syslog.h>
+
+#include <openais/saAis.h>
+#include <openais/saLck.h>
+
+#include <corosync/corotypes.h>
+#include <corosync/cpg.h>
+
+#include "locking.h"
+#include "clvm.h"
+#include "clvmd-comms.h"
+#include "lvm-functions.h"
+#include "clvmd.h"
+
+/* Timeout value for several openais calls */
+#define TIMEOUT 10
+
+static void openais_cpg_deliver_callback (cpg_handle_t handle,
+				  const struct cpg_name *groupName,
+				  uint32_t nodeid,
+				  uint32_t pid,
+				  void *msg,
+				  size_t msg_len);
+static void openais_cpg_confchg_callback(cpg_handle_t handle,
+				 const struct cpg_name *groupName,
+				 const struct cpg_address *member_list, size_t member_list_entries,
+				 const struct cpg_address *left_list, size_t left_list_entries,
+				 const struct cpg_address *joined_list, size_t joined_list_entries);
+
+static void _cluster_closedown(void);
+
+/* Hash list of nodes in the cluster */
+static struct dm_hash_table *node_hash;
+
+/* For associating lock IDs & resource handles */
+static struct dm_hash_table *lock_hash;
+
+/* Number of active nodes */
+static int num_nodes;
+static unsigned int our_nodeid;
+
+static struct local_client *cluster_client;
+
+/* OpenAIS handles */
+static cpg_handle_t cpg_handle;
+static SaLckHandleT lck_handle;
+
+static struct cpg_name cpg_group_name;
+
+/* Openais callback structs */
+cpg_callbacks_t openais_cpg_callbacks = {
+	.cpg_deliver_fn =            openais_cpg_deliver_callback,
+	.cpg_confchg_fn =            openais_cpg_confchg_callback,
+};
+
+struct node_info
+{
+	enum {NODE_UNKNOWN, NODE_DOWN, NODE_UP, NODE_CLVMD} state;
+	int nodeid;
+};
+
+struct lock_info
+{
+	SaLckResourceHandleT res_handle;
+	SaLckLockIdT         lock_id;
+	SaNameT              lock_name;
+};
+
+/* Set errno to something approximating the right value and return 0 or -1 */
+static int ais_to_errno(SaAisErrorT err)
+{
+	switch(err)
+	{
+	case SA_AIS_OK:
+		return 0;
+        case SA_AIS_ERR_LIBRARY:
+		errno = EINVAL;
+		break;
+        case SA_AIS_ERR_VERSION:
+		errno = EINVAL;
+		break;
+        case SA_AIS_ERR_INIT:
+		errno = EINVAL;
+		break;
+        case SA_AIS_ERR_TIMEOUT:
+		errno = ETIME;
+		break;
+        case SA_AIS_ERR_TRY_AGAIN:
+		errno = EAGAIN;
+		break;
+        case SA_AIS_ERR_INVALID_PARAM:
+		errno = EINVAL;
+		break;
+        case SA_AIS_ERR_NO_MEMORY:
+		errno = ENOMEM;
+		break;
+        case SA_AIS_ERR_BAD_HANDLE:
+		errno = EINVAL;
+		break;
+        case SA_AIS_ERR_BUSY:
+		errno = EBUSY;
+		break;
+        case SA_AIS_ERR_ACCESS:
+		errno = EPERM;
+		break;
+        case SA_AIS_ERR_NOT_EXIST:
+		errno = ENOENT;
+		break;
+        case SA_AIS_ERR_NAME_TOO_LONG:
+		errno = ENAMETOOLONG;
+		break;
+        case SA_AIS_ERR_EXIST:
+		errno = EEXIST;
+		break;
+        case SA_AIS_ERR_NO_SPACE:
+		errno = ENOSPC;
+		break;
+        case SA_AIS_ERR_INTERRUPT:
+		errno = EINTR;
+		break;
+	case SA_AIS_ERR_NAME_NOT_FOUND:
+		errno = ENOENT;
+		break;
+        case SA_AIS_ERR_NO_RESOURCES:
+		errno = ENOMEM;
+		break;
+        case SA_AIS_ERR_NOT_SUPPORTED:
+		errno = EOPNOTSUPP;
+		break;
+        case SA_AIS_ERR_BAD_OPERATION:
+		errno = EINVAL;
+		break;
+        case SA_AIS_ERR_FAILED_OPERATION:
+		errno = EIO;
+		break;
+        case SA_AIS_ERR_MESSAGE_ERROR:
+		errno = EIO;
+		break;
+        case SA_AIS_ERR_QUEUE_FULL:
+		errno = EXFULL;
+		break;
+        case SA_AIS_ERR_QUEUE_NOT_AVAILABLE:
+		errno = EINVAL;
+		break;
+        case SA_AIS_ERR_BAD_FLAGS:
+		errno = EINVAL;
+		break;
+        case SA_AIS_ERR_TOO_BIG:
+		errno = E2BIG;
+		break;
+        case SA_AIS_ERR_NO_SECTIONS:
+		errno = ENOMEM;
+		break;
+	default:
+		errno = EINVAL;
+		break;
+	}
+	return -1;
+}
+
+static char *print_openais_csid(const char *csid)
+{
+	static char buf[128];
+	int id;
+
+	memcpy(&id, csid, sizeof(int));
+	sprintf(buf, "%d", id);
+	return buf;
+}
+
+static int add_internal_client(int fd, fd_callback_t callback)
+{
+	struct local_client *client;
+
+	DEBUGLOG("Add_internal_client, fd = %d\n", fd);
+
+	client = calloc(1, sizeof(struct local_client));
+	if (!client)
+	{
+		DEBUGLOG("malloc failed\n");
+		return -1;
+	}
+
+	client->fd = fd;
+	client->type = CLUSTER_INTERNAL;
+	client->callback = callback;
+	add_client(client);
+
+	/* Set Close-on-exec */
+	fcntl(fd, F_SETFD, 1);
+
+	return 0;
+}
+
+static void openais_cpg_deliver_callback (cpg_handle_t handle,
+				  const struct cpg_name *groupName,
+				  uint32_t nodeid,
+				  uint32_t pid,
+				  void *msg,
+				  size_t msg_len)
+{
+	int target_nodeid;
+
+	memcpy(&target_nodeid, msg, OPENAIS_CSID_LEN);
+
+	DEBUGLOG("%u got message from nodeid %d for %d. len %" PRIsize_t "\n",
+		 our_nodeid, nodeid, target_nodeid, msg_len-4);
+
+	if (nodeid != our_nodeid)
+		if (target_nodeid == our_nodeid || target_nodeid == 0)
+			process_message(cluster_client, (char *)msg+OPENAIS_CSID_LEN,
+					msg_len-OPENAIS_CSID_LEN, (char*)&nodeid);
+}
+
+static void openais_cpg_confchg_callback(cpg_handle_t handle,
+				 const struct cpg_name *groupName,
+				 const struct cpg_address *member_list, size_t member_list_entries,
+				 const struct cpg_address *left_list, size_t left_list_entries,
+				 const struct cpg_address *joined_list, size_t joined_list_entries)
+{
+	int i;
+	struct node_info *ninfo;
+
+	DEBUGLOG("confchg callback. %" PRIsize_t " joined, "
+		 "%" PRIsize_t " left, %" PRIsize_t " members\n",
+		 joined_list_entries, left_list_entries, member_list_entries);
+
+	for (i=0; i<joined_list_entries; i++) {
+		ninfo = dm_hash_lookup_binary(node_hash,
+					      (char *)&joined_list[i].nodeid,
+					      OPENAIS_CSID_LEN);
+		if (!ninfo) {
+			ninfo = malloc(sizeof(struct node_info));
+			if (!ninfo) {
+				break;
+			}
+			else {
+				ninfo->nodeid = joined_list[i].nodeid;
+				dm_hash_insert_binary(node_hash,
+						      (char *)&ninfo->nodeid,
+						      OPENAIS_CSID_LEN, ninfo);
+			}
+		}
+		ninfo->state = NODE_CLVMD;
+	}
+
+	for (i=0; i<left_list_entries; i++) {
+		ninfo = dm_hash_lookup_binary(node_hash,
+					      (char *)&left_list[i].nodeid,
+					      OPENAIS_CSID_LEN);
+		if (ninfo)
+			ninfo->state = NODE_DOWN;
+	}
+
+	for (i=0; i<member_list_entries; i++) {
+		if (member_list[i].nodeid == 0) continue;
+		ninfo = dm_hash_lookup_binary(node_hash,
+				(char *)&member_list[i].nodeid,
+				OPENAIS_CSID_LEN);
+		if (!ninfo) {
+			ninfo = malloc(sizeof(struct node_info));
+			if (!ninfo) {
+				break;
+			}
+			else {
+				ninfo->nodeid = member_list[i].nodeid;
+				dm_hash_insert_binary(node_hash,
+						(char *)&ninfo->nodeid,
+						OPENAIS_CSID_LEN, ninfo);
+			}
+		}
+		ninfo->state = NODE_CLVMD;
+	}
+
+	num_nodes = member_list_entries;
+}
+
+static int lck_dispatch(struct local_client *client, char *buf, int len,
+			const char *csid, struct local_client **new_client)
+{
+	*new_client = NULL;
+	saLckDispatch(lck_handle, SA_DISPATCH_ONE);
+	return 1;
+}
+
+static int _init_cluster(void)
+{
+	SaAisErrorT err;
+	SaVersionT  ver = { 'B', 1, 1 };
+	int select_fd;
+
+	node_hash = dm_hash_create(100);
+	lock_hash = dm_hash_create(10);
+
+	err = cpg_initialize(&cpg_handle,
+			     &openais_cpg_callbacks);
+	if (err != SA_AIS_OK) {
+		syslog(LOG_ERR, "Cannot initialise OpenAIS CPG service: %d",
+		       err);
+		DEBUGLOG("Cannot initialise OpenAIS CPG service: %d", err);
+		return ais_to_errno(err);
+	}
+
+	err = saLckInitialize(&lck_handle,
+					NULL,
+			      &ver);
+	if (err != SA_AIS_OK) {
+		cpg_initialize(&cpg_handle, &openais_cpg_callbacks);
+		syslog(LOG_ERR, "Cannot initialise OpenAIS lock service: %d",
+		       err);
+		DEBUGLOG("Cannot initialise OpenAIS lock service: %d\n\n", err);
+		return ais_to_errno(err);
+	}
+
+	/* Connect to the clvmd group */
+	strcpy((char *)cpg_group_name.value, "clvmd");
+	cpg_group_name.length = strlen((char *)cpg_group_name.value);
+	err = cpg_join(cpg_handle, &cpg_group_name);
+	if (err != SA_AIS_OK) {
+		cpg_finalize(cpg_handle);
+		saLckFinalize(lck_handle);
+		syslog(LOG_ERR, "Cannot join clvmd process group");
+		DEBUGLOG("Cannot join clvmd process group: %d\n", err);
+		return ais_to_errno(err);
+	}
+
+	err = cpg_local_get(cpg_handle,
+			    &our_nodeid);
+	if (err != SA_AIS_OK) {
+		cpg_finalize(cpg_handle);
+		saLckFinalize(lck_handle);
+		syslog(LOG_ERR, "Cannot get local node id\n");
+		return ais_to_errno(err);
+	}
+	DEBUGLOG("Our local node id is %d\n", our_nodeid);
+
+	saLckSelectionObjectGet(lck_handle, (SaSelectionObjectT *)&select_fd);
+	add_internal_client(select_fd, lck_dispatch);
+
+	DEBUGLOG("Connected to OpenAIS\n");
+
+	return 0;
+}
+
+static void _cluster_closedown(void)
+{
+	DEBUGLOG("cluster_closedown\n");
+	destroy_lvhash();
+
+	saLckFinalize(lck_handle);
+	cpg_finalize(cpg_handle);
+}
+
+static void _get_our_csid(char *csid)
+{
+	memcpy(csid, &our_nodeid, sizeof(int));
+}
+
+/* OpenAIS doesn't really have nmode names so we
+   just use the node ID in hex instead */
+static int _csid_from_name(char *csid, const char *name)
+{
+	int nodeid;
+	struct node_info *ninfo;
+
+	if (sscanf(name, "%x", &nodeid) == 1) {
+		ninfo = dm_hash_lookup_binary(node_hash, csid, OPENAIS_CSID_LEN);
+		if (ninfo)
+			return nodeid;
+	}
+	return -1;
+}
+
+static int _name_from_csid(const char *csid, char *name)
+{
+	struct node_info *ninfo;
+
+	ninfo = dm_hash_lookup_binary(node_hash, csid, OPENAIS_CSID_LEN);
+	if (!ninfo)
+	{
+		sprintf(name, "UNKNOWN %s", print_openais_csid(csid));
+		return -1;
+	}
+
+	sprintf(name, "%x", ninfo->nodeid);
+	return 0;
+}
+
+static int _get_num_nodes()
+{
+	DEBUGLOG("num_nodes = %d\n", num_nodes);
+	return num_nodes;
+}
+
+/* Node is now known to be running a clvmd */
+static void _add_up_node(const char *csid)
+{
+	struct node_info *ninfo;
+
+	ninfo = dm_hash_lookup_binary(node_hash, csid, OPENAIS_CSID_LEN);
+	if (!ninfo) {
+		DEBUGLOG("openais_add_up_node no node_hash entry for csid %s\n",
+			 print_openais_csid(csid));
+		return;
+	}
+
+	DEBUGLOG("openais_add_up_node %d\n", ninfo->nodeid);
+
+	ninfo->state = NODE_CLVMD;
+
+	return;
+}
+
+/* Call a callback for each node, so the caller knows whether it's up or down */
+static int _cluster_do_node_callback(struct local_client *master_client,
+				     void (*callback)(struct local_client *,
+						      const char *csid, int node_up))
+{
+	struct dm_hash_node *hn;
+	struct node_info *ninfo;
+	int somedown = 0;
+
+	dm_hash_iterate(hn, node_hash)
+	{
+		char csid[OPENAIS_CSID_LEN];
+
+		ninfo = dm_hash_get_data(node_hash, hn);
+		memcpy(csid, dm_hash_get_key(node_hash, hn), OPENAIS_CSID_LEN);
+
+		DEBUGLOG("down_callback. node %d, state = %d\n", ninfo->nodeid,
+			 ninfo->state);
+
+		if (ninfo->state != NODE_DOWN)
+			callback(master_client, csid, ninfo->state == NODE_CLVMD);
+		if (ninfo->state != NODE_CLVMD)
+			somedown = -1;
+	}
+	return somedown;
+}
+
+/* Real locking */
+static int _lock_resource(char *resource, int mode, int flags, int *lockid)
+{
+	struct lock_info *linfo;
+	SaLckResourceHandleT res_handle;
+	SaAisErrorT err;
+	SaLckLockIdT lock_id;
+	SaLckLockStatusT lockStatus;
+
+	/* This needs to be converted from DLM/LVM2 value for OpenAIS LCK */
+	if (flags & LCK_NONBLOCK) flags = SA_LCK_LOCK_NO_QUEUE;
+
+	linfo = malloc(sizeof(struct lock_info));
+	if (!linfo)
+		return -1;
+
+	DEBUGLOG("lock_resource '%s', flags=%d, mode=%d\n", resource, flags, mode);
+
+	linfo->lock_name.length = strlen(resource)+1;
+	strcpy((char *)linfo->lock_name.value, resource);
+
+	err = saLckResourceOpen(lck_handle, &linfo->lock_name,
+				SA_LCK_RESOURCE_CREATE, TIMEOUT, &res_handle);
+	if (err != SA_AIS_OK)
+	{
+		DEBUGLOG("ResourceOpen returned %d\n", err);
+		free(linfo);
+		return ais_to_errno(err);
+	}
+
+	err = saLckResourceLock(
+			res_handle,
+			&lock_id,
+			mode,
+			flags,
+			0,
+			SA_TIME_END,
+			&lockStatus);
+	if (err != SA_AIS_OK && lockStatus != SA_LCK_LOCK_GRANTED)
+	{
+		free(linfo);
+		saLckResourceClose(res_handle);
+		return ais_to_errno(err);
+	}
+
+	/* Wait for it to complete */
+
+	DEBUGLOG("lock_resource returning %d, lock_id=%" PRIx64 "\n",
+		 err, lock_id);
+
+	linfo->lock_id = lock_id;
+	linfo->res_handle = res_handle;
+
+	dm_hash_insert(lock_hash, resource, linfo);
+
+	return ais_to_errno(err);
+}
+
+
+static int _unlock_resource(char *resource, int lockid)
+{
+	SaAisErrorT err;
+	struct lock_info *linfo;
+
+	DEBUGLOG("unlock_resource %s\n", resource);
+	linfo = dm_hash_lookup(lock_hash, resource);
+	if (!linfo)
+		return 0;
+
+	DEBUGLOG("unlock_resource: lockid: %" PRIx64 "\n", linfo->lock_id);
+	err = saLckResourceUnlock(linfo->lock_id, SA_TIME_END);
+	if (err != SA_AIS_OK)
+	{
+		DEBUGLOG("Unlock returned %d\n", err);
+		return ais_to_errno(err);
+	}
+
+	/* Release the resource */
+	dm_hash_remove(lock_hash, resource);
+	saLckResourceClose(linfo->res_handle);
+	free(linfo);
+
+	return ais_to_errno(err);
+}
+
+static int _sync_lock(const char *resource, int mode, int flags, int *lockid)
+{
+	int status;
+	char lock1[strlen(resource)+3];
+	char lock2[strlen(resource)+3];
+
+	snprintf(lock1, sizeof(lock1), "%s-1", resource);
+	snprintf(lock2, sizeof(lock2), "%s-2", resource);
+
+	switch (mode)
+	{
+	case LCK_EXCL:
+		status = _lock_resource(lock1, SA_LCK_EX_LOCK_MODE, flags, lockid);
+		if (status)
+			goto out;
+
+		/* If we can't get this lock too then bail out */
+		status = _lock_resource(lock2, SA_LCK_EX_LOCK_MODE, LCK_NONBLOCK,
+					lockid);
+		if (status == SA_LCK_LOCK_NOT_QUEUED)
+		{
+			_unlock_resource(lock1, *lockid);
+			status = -1;
+			errno = EAGAIN;
+		}
+		break;
+
+	case LCK_PREAD:
+	case LCK_READ:
+		status = _lock_resource(lock1, SA_LCK_PR_LOCK_MODE, flags, lockid);
+		if (status)
+			goto out;
+		_unlock_resource(lock2, *lockid);
+		break;
+
+	case LCK_WRITE:
+		status = _lock_resource(lock2, SA_LCK_EX_LOCK_MODE, flags, lockid);
+		if (status)
+			goto out;
+		_unlock_resource(lock1, *lockid);
+		break;
+
+	default:
+		status = -1;
+		errno = EINVAL;
+		break;
+	}
+out:
+	*lockid = mode;
+	return status;
+}
+
+static int _sync_unlock(const char *resource, int lockid)
+{
+	int status = 0;
+	char lock1[strlen(resource)+3];
+	char lock2[strlen(resource)+3];
+
+	snprintf(lock1, sizeof(lock1), "%s-1", resource);
+	snprintf(lock2, sizeof(lock2), "%s-2", resource);
+
+	_unlock_resource(lock1, lockid);
+	_unlock_resource(lock2, lockid);
+
+	return status;
+}
+
+/* We are always quorate ! */
+static int _is_quorate()
+{
+	return 1;
+}
+
+static int _get_main_cluster_fd(void)
+{
+	int select_fd;
+
+	cpg_fd_get(cpg_handle, &select_fd);
+	return select_fd;
+}
+
+static int _cluster_fd_callback(struct local_client *fd, char *buf, int len,
+				const char *csid,
+				struct local_client **new_client)
+{
+	cluster_client = fd;
+	*new_client = NULL;
+	cpg_dispatch(cpg_handle, SA_DISPATCH_ONE);
+	return 1;
+}
+
+static int _cluster_send_message(const void *buf, int msglen, const char *csid,
+				 const char *errtext)
+{
+	struct iovec iov[2];
+	SaAisErrorT err;
+	int target_node;
+
+	if (csid)
+		memcpy(&target_node, csid, OPENAIS_CSID_LEN);
+	else
+		target_node = 0;
+
+	iov[0].iov_base = &target_node;
+	iov[0].iov_len = sizeof(int);
+	iov[1].iov_base = (char *)buf;
+	iov[1].iov_len = msglen;
+
+	err = cpg_mcast_joined(cpg_handle, CPG_TYPE_AGREED, iov, 2);
+	return ais_to_errno(err);
+}
+
+/* We don't have a cluster name to report here */
+static int _get_cluster_name(char *buf, int buflen)
+{
+	strncpy(buf, "OpenAIS", buflen);
+	return 0;
+}
+
+static struct cluster_ops _cluster_openais_ops = {
+	.name                     = "openais",
+	.cluster_init_completed   = NULL,
+	.cluster_send_message     = _cluster_send_message,
+	.name_from_csid           = _name_from_csid,
+	.csid_from_name           = _csid_from_name,
+	.get_num_nodes            = _get_num_nodes,
+	.cluster_fd_callback      = _cluster_fd_callback,
+	.get_main_cluster_fd      = _get_main_cluster_fd,
+	.cluster_do_node_callback = _cluster_do_node_callback,
+	.is_quorate               = _is_quorate,
+	.get_our_csid             = _get_our_csid,
+	.add_up_node              = _add_up_node,
+	.reread_config            = NULL,
+	.cluster_closedown        = _cluster_closedown,
+	.get_cluster_name         = _get_cluster_name,
+	.sync_lock                = _sync_lock,
+	.sync_unlock              = _sync_unlock,
+};
+
+struct cluster_ops *init_openais_cluster(void)
+{
+	if (!_init_cluster())
+		return &_cluster_openais_ops;
+	else
+		return NULL;
+}

daemons/clvmd/clvmd-singlenode.c

@@ -0,0 +1,316 @@
+/*
+ * Copyright (C) 2009 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include "clvmd-common.h"
+
+#include <pthread.h>
+
+#include "locking.h"
+#include "clvm.h"
+#include "clvmd-comms.h"
+#include "lvm-functions.h"
+#include "clvmd.h"
+
+#include <sys/un.h>
+#include <sys/socket.h>
+#include <fcntl.h>
+
+static const char SINGLENODE_CLVMD_SOCKNAME[] = DEFAULT_RUN_DIR "/clvmd_singlenode.sock";
+static int listen_fd = -1;
+
+static struct dm_hash_table *_locks;
+static int _lockid;
+
+struct lock {
+	int lockid;
+	int mode;
+	int excl;
+};
+
+static void close_comms(void)
+{
+	if (listen_fd != -1 && close(listen_fd))
+		stack;
+	(void)unlink(SINGLENODE_CLVMD_SOCKNAME);
+	listen_fd = -1;
+}
+
+static int init_comms(void)
+{
+	mode_t old_mask;
+	struct sockaddr_un addr = { .sun_family = AF_UNIX };
+
+	if (!dm_strncpy(addr.sun_path, SINGLENODE_CLVMD_SOCKNAME,
+			sizeof(addr.sun_path))) {
+		DEBUGLOG("%s: singlenode socket name too long.",
+			 SINGLENODE_CLVMD_SOCKNAME);
+		return -1;
+	}
+
+	close_comms();
+
+	(void) dm_prepare_selinux_context(SINGLENODE_CLVMD_SOCKNAME, S_IFSOCK);
+	old_mask = umask(0077);
+
+	listen_fd = socket(PF_UNIX, SOCK_STREAM, 0);
+	if (listen_fd < 0) {
+		DEBUGLOG("Can't create local socket: %s\n", strerror(errno));
+		goto error;
+	}
+	/* Set Close-on-exec */
+	if (fcntl(listen_fd, F_SETFD, 1)) {
+		DEBUGLOG("Setting CLOEXEC on client fd failed: %s\n", strerror(errno));
+		goto error;
+	}
+
+	if (bind(listen_fd, (struct sockaddr *)&addr, sizeof(addr)) < 0) {
+		DEBUGLOG("Can't bind local socket: %s\n", strerror(errno));
+		goto error;
+	}
+	if (listen(listen_fd, 10) < 0) {
+		DEBUGLOG("Can't listen local socket: %s\n", strerror(errno));
+		goto error;
+	}
+
+	umask(old_mask);
+	(void) dm_prepare_selinux_context(NULL, 0);
+	return 0;
+error:
+	umask(old_mask);
+	(void) dm_prepare_selinux_context(NULL, 0);
+	close_comms();
+	return -1;
+}
+
+static int _init_cluster(void)
+{
+	int r;
+
+	if (!(_locks = dm_hash_create(128))) {
+		DEBUGLOG("Failed to allocate single-node hash table.\n");
+		return 1;
+	}
+
+	r = init_comms();
+	if (r) {
+		dm_hash_destroy(_locks);
+		return r;
+	}
+
+	DEBUGLOG("Single-node cluster initialised.\n");
+	return 0;
+}
+
+static void _cluster_closedown(void)
+{
+	close_comms();
+
+	DEBUGLOG("cluster_closedown\n");
+	destroy_lvhash();
+	dm_hash_destroy(_locks);
+	_locks = NULL;
+	_lockid = 0;
+}
+
+static void _get_our_csid(char *csid)
+{
+	int nodeid = 1;
+	memcpy(csid, &nodeid, sizeof(int));
+}
+
+static int _csid_from_name(char *csid, const char *name)
+{
+	return 1;
+}
+
+static int _name_from_csid(const char *csid, char *name)
+{
+	strcpy(name, "SINGLENODE");
+	return 0;
+}
+
+static int _get_num_nodes(void)
+{
+	return 1;
+}
+
+/* Node is now known to be running a clvmd */
+static void _add_up_node(const char *csid)
+{
+}
+
+/* Call a callback for each node, so the caller knows whether it's up or down */
+static int _cluster_do_node_callback(struct local_client *master_client,
+				     void (*callback)(struct local_client *,
+				     const char *csid, int node_up))
+{
+	return 0;
+}
+
+int _lock_file(const char *file, uint32_t flags);
+
+static pthread_mutex_t _lock_mutex = PTHREAD_MUTEX_INITIALIZER;
+/* Using one common condition for all locks for simplicity */
+static pthread_cond_t _lock_cond = PTHREAD_COND_INITIALIZER;
+
+/* Real locking */
+static int _lock_resource(const char *resource, int mode, int flags, int *lockid)
+{
+	struct lock *lck;
+
+	DEBUGLOG("Locking resource %s, flags=%d, mode=%d\n",
+		 resource, flags, mode);
+
+	mode &= LCK_TYPE_MASK;
+	pthread_mutex_lock(&_lock_mutex);
+retry:
+	if (!(lck = dm_hash_lookup(_locks, resource))) {
+		/* Add new locked resource */
+		if (!(lck = dm_zalloc(sizeof(struct lock))) ||
+		    !dm_hash_insert(_locks, resource, lck))
+			goto bad;
+
+		lck->lockid = ++_lockid;
+		goto out;
+	}
+
+        /* Update/convert lock */
+	if (flags == LCKF_CONVERT) {
+		if (lck->excl)
+			mode = LCK_EXCL;
+	} else if ((lck->mode == LCK_WRITE) || (lck->mode == LCK_EXCL)) {
+		DEBUGLOG("Resource %s already %s locked (%d)...\n", resource,
+			 (lck->mode == LCK_WRITE) ? "write" : "exclusively", lck->lockid);
+		goto maybe_retry;
+	} else if (lck->mode > mode) {
+		DEBUGLOG("Resource %s already locked and %s lock requested...\n",
+			 resource,
+			 (mode == LCK_READ) ? "READ" :
+			 (mode == LCK_WRITE) ? "WRITE" : "EXCLUSIVE");
+		goto maybe_retry;
+	}
+
+out:
+	*lockid = lck->lockid;
+	lck->mode = mode;
+	lck->excl |= (mode == LCK_EXCL);
+	DEBUGLOG("Locked resource %s, lockid=%d, mode=%d\n", resource, lck->lockid, mode);
+	pthread_cond_broadcast(&_lock_cond); /* wakeup waiters */
+	pthread_mutex_unlock(&_lock_mutex);
+
+	return 0;
+
+maybe_retry:
+	if (!(flags & LCK_NONBLOCK)) {
+		pthread_cond_wait(&_lock_cond, &_lock_mutex);
+		DEBUGLOG("Resource %s RETRYING lock...\n", resource);
+		goto retry;
+	}
+bad:
+	pthread_mutex_unlock(&_lock_mutex);
+	DEBUGLOG("Failed to lock resource %s\n", resource);
+
+	return 1; /* fail */
+}
+
+static int _unlock_resource(const char *resource, int lockid)
+{
+	struct lock *lck;
+
+	if (lockid < 0) {
+		DEBUGLOG("Not tracking unlock of lockid -1: %s, lockid=%d\n",
+			 resource, lockid);
+		return 0;
+	}
+
+	DEBUGLOG("Unlocking resource %s, lockid=%d\n", resource, lockid);
+	pthread_mutex_lock(&_lock_mutex);
+
+	if (!(lck = dm_hash_lookup(_locks, resource))) {
+		pthread_mutex_unlock(&_lock_mutex);
+		DEBUGLOG("Resource %s, lockid=%d is not locked.\n", resource, lockid);
+		return 1;
+	}
+
+	if (lck->lockid != lockid) {
+		pthread_mutex_unlock(&_lock_mutex);
+		DEBUGLOG("Resource %s has wrong lockid %d, expected %d.\n",
+			 resource, lck->lockid, lockid);
+		return 1;
+	}
+
+	dm_hash_remove(_locks, resource);
+	pthread_cond_broadcast(&_lock_cond); /* wakeup waiters */
+	pthread_mutex_unlock(&_lock_mutex);
+	dm_free(lck);
+
+	return 0;
+}
+
+static int _is_quorate(void)
+{
+	return 1;
+}
+
+static int _get_main_cluster_fd(void)
+{
+	return listen_fd;
+}
+
+static int _cluster_fd_callback(struct local_client *fd, char *buf, int len,
+				const char *csid,
+				struct local_client **new_client)
+{
+	return 1;
+}
+
+static int _cluster_send_message(const void *buf, int msglen,
+				 const char *csid,
+				 const char *errtext)
+{
+	return 0;
+}
+
+static int _get_cluster_name(char *buf, int buflen)
+{
+	return dm_strncpy(buf, "localcluster", buflen) ? 0 : 1;
+}
+
+static struct cluster_ops _cluster_singlenode_ops = {
+	.name                     = "singlenode",
+	.cluster_init_completed   = NULL,
+	.cluster_send_message     = _cluster_send_message,
+	.name_from_csid           = _name_from_csid,
+	.csid_from_name           = _csid_from_name,
+	.get_num_nodes            = _get_num_nodes,
+	.cluster_fd_callback      = _cluster_fd_callback,
+	.get_main_cluster_fd      = _get_main_cluster_fd,
+	.cluster_do_node_callback = _cluster_do_node_callback,
+	.is_quorate               = _is_quorate,
+	.get_our_csid             = _get_our_csid,
+	.add_up_node              = _add_up_node,
+	.reread_config            = NULL,
+	.cluster_closedown        = _cluster_closedown,
+	.get_cluster_name         = _get_cluster_name,
+	.sync_lock                = _lock_resource,
+	.sync_unlock              = _unlock_resource,
+};
+
+struct cluster_ops *init_singlenode_cluster(void)
+{
+	if (!_init_cluster())
+		return &_cluster_singlenode_ops;
+
+	return NULL;
+}

daemons/clvmd/clvmd.h

@@ -0,0 +1,127 @@
+/*
+ * Copyright (C) 2002-2004 Sistina Software, Inc. All rights reserved.
+ * Copyright (C) 2004 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU General Public License v.2.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#ifndef _CLVMD_H
+#define _CLVMD_H
+
+#define CLVMD_MAJOR_VERSION 0
+#define CLVMD_MINOR_VERSION 2
+#define CLVMD_PATCH_VERSION 1
+
+/* Default time (in seconds) we will wait for all remote commands to execute
+   before declaring them dead */
+#define DEFAULT_CMD_TIMEOUT 60
+
+/* One of these for each reply we get from command execution on a node */
+struct node_reply {
+	char node[MAX_CLUSTER_MEMBER_NAME_LEN];
+	char *replymsg;
+	int status;
+	struct node_reply *next;
+};
+
+typedef enum {DEBUG_OFF, DEBUG_STDERR, DEBUG_SYSLOG} debug_t;
+
+/*
+ * These exist for the use of local sockets only when we are
+ * collecting responses from all cluster nodes
+ */
+struct localsock_bits {
+	struct node_reply *replies;
+	int num_replies;
+	int expected_replies;
+	time_t sent_time;	/* So we can check for timeouts */
+	int in_progress;	/* Only execute one cmd at a time per client */
+	int sent_out;		/* Flag to indicate that a command was sent
+				   to remote nodes */
+	void *private;		/* Private area for command processor use */
+	void *cmd;		/* Whole command as passed down local socket */
+	int cmd_len;		/* Length of above */
+	int pipe;		/* Pipe to send PRE completion status down */
+	int finished;		/* Flag to tell subthread to exit */
+	int all_success;	/* Set to 0 if any node (or the pre_command)
+				   failed */
+	struct local_client *pipe_client;
+	pthread_t threadid;
+	enum { PRE_COMMAND, POST_COMMAND, QUIT } state;
+	pthread_mutex_t mutex;	/* Main thread and worker synchronisation */
+	pthread_cond_t cond;
+
+	pthread_mutex_t reply_mutex;	/* Protect reply structure */
+};
+
+/* Entries for PIPE clients */
+struct pipe_bits {
+	struct local_client *client;	/* Actual (localsock) client */
+	pthread_t threadid;		/* Our own copy of the thread id */
+};
+
+/* Entries for Network socket clients */
+struct netsock_bits {
+	void *private;
+	int flags;
+};
+
+typedef int (*fd_callback_t) (struct local_client * fd, char *buf, int len,
+			      const char *csid,
+			      struct local_client ** new_client);
+
+/* One of these for each fd we are listening on */
+struct local_client {
+	int fd;
+	enum { CLUSTER_MAIN_SOCK, CLUSTER_DATA_SOCK, LOCAL_RENDEZVOUS,
+		    LOCAL_SOCK, THREAD_PIPE, CLUSTER_INTERNAL } type;
+	struct local_client *next;
+	unsigned short xid;
+	fd_callback_t callback;
+	uint8_t removeme;
+
+	union {
+		struct localsock_bits localsock;
+		struct pipe_bits pipe;
+		struct netsock_bits net;
+	} bits;
+};
+
+#define DEBUGLOG(fmt, args...) debuglog(fmt, ## args)
+
+#ifndef max
+#define max(a,b) ((a)>(b)?(a):(b))
+#endif
+
+/* The real command processor is in clvmd-command.c */
+extern int do_command(struct local_client *client, struct clvm_header *msg,
+		      int msglen, char **buf, int buflen, int *retlen);
+
+/* Pre and post command routines are called only on the local node */
+extern int do_pre_command(struct local_client *client);
+extern int do_post_command(struct local_client *client);
+extern void cmd_client_cleanup(struct local_client *client);
+extern int add_client(struct local_client *new_client);
+
+extern void clvmd_cluster_init_completed(void);
+extern void process_message(struct local_client *client, char *buf,
+			    int len, const char *csid);
+extern void debuglog(const char *fmt, ... )
+  __attribute__ ((format(printf, 1, 2)));
+
+void clvmd_set_debug(debug_t new_de);
+debug_t clvmd_get_debug(void);
+int clvmd_get_foreground(void);
+
+int sync_lock(const char *resource, int mode, int flags, int *lockid);
+int sync_unlock(const char *resource, int lockid);
+
+#endif

daemons/clvmd/lvm-functions.c

@@ -0,0 +1,921 @@
+/*
+ * Copyright (C) 2002-2004 Sistina Software, Inc. All rights reserved.
+ * Copyright (C) 2004-2012 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU General Public License v.2.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include "clvmd-common.h"
+
+#include <pthread.h>
+
+#include "lvm-types.h"
+#include "clvm.h"
+#include "clvmd-comms.h"
+#include "clvmd.h"
+#include "lvm-functions.h"
+
+/* LVM2 headers */
+#include "toolcontext.h"
+#include "lvmcache.h"
+#include "lvm-globals.h"
+#include "activate.h"
+#include "archiver.h"
+#include "memlock.h"
+
+#include <syslog.h>
+
+static struct cmd_context *cmd = NULL;
+static struct dm_hash_table *lv_hash = NULL;
+static pthread_mutex_t lv_hash_lock;
+static pthread_mutex_t lvm_lock;
+static char last_error[1024];
+
+struct lv_info {
+	int lock_id;
+	int lock_mode;
+};
+
+static const char *decode_full_locking_cmd(uint32_t cmdl)
+{
+	static char buf[128];
+	const char *type;
+	const char *scope;
+	const char *command;
+
+	switch (cmdl & LCK_TYPE_MASK) {
+	case LCK_NULL:
+		type = "NULL";
+		break;
+	case LCK_READ:
+		type = "READ";
+		break;
+	case LCK_PREAD:
+		type = "PREAD";
+		break;
+	case LCK_WRITE:
+		type = "WRITE";
+		break;
+	case LCK_EXCL:
+		type = "EXCL";
+		break;
+	case LCK_UNLOCK:
+		type = "UNLOCK";
+		break;
+	default:
+		type = "unknown";
+		break;
+	}
+
+	switch (cmdl & LCK_SCOPE_MASK) {
+	case LCK_VG:
+		scope = "VG";
+		command = "LCK_VG";
+		break;
+	case LCK_LV:
+		scope = "LV";
+		switch (cmdl & LCK_MASK) {
+		case LCK_LV_EXCLUSIVE & LCK_MASK:
+			command = "LCK_LV_EXCLUSIVE";
+			break;
+		case LCK_LV_SUSPEND & LCK_MASK:
+			command = "LCK_LV_SUSPEND";
+			break;
+		case LCK_LV_RESUME & LCK_MASK:
+			command = "LCK_LV_RESUME";
+			break;
+		case LCK_LV_ACTIVATE & LCK_MASK:
+			command = "LCK_LV_ACTIVATE";
+			break;
+		case LCK_LV_DEACTIVATE & LCK_MASK:
+			command = "LCK_LV_DEACTIVATE";
+			break;
+		default:
+			command = "unknown";
+			break;
+		}
+		break;
+	default:
+		scope = "unknown";
+		command = "unknown";
+		break;
+	}
+
+	sprintf(buf, "0x%x %s (%s|%s%s%s%s%s)", cmdl, command, type, scope,
+		cmdl & LCK_NONBLOCK   ? "|NONBLOCK" : "",
+		cmdl & LCK_HOLD       ? "|HOLD" : "",
+		cmdl & LCK_CLUSTER_VG ? "|CLUSTER_VG" : "",
+		cmdl & LCK_CACHE      ? "|CACHE" : "");
+
+	return buf;
+}
+
+/*
+ * Only processes 8 bits: excludes LCK_CACHE.
+ */
+static const char *decode_locking_cmd(unsigned char cmdl)
+{
+	return decode_full_locking_cmd((uint32_t) cmdl);
+}
+
+static const char *decode_flags(unsigned char flags)
+{
+	static char buf[128];
+	int len;
+
+	len = sprintf(buf, "0x%x ( %s%s%s%s%s%s%s)", flags,
+		flags & LCK_PARTIAL_MODE	  ? "PARTIAL_MODE|" : "",
+		flags & LCK_MIRROR_NOSYNC_MODE	  ? "MIRROR_NOSYNC|" : "",
+		flags & LCK_DMEVENTD_MONITOR_MODE ? "DMEVENTD_MONITOR|" : "",
+		flags & LCK_ORIGIN_ONLY_MODE ? "ORIGIN_ONLY|" : "",
+		flags & LCK_TEST_MODE ? "TEST|" : "",
+		flags & LCK_CONVERT ? "CONVERT|" : "",
+		flags & LCK_DMEVENTD_MONITOR_IGNORE ? "DMEVENTD_MONITOR_IGNORE|" : "");
+
+	if (len > 1)
+		buf[len - 2] = ' ';
+	else
+		buf[0] = '\0';
+
+	return buf;
+}
+
+char *get_last_lvm_error(void)
+{
+	return last_error;
+}
+
+/*
+ * Hash lock info helpers
+ */
+static struct lv_info *lookup_info(const char *resource)
+{
+	struct lv_info *lvi;
+
+	pthread_mutex_lock(&lv_hash_lock);
+	lvi = dm_hash_lookup(lv_hash, resource);
+	pthread_mutex_unlock(&lv_hash_lock);
+
+	return lvi;
+}
+
+static int insert_info(const char *resource, struct lv_info *lvi)
+{
+	int ret;
+
+	pthread_mutex_lock(&lv_hash_lock);
+	ret = dm_hash_insert(lv_hash, resource, lvi);
+	pthread_mutex_unlock(&lv_hash_lock);
+
+	return ret;
+}
+
+static void remove_info(const char *resource)
+{
+	pthread_mutex_lock(&lv_hash_lock);
+	dm_hash_remove(lv_hash, resource);
+	pthread_mutex_unlock(&lv_hash_lock);
+}
+
+/*
+ * Return the mode a lock is currently held at (or -1 if not held)
+ */
+static int get_current_lock(char *resource)
+{
+	struct lv_info *lvi;
+
+	if ((lvi = lookup_info(resource)))
+		return lvi->lock_mode;
+
+	return -1;
+}
+
+
+void init_lvhash(void)
+{
+	/* Create hash table for keeping LV locks & status */
+	lv_hash = dm_hash_create(1024);
+	pthread_mutex_init(&lv_hash_lock, NULL);
+	pthread_mutex_init(&lvm_lock, NULL);
+}
+
+/* Called at shutdown to tidy the lockspace */
+void destroy_lvhash(void)
+{
+	struct dm_hash_node *v;
+	struct lv_info *lvi;
+	char *resource;
+	int status;
+
+	pthread_mutex_lock(&lv_hash_lock);
+
+	dm_hash_iterate(v, lv_hash) {
+		lvi = dm_hash_get_data(lv_hash, v);
+		resource = dm_hash_get_key(lv_hash, v);
+
+		if ((status = sync_unlock(resource, lvi->lock_id)))
+			DEBUGLOG("unlock_all. unlock failed(%d): %s\n",
+				 status,  strerror(errno));
+		free(lvi);
+	}
+
+	dm_hash_destroy(lv_hash);
+	lv_hash = NULL;
+
+	pthread_mutex_unlock(&lv_hash_lock);
+}
+
+/* Gets a real lock and keeps the info in the hash table */
+static int hold_lock(char *resource, int mode, int flags)
+{
+	int status;
+	int saved_errno;
+	struct lv_info *lvi;
+
+	if (test_mode())
+		return 0;
+
+	/* Mask off invalid options */
+	flags &= LCKF_NOQUEUE | LCKF_CONVERT;
+
+	lvi = lookup_info(resource);
+
+	if (lvi) {
+		if (lvi->lock_mode == mode) {
+			DEBUGLOG("hold_lock, lock mode %d already held\n",
+				 mode);
+			return 0;
+		}
+		if ((lvi->lock_mode == LCK_EXCL) && (mode == LCK_WRITE)) {
+			DEBUGLOG("hold_lock, lock already held LCK_EXCL, "
+				 "ignoring LCK_WRITE request");
+			return 0;
+		}
+	}
+
+	/* Only allow explicit conversions */
+	if (lvi && !(flags & LCKF_CONVERT)) {
+		errno = EBUSY;
+		return -1;
+	}
+	if (lvi) {
+		/* Already exists - convert it */
+		status =
+		    sync_lock(resource, mode, flags, &lvi->lock_id);
+		saved_errno = errno;
+		if (!status)
+			lvi->lock_mode = mode;
+
+		if (status) {
+			DEBUGLOG("hold_lock. convert to %d failed: %s\n", mode,
+				 strerror(errno));
+		}
+		errno = saved_errno;
+	} else {
+		lvi = malloc(sizeof(struct lv_info));
+		if (!lvi) {
+			errno = ENOMEM;
+			return -1;
+		}
+
+		lvi->lock_mode = mode;
+		status = sync_lock(resource, mode, flags & ~LCKF_CONVERT, &lvi->lock_id);
+		saved_errno = errno;
+		if (status) {
+			free(lvi);
+			DEBUGLOG("hold_lock. lock at %d failed: %s\n", mode,
+				 strerror(errno));
+		} else
+			if (!insert_info(resource, lvi)) {
+				errno = ENOMEM;
+				return -1;
+			}
+
+		errno = saved_errno;
+	}
+	return status;
+}
+
+/* Unlock and remove it from the hash table */
+static int hold_unlock(char *resource)
+{
+	struct lv_info *lvi;
+	int status;
+	int saved_errno;
+
+	if (test_mode())
+		return 0;
+
+	if (!(lvi = lookup_info(resource))) {
+		DEBUGLOG("hold_unlock, lock not already held\n");
+		return 0;
+	}
+
+	status = sync_unlock(resource, lvi->lock_id);
+	saved_errno = errno;
+	if (!status) {
+		remove_info(resource);
+		free(lvi);
+	} else {
+		DEBUGLOG("hold_unlock. unlock failed(%d): %s\n", status,
+			 strerror(errno));
+	}
+
+	errno = saved_errno;
+	return status;
+}
+
+/* Watch the return codes here.
+   liblvm API functions return 1(true) for success, 0(false) for failure and don't set errno.
+   libdlm API functions return 0 for success, -1 for failure and do set errno.
+   These functions here return 0 for success or >0 for failure (where the retcode is errno)
+*/
+
+/* Activate LV exclusive or non-exclusive */
+static int do_activate_lv(char *resource, unsigned char command, unsigned char lock_flags, int mode)
+{
+	int oldmode;
+	int status;
+	int activate_lv;
+	int exclusive = 0;
+	struct lvinfo lvi;
+
+	/* Is it already open ? */
+	oldmode = get_current_lock(resource);
+	if (oldmode == mode && (command & LCK_CLUSTER_VG)) {
+		DEBUGLOG("do_activate_lv, lock already held at %d\n", oldmode);
+		return 0;	/* Nothing to do */
+	}
+
+	/* Does the config file want us to activate this LV ? */
+	if (!lv_activation_filter(cmd, resource, &activate_lv))
+		return EIO;
+
+	if (!activate_lv)
+		return 0;	/* Success, we did nothing! */
+
+	/* Do we need to activate exclusively? */
+	if ((activate_lv == 2) || (mode == LCK_EXCL)) {
+		exclusive = 1;
+		mode = LCK_EXCL;
+	}
+
+	/*
+	 * Try to get the lock if it's a clustered volume group.
+	 * Use lock conversion only if requested, to prevent implicit conversion
+	 * of exclusive lock to shared one during activation.
+	 */
+	if (command & LCK_CLUSTER_VG) {
+		status = hold_lock(resource, mode, LCKF_NOQUEUE | (lock_flags & LCK_CONVERT ? LCKF_CONVERT:0));
+		if (status) {
+			/* Return an LVM-sensible error for this.
+			 * Forcing EIO makes the upper level return this text
+			 * rather than the strerror text for EAGAIN.
+			 */
+			if (errno == EAGAIN) {
+				sprintf(last_error, "Volume is busy on another node");
+				errno = EIO;
+			}
+			return errno;
+		}
+	}
+
+	/* If it's suspended then resume it */
+	if (!lv_info_by_lvid(cmd, resource, 0, &lvi, 0, 0))
+		goto error;
+
+	if (lvi.suspended) {
+		critical_section_inc(cmd, "resuming");
+		if (!lv_resume(cmd, resource, 0)) {
+			critical_section_dec(cmd, "resumed");
+			goto error;
+		}
+	}
+
+	/* Now activate it */
+	if (!lv_activate(cmd, resource, exclusive))
+		goto error;
+
+	return 0;
+
+error:
+	if (oldmode == -1 || oldmode != mode)
+		(void)hold_unlock(resource);
+	return EIO;
+}
+
+/* Resume the LV if it was active */
+static int do_resume_lv(char *resource, unsigned char command, unsigned char lock_flags)
+{
+	int oldmode, origin_only, exclusive, revert;
+
+	/* Is it open ? */
+	oldmode = get_current_lock(resource);
+	if (oldmode == -1 && (command & LCK_CLUSTER_VG)) {
+		DEBUGLOG("do_resume_lv, lock not already held\n");
+		return 0;	/* We don't need to do anything */
+	}
+	origin_only = (lock_flags & LCK_ORIGIN_ONLY_MODE) ? 1 : 0;
+	exclusive = (oldmode == LCK_EXCL) ? 1 : 0;
+	revert = (lock_flags & LCK_REVERT_MODE) ? 1 : 0;
+
+	if (!lv_resume_if_active(cmd, resource, origin_only, exclusive, revert))
+		return EIO;
+
+	return 0;
+}
+
+/* Suspend the device if active */
+static int do_suspend_lv(char *resource, unsigned char command, unsigned char lock_flags)
+{
+	int oldmode;
+	unsigned origin_only = (lock_flags & LCK_ORIGIN_ONLY_MODE) ? 1 : 0;
+	unsigned exclusive;
+
+	/* Is it open ? */
+	oldmode = get_current_lock(resource);
+	if (oldmode == -1 && (command & LCK_CLUSTER_VG)) {
+		DEBUGLOG("do_suspend_lv, lock not already held\n");
+		return 0; /* Not active, so it's OK */
+	}
+
+	exclusive = (oldmode == LCK_EXCL) ? 1 : 0;
+
+	/* Always call lv_suspend to read commited and precommited data */
+	if (!lv_suspend_if_active(cmd, resource, origin_only, exclusive))
+		return EIO;
+
+	return 0;
+}
+
+static int do_deactivate_lv(char *resource, unsigned char command, unsigned char lock_flags)
+{
+	int oldmode;
+	int status;
+
+	/* Is it open ? */
+	oldmode = get_current_lock(resource);
+	if (oldmode == -1 && (command & LCK_CLUSTER_VG)) {
+		DEBUGLOG("do_deactivate_lock, lock not already held\n");
+		return 0;	/* We don't need to do anything */
+	}
+
+	if (!lv_deactivate(cmd, resource))
+		return EIO;
+
+	if (command & LCK_CLUSTER_VG) {
+		status = hold_unlock(resource);
+		if (status)
+			return errno;
+	}
+
+	return 0;
+}
+
+const char *do_lock_query(char *resource)
+{
+	int mode;
+	const char *type = NULL;
+
+	mode = get_current_lock(resource);
+	switch (mode) {
+		case LCK_NULL: type = "NL"; break;
+		case LCK_READ: type = "CR"; break;
+		case LCK_PREAD:type = "PR"; break;
+		case LCK_WRITE:type = "PW"; break;
+		case LCK_EXCL: type = "EX"; break;
+	}
+
+	DEBUGLOG("do_lock_query: resource '%s', mode %i (%s)\n", resource, mode, type ?: "?");
+
+	return type;
+}
+
+/* This is the LOCK_LV part that happens on all nodes in the cluster -
+   it is responsible for the interaction with device-mapper and LVM */
+int do_lock_lv(unsigned char command, unsigned char lock_flags, char *resource)
+{
+	int status = 0;
+
+	DEBUGLOG("do_lock_lv: resource '%s', cmd = %s, flags = %s, critical_section = %d\n",
+		 resource, decode_locking_cmd(command), decode_flags(lock_flags), critical_section());
+
+	if (!cmd->config_valid || config_files_changed(cmd)) {
+		/* Reinitialise various settings inc. logging, filters */
+		if (do_refresh_cache()) {
+			log_error("Updated config file invalid. Aborting.");
+			return EINVAL;
+		}
+	}
+
+	pthread_mutex_lock(&lvm_lock);
+	if (lock_flags & LCK_MIRROR_NOSYNC_MODE)
+		init_mirror_in_sync(1);
+
+	if (lock_flags & LCK_DMEVENTD_MONITOR_IGNORE)
+		init_dmeventd_monitor(DMEVENTD_MONITOR_IGNORE);
+	else {
+		if (lock_flags & LCK_DMEVENTD_MONITOR_MODE)
+			init_dmeventd_monitor(1);
+		else
+			init_dmeventd_monitor(0);
+	}
+
+	cmd->partial_activation = (lock_flags & LCK_PARTIAL_MODE) ? 1 : 0;
+
+	/* clvmd should never try to read suspended device */
+	init_ignore_suspended_devices(1);
+
+	switch (command & LCK_MASK) {
+	case LCK_LV_EXCLUSIVE:
+		status = do_activate_lv(resource, command, lock_flags, LCK_EXCL);
+		break;
+
+	case LCK_LV_SUSPEND:
+		status = do_suspend_lv(resource, command, lock_flags);
+		break;
+
+	case LCK_UNLOCK:
+	case LCK_LV_RESUME:	/* if active */
+		status = do_resume_lv(resource, command, lock_flags);
+		break;
+
+	case LCK_LV_ACTIVATE:
+		status = do_activate_lv(resource, command, lock_flags, LCK_READ);
+		break;
+
+	case LCK_LV_DEACTIVATE:
+		status = do_deactivate_lv(resource, command, lock_flags);
+		break;
+
+	default:
+		DEBUGLOG("Invalid LV command 0x%x\n", command);
+		status = EINVAL;
+		break;
+	}
+
+	if (lock_flags & LCK_MIRROR_NOSYNC_MODE)
+		init_mirror_in_sync(0);
+
+	cmd->partial_activation = 0;
+
+	/* clean the pool for another command */
+	dm_pool_empty(cmd->mem);
+	pthread_mutex_unlock(&lvm_lock);
+
+	DEBUGLOG("Command return is %d, critical_section is %d\n", status, critical_section());
+	return status;
+}
+
+/* Functions to do on the local node only BEFORE the cluster-wide stuff above happens */
+int pre_lock_lv(unsigned char command, unsigned char lock_flags, char *resource)
+{
+	/* Nearly all the stuff happens cluster-wide. Apart from SUSPEND. Here we get the
+	   lock out on this node (because we are the node modifying the metadata)
+	   before suspending cluster-wide.
+	   LCKF_CONVERT is used always, local node is going to modify metadata
+	 */
+	if ((command & (LCK_SCOPE_MASK | LCK_TYPE_MASK)) == LCK_LV_SUSPEND &&
+	    (command & LCK_CLUSTER_VG)) {
+		DEBUGLOG("pre_lock_lv: resource '%s', cmd = %s, flags = %s\n",
+			 resource, decode_locking_cmd(command), decode_flags(lock_flags));
+
+		if (hold_lock(resource, LCK_WRITE, LCKF_NOQUEUE | LCKF_CONVERT))
+			return errno;
+	}
+	return 0;
+}
+
+/* Functions to do on the local node only AFTER the cluster-wide stuff above happens */
+int post_lock_lv(unsigned char command, unsigned char lock_flags,
+		 char *resource)
+{
+	int status;
+	unsigned origin_only = (lock_flags & LCK_ORIGIN_ONLY_MODE) ? 1 : 0;
+
+	/* Opposite of above, done on resume after a metadata update */
+	if ((command & (LCK_SCOPE_MASK | LCK_TYPE_MASK)) == LCK_LV_RESUME &&
+	    (command & LCK_CLUSTER_VG)) {
+		int oldmode;
+
+		DEBUGLOG
+		    ("post_lock_lv: resource '%s', cmd = %s, flags = %s\n",
+		     resource, decode_locking_cmd(command), decode_flags(lock_flags));
+
+		/* If the lock state is PW then restore it to what it was */
+		oldmode = get_current_lock(resource);
+		if (oldmode == LCK_WRITE) {
+			struct lvinfo lvi;
+
+			pthread_mutex_lock(&lvm_lock);
+			status = lv_info_by_lvid(cmd, resource, origin_only, &lvi, 0, 0);
+			pthread_mutex_unlock(&lvm_lock);
+			if (!status)
+				return EIO;
+
+			if (lvi.exists) {
+				if (hold_lock(resource, LCK_READ, LCKF_CONVERT))
+					return errno;
+			} else if (hold_unlock(resource))
+				return errno;
+		}
+	}
+	return 0;
+}
+
+/* Check if a VG is in use by LVM1 so we don't stomp on it */
+int do_check_lvm1(const char *vgname)
+{
+	int status;
+
+	status = check_lvm1_vg_inactive(cmd, vgname);
+
+	return status == 1 ? 0 : EBUSY;
+}
+
+int do_refresh_cache(void)
+{
+	DEBUGLOG("Refreshing context\n");
+	log_notice("Refreshing context");
+
+	pthread_mutex_lock(&lvm_lock);
+
+	if (!refresh_toolcontext(cmd)) {
+		pthread_mutex_unlock(&lvm_lock);
+		return -1;
+	}
+
+	init_full_scan_done(0);
+	init_ignore_suspended_devices(1);
+	lvmcache_label_scan(cmd, 2);
+	dm_pool_empty(cmd->mem);
+
+	pthread_mutex_unlock(&lvm_lock);
+
+	return 0;
+}
+
+/*
+ * Handle VG lock - drop metadata or update lvmcache state
+ */
+void do_lock_vg(unsigned char command, unsigned char lock_flags, char *resource)
+{
+	uint32_t lock_cmd = command;
+	char *vgname = resource + 2;
+
+	lock_cmd &= (LCK_SCOPE_MASK | LCK_TYPE_MASK | LCK_HOLD);
+
+	/*
+	 * Check if LCK_CACHE should be set. All P_ locks except # are cache related.
+	 */
+	if (strncmp(resource, "P_#", 3) && !strncmp(resource, "P_", 2))
+		lock_cmd |= LCK_CACHE;
+
+	DEBUGLOG("do_lock_vg: resource '%s', cmd = %s, flags = %s, critical_section = %d\n",
+		 resource, decode_full_locking_cmd(lock_cmd), decode_flags(lock_flags), critical_section());
+
+	/* P_#global causes a full cache refresh */
+	if (!strcmp(resource, "P_" VG_GLOBAL)) {
+		do_refresh_cache();
+		return;
+	}
+
+	pthread_mutex_lock(&lvm_lock);
+	switch (lock_cmd) {
+		case LCK_VG_COMMIT:
+			DEBUGLOG("vg_commit notification for VG %s\n", vgname);
+			lvmcache_commit_metadata(vgname);
+			break;
+		case LCK_VG_REVERT:
+			DEBUGLOG("vg_revert notification for VG %s\n", vgname);
+			lvmcache_drop_metadata(vgname, 1);
+			break;
+		case LCK_VG_DROP_CACHE:
+		default:
+			DEBUGLOG("Invalidating cached metadata for VG %s\n", vgname);
+			lvmcache_drop_metadata(vgname, 0);
+	}
+	pthread_mutex_unlock(&lvm_lock);
+}
+
+/*
+ * Ideally, clvmd should be started before any LVs are active
+ * but this may not be the case...
+ * I suppose this also comes in handy if clvmd crashes, not that it would!
+ */
+static int get_initial_state(struct dm_hash_table *excl_uuid)
+{
+	int lock_mode;
+	char lv[64], vg[64], flags[25], vg_flags[25];
+	char uuid[65];
+	char line[255];
+	char *lvs_cmd;
+	const char *lvm_binary = getenv("LVM_BINARY") ? : LVM_PATH;
+	FILE *lvs;
+
+	if (dm_asprintf(&lvs_cmd, "%s lvs  --config 'log{command_names=0 prefix=\"\"}' "
+			"--nolocking --noheadings -o vg_uuid,lv_uuid,lv_attr,vg_attr",
+			lvm_binary) < 0)
+		return_0;
+
+	/* FIXME: Maybe link and use liblvm2cmd directly instead of fork */
+	if (!(lvs = popen(lvs_cmd, "r"))) {
+		dm_free(lvs_cmd);
+		return 0;
+	}
+
+	while (fgets(line, sizeof(line), lvs)) {
+	        if (sscanf(line, "%64s %64s %25s %25s\n", vg, lv, flags, vg_flags) == 4) {
+
+			/* States: s:suspended a:active S:dropped snapshot I:invalid snapshot */
+		        if (strlen(vg) == 38 &&                         /* is is a valid UUID ? */
+			    (flags[4] == 'a' || flags[4] == 's') &&	/* is it active or suspended? */
+			    vg_flags[5] == 'c') {			/* is it clustered ? */
+				/* Convert hyphen-separated UUIDs into one */
+				memcpy(&uuid[0], &vg[0], 6);
+				memcpy(&uuid[6], &vg[7], 4);
+				memcpy(&uuid[10], &vg[12], 4);
+				memcpy(&uuid[14], &vg[17], 4);
+				memcpy(&uuid[18], &vg[22], 4);
+				memcpy(&uuid[22], &vg[27], 4);
+				memcpy(&uuid[26], &vg[32], 6);
+				memcpy(&uuid[32], &lv[0], 6);
+				memcpy(&uuid[38], &lv[7], 4);
+				memcpy(&uuid[42], &lv[12], 4);
+				memcpy(&uuid[46], &lv[17], 4);
+				memcpy(&uuid[50], &lv[22], 4);
+				memcpy(&uuid[54], &lv[27], 4);
+				memcpy(&uuid[58], &lv[32], 6);
+				uuid[64] = '\0';
+
+				/* Look for this lock in the list of EX locks
+				   we were passed on the command-line */
+				lock_mode = (dm_hash_lookup(excl_uuid, uuid)) ?
+					LCK_EXCL : LCK_READ;
+
+				DEBUGLOG("getting initial lock for %s\n", uuid);
+				if (hold_lock(uuid, lock_mode, LCKF_NOQUEUE))
+					DEBUGLOG("Failed to hold lock %s\n", uuid);
+			}
+		}
+	}
+	if (pclose(lvs))
+		DEBUGLOG("lvs pclose failed: %s\n", strerror(errno));
+
+	dm_free(lvs_cmd);
+
+	return 1;
+}
+
+static void lvm2_log_fn(int level, const char *file, int line, int dm_errno,
+			const char *message)
+{
+
+	/* Send messages to the normal LVM2 logging system too,
+	   so we get debug output when it's asked for.
+ 	   We need to NULL the function ptr otherwise it will just call
+	   back into here! */
+	init_log_fn(NULL);
+	print_log(level, file, line, dm_errno, "%s", message);
+	init_log_fn(lvm2_log_fn);
+
+	/*
+	 * Ignore non-error messages, but store the latest one for returning
+	 * to the user.
+	 */
+	if (level != _LOG_ERR && level != _LOG_FATAL)
+		return;
+
+	strncpy(last_error, message, sizeof(last_error));
+	last_error[sizeof(last_error)-1] = '\0';
+}
+
+/* This checks some basic cluster-LVM configuration stuff */
+static void check_config(void)
+{
+	int locking_type;
+
+	locking_type = find_config_tree_int(cmd, global_locking_type_CFG);
+
+	if (locking_type == 3) /* compiled-in cluster support */
+		return;
+
+	if (locking_type == 2) { /* External library, check name */
+		const char *libname;
+
+		libname = find_config_tree_str(cmd, global_locking_library_CFG);
+		if (libname && strstr(libname, "liblvm2clusterlock.so"))
+			return;
+
+		log_error("Incorrect LVM locking library specified in lvm.conf, cluster operations may not work.");
+		return;
+	}
+	log_error("locking_type not set correctly in lvm.conf, cluster operations will not work.");
+}
+
+/* Backups up the LVM metadata if it's changed */
+void lvm_do_backup(const char *vgname)
+{
+	struct volume_group * vg;
+	int consistent = 0;
+
+	DEBUGLOG("Triggering backup of VG metadata for %s.\n", vgname);
+
+	pthread_mutex_lock(&lvm_lock);
+
+	vg = vg_read_internal(cmd, vgname, NULL /*vgid*/, 1, &consistent);
+
+	if (vg && consistent)
+		check_current_backup(vg);
+	else
+		log_error("Error backing up metadata, can't find VG for group %s", vgname);
+
+	release_vg(vg);
+	dm_pool_empty(cmd->mem);
+
+	pthread_mutex_unlock(&lvm_lock);
+}
+
+struct dm_hash_node *get_next_excl_lock(struct dm_hash_node *v, char **name)
+{
+	struct lv_info *lvi;
+
+	*name = NULL;
+	if (!v)
+		v = dm_hash_get_first(lv_hash);
+
+	do {
+		if (v) {
+			lvi = dm_hash_get_data(lv_hash, v);
+			DEBUGLOG("Looking for EX locks. found %x mode %d\n", lvi->lock_id, lvi->lock_mode);
+
+			if (lvi->lock_mode == LCK_EXCL) {
+				*name = dm_hash_get_key(lv_hash, v);
+			}
+			v = dm_hash_get_next(lv_hash, v);
+		}
+	} while (v && !*name);
+
+	if (*name)
+		DEBUGLOG("returning EXclusive UUID %s\n", *name);
+	return v;
+}
+
+void lvm_do_fs_unlock(void)
+{
+	pthread_mutex_lock(&lvm_lock);
+	DEBUGLOG("Syncing device names\n");
+	fs_unlock();
+	pthread_mutex_unlock(&lvm_lock);
+}
+
+/* Called to initialise the LVM context of the daemon */
+int init_clvm(struct dm_hash_table *excl_uuid)
+{
+	/* Use LOG_DAEMON for syslog messages instead of LOG_USER */
+	init_syslog(LOG_DAEMON);
+	openlog("clvmd", LOG_PID, LOG_DAEMON);
+
+	/* Initialise already held locks */
+	if (!get_initial_state(excl_uuid))
+		log_error("Cannot load initial lock states.");
+
+	if (!(cmd = create_toolcontext(1, NULL, 0, 1))) {
+		log_error("Failed to allocate command context");
+		return 0;
+	}
+
+	if (stored_errno()) {
+		destroy_toolcontext(cmd);
+		return 0;
+	}
+
+	cmd->cmd_line = "clvmd";
+
+	/* Check lvm.conf is setup for cluster-LVM */
+	check_config();
+	init_ignore_suspended_devices(1);
+
+	/* Trap log messages so we can pass them back to the user */
+	init_log_fn(lvm2_log_fn);
+	memlock_inc_daemon(cmd);
+
+	return 1;
+}
+
+void destroy_lvm(void)
+{
+	if (cmd) {
+		memlock_dec_daemon(cmd);
+		destroy_toolcontext(cmd);
+	}
+	cmd = NULL;
+}

daemons/clvmd/lvm-functions.h

@@ -0,0 +1,41 @@
+/*
+ * Copyright (C) 2002-2004 Sistina Software, Inc. All rights reserved.
+ * Copyright (C) 2004-2010 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU General Public License v.2.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+/* Functions in lvm-functions.c */
+
+#ifndef _LVM_FUNCTIONS_H
+#define _LVM_FUNCTIONS_H
+
+extern int pre_lock_lv(unsigned char lock_cmd, unsigned char lock_flags,
+		       char *resource);
+extern int do_lock_lv(unsigned char lock_cmd, unsigned char lock_flags,
+		      char *resource);
+extern const char *do_lock_query(char *resource);
+extern int post_lock_lv(unsigned char lock_cmd, unsigned char lock_flags,
+			char *resource);
+extern int do_check_lvm1(const char *vgname);
+extern int do_refresh_cache(void);
+extern int init_clvm(struct dm_hash_table *excl_uuid);
+extern void destroy_lvm(void);
+extern void init_lvhash(void);
+extern void destroy_lvhash(void);
+extern void lvm_do_backup(const char *vgname);
+extern char *get_last_lvm_error(void);
+extern void do_lock_vg(unsigned char command, unsigned char lock_flags,
+		      char *resource);
+extern struct dm_hash_node *get_next_excl_lock(struct dm_hash_node *v, char **name);
+void lvm_do_fs_unlock(void);
+
+#endif

daemons/clvmd/refresh_clvmd.c

@@ -0,0 +1,382 @@
+/*
+ * Copyright (C) 2002-2004 Sistina Software, Inc. All rights reserved.
+ * Copyright (C) 2004-2010 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU General Public License v.2.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+/* FIXME Remove duplicated functions from this file. */
+
+/*
+ * Send a command to a running clvmd from the command-line
+ */
+
+#include "clvmd-common.h"
+
+#include "clvm.h"
+#include "refresh_clvmd.h"
+
+#include <stddef.h>
+#include <sys/socket.h>
+#include <sys/un.h>
+
+typedef struct lvm_response {
+	char node[255];
+	char *response;
+	int status;
+	int len;
+} lvm_response_t;
+
+/*
+ * This gets stuck at the start of memory we allocate so we
+ * can sanity-check it at deallocation time
+ */
+#define LVM_SIGNATURE 0x434C564D
+
+static int _clvmd_sock = -1;
+
+/* Open connection to the clvm daemon */
+static int _open_local_sock(void)
+{
+	int local_socket;
+	struct sockaddr_un sockaddr = { .sun_family = AF_UNIX };
+
+	if (!dm_strncpy(sockaddr.sun_path, CLVMD_SOCKNAME, sizeof(sockaddr.sun_path))) {
+		fprintf(stderr, "%s: clvmd socket name too long.", CLVMD_SOCKNAME);
+		return -1;
+	}
+
+	/* Open local socket */
+	if ((local_socket = socket(PF_UNIX, SOCK_STREAM, 0)) < 0) {
+		fprintf(stderr, "Local socket creation failed: %s", strerror(errno));
+		return -1;
+	}
+
+	if (connect(local_socket,(struct sockaddr *) &sockaddr,
+		    sizeof(sockaddr))) {
+		int saved_errno = errno;
+
+		fprintf(stderr, "connect() failed on local socket: %s\n",
+			  strerror(errno));
+		if (close(local_socket))
+			return -1;
+
+		errno = saved_errno;
+		return -1;
+	}
+
+	return local_socket;
+}
+
+/* Send a request and return the status */
+static int _send_request(const char *inbuf, int inlen, char **retbuf, int no_response)
+{
+	char outbuf[PIPE_BUF];
+	struct clvm_header *outheader = (struct clvm_header *) outbuf;
+	int len;
+	unsigned off;
+	int buflen;
+	int err;
+
+	/* Send it to CLVMD */
+ rewrite:
+	if ( (err = write(_clvmd_sock, inbuf, inlen)) != inlen) {
+	        if (err == -1 && errno == EINTR)
+		        goto rewrite;
+		fprintf(stderr, "Error writing data to clvmd: %s", strerror(errno));
+		return 0;
+	}
+	if (no_response)
+		return 1;
+
+	/* Get the response */
+ reread:
+	if ((len = read(_clvmd_sock, outbuf, sizeof(struct clvm_header))) < 0) {
+	        if (errno == EINTR)
+		        goto reread;
+		fprintf(stderr, "Error reading data from clvmd: %s", strerror(errno));
+		return 0;
+	}
+
+	if (len == 0) {
+		fprintf(stderr, "EOF reading CLVMD");
+		errno = ENOTCONN;
+		return 0;
+	}
+
+	/* Allocate buffer */
+	buflen = len + outheader->arglen;
+	*retbuf = dm_malloc(buflen);
+	if (!*retbuf) {
+		errno = ENOMEM;
+		return 0;
+	}
+
+	/* Copy the header */
+	memcpy(*retbuf, outbuf, len);
+	outheader = (struct clvm_header *) *retbuf;
+
+	/* Read the returned values */
+	off = 1;		/* we've already read the first byte */
+	while (off <= outheader->arglen && len > 0) {
+		len = read(_clvmd_sock, outheader->args + off,
+			   buflen - off - offsetof(struct clvm_header, args));
+		if (len > 0)
+			off += len;
+	}
+
+	/* Was it an error ? */
+	if (outheader->status != 0) {
+		errno = outheader->status;
+
+		/* Only return an error here if there are no node-specific
+		   errors present in the message that might have more detail */
+		if (!(outheader->flags & CLVMD_FLAG_NODEERRS)) {
+			fprintf(stderr, "cluster request failed: %s\n", strerror(errno));
+			return 0;
+		}
+
+	}
+
+	return 1;
+}
+
+/* Build the structure header and parse-out wildcard node names */
+static void _build_header(struct clvm_header *head, int cmd, const char *node,
+			  unsigned int len)
+{
+	head->cmd = cmd;
+	head->status = 0;
+	head->flags = 0;
+	head->xid = 0;
+	head->clientid = 0;
+	if (len)
+		/* 1 byte is used from struct clvm_header.args[1], so -> len - 1 */
+		head->arglen = len - 1;
+	else {
+		head->arglen = 0;
+		*head->args = '\0';
+	}
+
+	/*
+	 * Translate special node names.
+	 */
+	if (!node || !strcmp(node, NODE_ALL))
+		head->node[0] = '\0';
+	else if (!strcmp(node, NODE_LOCAL)) {
+		head->node[0] = '\0';
+		head->flags = CLVMD_FLAG_LOCAL;
+	} else
+		strcpy(head->node, node);
+}
+
+/*
+ * Send a message to a(or all) node(s) in the cluster and wait for replies
+ */
+static int _cluster_request(char cmd, const char *node, void *data, int len,
+			    lvm_response_t ** response, int *num, int no_response)
+{
+	char outbuf[sizeof(struct clvm_header) + len + strlen(node) + 1];
+	char *inptr;
+	char *retbuf = NULL;
+	int status;
+	int i;
+	int num_responses = 0;
+	struct clvm_header *head = (struct clvm_header *) outbuf;
+	lvm_response_t *rarray;
+
+	*num = 0;
+
+	if (_clvmd_sock == -1)
+		_clvmd_sock = _open_local_sock();
+
+	if (_clvmd_sock == -1)
+		return 0;
+
+	_build_header(head, cmd, node, len);
+	if (len)
+		memcpy(head->node + strlen(head->node) + 1, data, len);
+
+	status = _send_request(outbuf, sizeof(struct clvm_header) +
+			       strlen(head->node) + len, &retbuf, no_response);
+	if (!status || no_response)
+		goto out;
+
+	/* Count the number of responses we got */
+	head = (struct clvm_header *) retbuf;
+	inptr = head->args;
+	while (inptr[0]) {
+		num_responses++;
+		inptr += strlen(inptr) + 1;
+		inptr += sizeof(int);
+		inptr += strlen(inptr) + 1;
+	}
+
+	/*
+	 * Allocate response array.
+	 * With an extra pair of INTs on the front to sanity
+	 * check the pointer when we are given it back to free
+	 */
+	*response = NULL;
+	if (!(rarray = dm_malloc(sizeof(lvm_response_t) * num_responses +
+				 sizeof(int) * 2))) {
+		errno = ENOMEM;
+		status = 0;
+		goto out;
+	}
+
+	/* Unpack the response into an lvm_response_t array */
+	inptr = head->args;
+	i = 0;
+	while (inptr[0]) {
+		strcpy(rarray[i].node, inptr);
+		inptr += strlen(inptr) + 1;
+
+		memcpy(&rarray[i].status, inptr, sizeof(int));
+		inptr += sizeof(int);
+
+		rarray[i].response = dm_malloc(strlen(inptr) + 1);
+		if (rarray[i].response == NULL) {
+			/* Free up everything else and return error */
+			int j;
+			for (j = 0; j < i; j++)
+				dm_free(rarray[i].response);
+			dm_free(rarray);
+			errno = ENOMEM;
+			status = 0;
+			goto out;
+		}
+
+		strcpy(rarray[i].response, inptr);
+		rarray[i].len = strlen(inptr);
+		inptr += strlen(inptr) + 1;
+		i++;
+	}
+	*num = num_responses;
+	*response = rarray;
+
+      out:
+	dm_free(retbuf);
+
+	return status;
+}
+
+/* Free reply array */
+static int _cluster_free_request(lvm_response_t * response, int num)
+{
+	int i;
+
+	for (i = 0; i < num; i++) {
+		dm_free(response[i].response);
+	}
+
+	dm_free(response);
+
+	return 1;
+}
+
+int refresh_clvmd(int all_nodes)
+{
+	int num_responses;
+	char args[1]; // No args really.
+	lvm_response_t *response = NULL;
+	int saved_errno;
+	int status;
+	int i;
+
+	status = _cluster_request(CLVMD_CMD_REFRESH, all_nodes ? NODE_ALL : NODE_LOCAL, args, 0, &response, &num_responses, 0);
+
+	/* If any nodes were down then display them and return an error */
+	for (i = 0; i < num_responses; i++) {
+		if (response[i].status == EHOSTDOWN) {
+			fprintf(stderr, "clvmd not running on node %s",
+				  response[i].node);
+			status = 0;
+			errno = response[i].status;
+		} else if (response[i].status) {
+			fprintf(stderr, "Error resetting node %s: %s",
+				  response[i].node,
+				  response[i].response[0] ?
+				  	response[i].response :
+				  	strerror(response[i].status));
+			status = 0;
+			errno = response[i].status;
+		}
+	}
+
+	saved_errno = errno;
+	_cluster_free_request(response, num_responses);
+	errno = saved_errno;
+
+	return status;
+}
+
+int restart_clvmd(int all_nodes)
+{
+	int dummy, status;
+
+	status = _cluster_request(CLVMD_CMD_RESTART, all_nodes ? NODE_ALL : NODE_LOCAL, NULL, 0, NULL, &dummy, 1);
+
+	/*
+	 * FIXME: we cannot receive response, clvmd re-exec before it.
+	 *        but also should not close socket too early (the whole rq is dropped then).
+	 * FIXME: This should be handled this way:
+	 *  - client waits for RESTART ack (and socket close)
+	 *  - server restarts
+	 *  - client checks that server is ready again (VERSION command?)
+	 */
+	usleep(500000);
+
+	return status;
+}
+
+int debug_clvmd(int level, int clusterwide)
+{
+	int num_responses;
+	char args[1];
+	const char *nodes;
+	lvm_response_t *response = NULL;
+	int saved_errno;
+	int status;
+	int i;
+
+	args[0] = level;
+	if (clusterwide)
+		nodes = NODE_ALL;
+	else
+		nodes = NODE_LOCAL;
+
+	status = _cluster_request(CLVMD_CMD_SET_DEBUG, nodes, args, 1, &response, &num_responses, 0);
+
+	/* If any nodes were down then display them and return an error */
+	for (i = 0; i < num_responses; i++) {
+		if (response[i].status == EHOSTDOWN) {
+			fprintf(stderr, "clvmd not running on node %s",
+				  response[i].node);
+			status = 0;
+			errno = response[i].status;
+		} else if (response[i].status) {
+			fprintf(stderr, "Error setting debug on node %s: %s",
+				  response[i].node,
+				  response[i].response[0] ?
+				  	response[i].response :
+				  	strerror(response[i].status));
+			status = 0;
+			errno = response[i].status;
+		}
+	}
+
+	saved_errno = errno;
+	_cluster_free_request(response, num_responses);
+	errno = saved_errno;
+
+	return status;
+}

daemons/clvmd/refresh_clvmd.h

@@ -0,0 +1,19 @@
+/*
+ * Copyright (C) 2007 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU General Public License v.2.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+
+int refresh_clvmd(int all_nodes);
+int restart_clvmd(int all_nodes);
+int debug_clvmd(int level, int clusterwide);
+

daemons/cmirrord/Makefile.in

@@ -0,0 +1,38 @@
+#
+# Copyright (C) 2009-2010 Red Hat, Inc. All rights reserved.
+#
+# This file is part of LVM2.
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions
+# of the GNU General Public License v.2.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+srcdir = @srcdir@
+top_srcdir = @top_srcdir@
+top_builddir = @top_builddir@
+
+CPG_LIBS = @CPG_LIBS@
+CPG_CFLAGS = @CPG_CFLAGS@
+SACKPT_LIBS = @SACKPT_LIBS@
+SACKPT_CFLAGS = @SACKPT_CFLAGS@
+
+SOURCES = clogd.c cluster.c compat.c functions.c link_mon.c local.c logging.c
+
+TARGETS = cmirrord
+
+include $(top_builddir)/make.tmpl
+
+LIBS += -ldevmapper
+LMLIBS += $(CPG_LIBS) $(SACKPT_LIBS)
+CFLAGS += $(CPG_CFLAGS) $(SACKPT_CFLAGS)
+
+cmirrord: $(OBJECTS) $(top_builddir)/lib/liblvm-internal.a
+	$(CC) $(CFLAGS) $(LDFLAGS) -o $@ $(OBJECTS) \
+		$(LVMLIBS) $(LMLIBS) $(LIBS)
+
+install: $(TARGETS)
+	$(INSTALL_PROGRAM) -D cmirrord $(usrsbindir)/cmirrord

daemons/cmirrord/clogd.c

@@ -0,0 +1,250 @@
+/*
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU General Public License v.2.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+#include "logging.h"
+#include "common.h"
+#include "functions.h"
+#include "link_mon.h"
+#include "local.h"
+
+#include <errno.h>
+#include <fcntl.h>
+#include <sys/socket.h>
+#include <sys/stat.h>
+#include <sys/wait.h>
+#include <unistd.h>
+
+static volatile sig_atomic_t exit_now = 0;
+/* FIXME Review signal handling.  Should be volatile sig_atomic_t */
+static sigset_t signal_mask;
+static volatile sig_atomic_t signal_received;
+
+static void process_signals(void);
+static void daemonize(void);
+static void init_all(void);
+static void cleanup_all(void);
+
+int main(int argc __attribute__((unused)), char *argv[] __attribute__((unused)))
+{
+	daemonize();
+
+	init_all();
+
+	/* Parent can now exit, we're ready to handle requests */
+	kill(getppid(), SIGTERM);
+
+	LOG_PRINT("Starting cmirrord:");
+	LOG_PRINT(" Built: "__DATE__" "__TIME__"\n");
+	LOG_DBG(" Compiled with debugging.");
+
+	while (!exit_now) {
+		links_monitor();
+
+		links_issue_callbacks();
+
+		process_signals();
+	}
+	exit(EXIT_SUCCESS);
+}
+
+/*
+ * parent_exit_handler: exit the parent
+ * @sig: the signal
+ *
+ */
+static void parent_exit_handler(int sig __attribute__((unused)))
+{
+	exit_now = 1;
+}
+
+static void sig_handler(int sig)
+{
+	/* FIXME Races - don't touch signal_mask here. */
+	sigaddset(&signal_mask, sig);
+	signal_received = 1;
+}
+
+static void process_signal(int sig){
+	int r = 0;
+
+	switch(sig) {
+	case SIGINT:
+	case SIGQUIT:
+	case SIGTERM:
+	case SIGHUP:
+		r += log_status();
+		break;
+	case SIGUSR1:
+	case SIGUSR2:
+		log_debug();
+		/*local_debug();*/
+		cluster_debug();
+		return;
+	default:
+		LOG_PRINT("Unknown signal received... ignoring");
+		return;
+	}
+
+	if (!r) {
+		LOG_DBG("No current cluster logs... safe to exit.");
+		cleanup_all();
+		exit(EXIT_SUCCESS);
+	}
+
+	LOG_ERROR("Cluster logs exist.  Refusing to exit.");
+}
+
+static void process_signals(void)
+{
+	int x;
+
+	if (!signal_received)
+		return;
+
+	signal_received = 0;
+
+	for (x = 1; x < _NSIG; x++) {
+		if (sigismember(&signal_mask, x)) {
+			sigdelset(&signal_mask, x);
+			process_signal(x);
+		}
+	}
+}
+
+static void remove_lockfile(void)
+{
+	if (unlink(CMIRRORD_PIDFILE))
+		LOG_ERROR("Unable to remove \"" CMIRRORD_PIDFILE "\" %s", strerror(errno));
+}
+
+/*
+ * daemonize
+ *
+ * Performs the steps necessary to become a daemon.
+ */
+static void daemonize(void)
+{
+	int pid;
+	int status;
+	int devnull;
+
+	if ((devnull = open("/dev/null", O_RDWR)) == -1) {
+		LOG_ERROR("Can't open /dev/null: %s", strerror(errno));
+		exit(EXIT_FAILURE);
+	}
+
+	signal(SIGTERM, &parent_exit_handler);
+
+	pid = fork();
+
+	if (pid < 0) {
+		LOG_ERROR("Unable to fork()");
+		exit(EXIT_FAILURE);
+	}
+
+	if (pid) {
+		/* Parent waits here for child to get going */
+		while (!waitpid(pid, &status, WNOHANG) && !exit_now);
+		if (exit_now)
+			exit(EXIT_SUCCESS);
+
+		switch (WEXITSTATUS(status)) {
+		case EXIT_LOCKFILE:
+			LOG_ERROR("Failed to create lockfile");
+			LOG_ERROR("Process already running?");
+			break;
+		case EXIT_KERNEL_SOCKET:
+			LOG_ERROR("Unable to create netlink socket");
+			break;
+		case EXIT_KERNEL_BIND:
+			LOG_ERROR("Unable to bind to netlink socket");
+			break;
+		case EXIT_KERNEL_SETSOCKOPT:
+			LOG_ERROR("Unable to setsockopt on netlink socket");
+			break;
+		case EXIT_CLUSTER_CKPT_INIT:
+			LOG_ERROR("Unable to initialize checkpoint service");
+			LOG_ERROR("Has the cluster infrastructure been started?");
+			break;
+		case EXIT_FAILURE:
+			LOG_ERROR("Failed to start: Generic error");
+			break;
+		default:
+			LOG_ERROR("Failed to start: Unknown error");
+			break;
+		}
+		exit(EXIT_FAILURE);
+	}
+
+	setsid();
+	if (chdir("/")) {
+		LOG_ERROR("Failed to chdir /: %s", strerror(errno));
+		exit(EXIT_FAILURE);
+	}
+
+	umask(0);
+
+	if (close(0) || close(1) || close(2)) {
+		LOG_ERROR("Failed to close terminal FDs");
+		exit(EXIT_FAILURE);
+	}
+
+	if ((dup2(devnull, 0) < 0) || /* reopen stdin */
+	    (dup2(devnull, 1) < 0) || /* reopen stdout */
+	    (dup2(devnull, 2) < 0))   /* reopen stderr */
+		exit(EXIT_FAILURE);
+
+	LOG_OPEN("cmirrord", LOG_PID, LOG_DAEMON);
+
+	(void) dm_prepare_selinux_context(CMIRRORD_PIDFILE, S_IFREG);
+	if (dm_create_lockfile(CMIRRORD_PIDFILE) == 0)
+		exit(EXIT_LOCKFILE);
+	(void) dm_prepare_selinux_context(NULL, 0);
+
+	atexit(remove_lockfile);
+
+	/* FIXME Replace with sigaction. (deprecated) */
+	signal(SIGINT, &sig_handler);
+	signal(SIGQUIT, &sig_handler);
+	signal(SIGTERM, &sig_handler);
+	signal(SIGHUP, &sig_handler);
+	signal(SIGPIPE, SIG_IGN);
+	signal(SIGUSR1, &sig_handler);
+	signal(SIGUSR2, &sig_handler);
+	sigemptyset(&signal_mask);
+	signal_received = 0;
+}
+
+/*
+ * init_all
+ *
+ * Initialize modules.  Exit on failure.
+ */
+static void init_all(void)
+{
+	int r;
+
+	if ((r = init_local()) ||
+	    (r = init_cluster())) {
+		exit(r);
+	}
+}
+
+/*
+ * cleanup_all
+ *
+ * Clean up before exiting
+ */
+static void cleanup_all(void)
+{
+	cleanup_local();
+	cleanup_cluster();
+}

daemons/cmirrord/cluster.h

@@ -0,0 +1,76 @@
+/*
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+#ifndef _LVM_CLOG_CLUSTER_H
+#define _LVM_CLOG_CLUSTER_H
+
+#include "dm-log-userspace.h"
+#include "libdevmapper.h"
+
+#define DM_ULOG_RESPONSE 0x1000U /* in last byte of 32-bit value */
+#define DM_ULOG_CHECKPOINT_READY 21
+#define DM_ULOG_MEMBER_JOIN      22
+
+/*
+ * There is other information in addition to what can
+ * be found in the dm_ulog_request structure that we
+ * need for processing.  'clog_request' is the wrapping
+ * structure we use to make the additional fields
+ * available.
+ */
+struct clog_request {
+	/*
+	 * If we don't use a union, the structure size will
+	 * vary between 32-bit and 64-bit machines.  So, we
+	 * pack two 64-bit version numbers in there to force
+	 * the size of the structure to be the same.
+	 *
+	 * The two version numbers also help us with endian
+	 * issues.  The first is always little endian, while
+	 * the second is in native format of the sending
+	 * machine.  If the two are equal, there is no need
+	 * to do endian conversions.
+	 */
+	union {
+		uint64_t version[2]; /* LE version and native version */
+		struct dm_list list;
+	} u;
+
+	/*
+	 * 'originator' is the machine from which the requests
+	 * was made.
+	 */
+	uint32_t originator;
+
+	/*
+	 * 'pit_server' is the "point-in-time" server for the
+	 * request.  (I.e.  The machine that was the server at
+	 * the time the request was issued - only important during
+	 * startup.
+	 */
+	uint32_t pit_server;
+
+	/*
+	 * The request from the kernel that is being processed
+	 */
+	struct dm_ulog_request u_rq;
+};
+
+int init_cluster(void);
+void cleanup_cluster(void);
+void cluster_debug(void);
+
+int create_cluster_cpg(char *uuid, uint64_t luid);
+int destroy_cluster_cpg(char *uuid);
+
+int cluster_send(struct clog_request *rq);
+
+#endif /* _LVM_CLOG_CLUSTER_H */

daemons/cmirrord/common.h

@@ -0,0 +1,33 @@
+/*
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+#ifndef _LVM_CLOG_COMMON_H
+#define _LVM_CLOG_COMMON_H
+
+/*
+ * If there are problems when forking off to become a daemon,
+ * the child will exist with one of these codes.  This allows
+ * the parent to know the reason for the failure and print it
+ * to the launching terminal.
+ *
+ * #define EXIT_SUCCESS 0 (from stdlib.h)
+ * #define EXIT_FAILURE 1 (from stdlib.h)
+ */
+#define EXIT_LOCKFILE              2
+#define EXIT_KERNEL_SOCKET         3 /* Failed netlink socket create */
+#define EXIT_KERNEL_BIND           4
+#define EXIT_KERNEL_SETSOCKOPT     5
+#define EXIT_CLUSTER_CKPT_INIT     6 /* Failed to init checkpoint */
+#define EXIT_QUEUE_NOMEM           7
+
+#define DM_ULOG_REQUEST_SIZE 1024
+
+#endif /* _LVM_CLOG_COMMON_H */

daemons/cmirrord/compat.c

@@ -0,0 +1,210 @@
+/*
+ * Copyright (C) 2010 Red Hat, Inc. All rights reserved.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ */
+#include "logging.h"
+#include "cluster.h"
+#include "compat.h"
+#include "xlate.h"
+
+#include <errno.h>
+
+/*
+ * Older versions of the log daemon communicate with different
+ * versions of the inter-machine communication structure, which
+ * varies in size and fields.  The older versions append the
+ * standard upstream version of the structure to every request.
+ * COMPAT_OFFSET is where the upstream structure starts.
+ */
+#define COMPAT_OFFSET 256
+
+static void v5_data_endian_switch(struct clog_request *rq, int to_network __attribute__((unused)))
+{
+	int i, end;
+	int64_t *pi64;
+	uint64_t *pu64;
+	uint32_t rq_type = rq->u_rq.request_type & ~DM_ULOG_RESPONSE;
+
+	if (rq->u_rq.request_type & DM_ULOG_RESPONSE) {
+		switch (rq_type) {
+		case DM_ULOG_CTR:
+		case DM_ULOG_DTR:
+			LOG_ERROR("Invalid response type in endian switch");
+			exit(EXIT_FAILURE);
+
+		case DM_ULOG_PRESUSPEND:
+		case DM_ULOG_POSTSUSPEND:
+		case DM_ULOG_RESUME:
+		case DM_ULOG_FLUSH:
+		case DM_ULOG_MARK_REGION:
+		case DM_ULOG_CLEAR_REGION:
+		case DM_ULOG_SET_REGION_SYNC:
+		case DM_ULOG_CHECKPOINT_READY:
+		case DM_ULOG_MEMBER_JOIN:
+		case DM_ULOG_STATUS_INFO:
+		case DM_ULOG_STATUS_TABLE:
+			/* No outbound data */
+			break;
+
+		case DM_ULOG_GET_REGION_SIZE:
+		case DM_ULOG_GET_SYNC_COUNT:
+			pu64 = (uint64_t *)rq->u_rq.data;
+			*pu64 = xlate64(*pu64);
+			break;
+		case DM_ULOG_IS_CLEAN:
+		case DM_ULOG_IN_SYNC:
+			pi64 = (int64_t *)rq->u_rq.data;
+			*pi64 = xlate64(*pi64);
+			break;
+		case DM_ULOG_GET_RESYNC_WORK:
+		case DM_ULOG_IS_REMOTE_RECOVERING:
+			pi64 = (int64_t *)rq->u_rq.data;
+			pu64 = ((uint64_t *)rq->u_rq.data) + 1;
+			*pi64 = xlate64(*pi64);
+			*pu64 = xlate64(*pu64);
+			break;
+		default:
+			LOG_ERROR("Unknown request type, %u", rq_type);
+			return;
+		}
+	} else {
+		switch (rq_type) {
+		case DM_ULOG_CTR:
+		case DM_ULOG_DTR:
+			LOG_ERROR("Invalid request type in endian switch");
+			exit(EXIT_FAILURE);
+
+		case DM_ULOG_PRESUSPEND:
+		case DM_ULOG_POSTSUSPEND:
+		case DM_ULOG_RESUME:
+		case DM_ULOG_GET_REGION_SIZE:
+		case DM_ULOG_FLUSH:
+		case DM_ULOG_GET_RESYNC_WORK:
+		case DM_ULOG_GET_SYNC_COUNT:
+		case DM_ULOG_STATUS_INFO:
+		case DM_ULOG_STATUS_TABLE:
+		case DM_ULOG_CHECKPOINT_READY:
+		case DM_ULOG_MEMBER_JOIN:
+			/* No incoming data */
+			break;
+		case DM_ULOG_IS_CLEAN:
+		case DM_ULOG_IN_SYNC:
+		case DM_ULOG_IS_REMOTE_RECOVERING:
+			pu64 = (uint64_t *)rq->u_rq.data;
+			*pu64 = xlate64(*pu64);
+			break;
+		case DM_ULOG_MARK_REGION:
+		case DM_ULOG_CLEAR_REGION:
+			end = rq->u_rq.data_size/sizeof(uint64_t);
+
+			pu64 = (uint64_t *)rq->u_rq.data;
+			for (i = 0; i < end; i++)
+				pu64[i] = xlate64(pu64[i]);
+			break;
+		case DM_ULOG_SET_REGION_SYNC:
+			pu64 = (uint64_t *)rq->u_rq.data;
+			pi64 = ((int64_t *)rq->u_rq.data) + 1;
+			*pu64 = xlate64(*pu64);
+			*pi64 = xlate64(*pi64);
+			break;
+		default:
+			LOG_ERROR("Unknown request type, %u", rq_type);
+			exit(EXIT_FAILURE);
+		}
+	}
+}
+
+static int v5_endian_to_network(struct clog_request *rq)
+{
+	int size;
+	struct dm_ulog_request *u_rq = &rq->u_rq;
+
+	size = sizeof(*rq) + u_rq->data_size;
+
+	u_rq->error = xlate32(u_rq->error);
+	u_rq->seq = xlate32(u_rq->seq);
+	u_rq->request_type = xlate32(u_rq->request_type);
+	u_rq->data_size = xlate64(u_rq->data_size);
+
+	rq->originator = xlate32(rq->originator);
+
+	v5_data_endian_switch(rq, 1);
+
+	return size;
+}
+
+int clog_request_to_network(struct clog_request *rq)
+{
+	int r;
+
+	/* FIXME: Remove this safety check */
+	if (rq->u.version[0] != xlate64(rq->u.version[1])) {
+		LOG_ERROR("Programmer error:  version[0] must be LE");
+		exit(EXIT_FAILURE);
+	}
+
+	/*
+	 * Are we already running in the endian mode we send
+	 * over the wire?
+	 */
+	if (rq->u.version[0] == rq->u.version[1])
+		return 0;
+
+	r = v5_endian_to_network(rq);
+	if (r < 0)
+		return r;
+	return 0;
+}
+
+static int v5_endian_from_network(struct clog_request *rq)
+{
+	int size;
+	struct dm_ulog_request *u_rq = &rq->u_rq;
+
+	u_rq->error = xlate32(u_rq->error);
+	u_rq->seq = xlate32(u_rq->seq);
+	u_rq->request_type = xlate32(u_rq->request_type);
+	u_rq->data_size = xlate64(u_rq->data_size);
+
+	rq->originator = xlate32(rq->originator);
+
+	size = sizeof(*rq) + u_rq->data_size;
+
+	v5_data_endian_switch(rq, 0);
+
+	return size;
+}
+
+int clog_request_from_network(void *data, size_t data_len)
+{
+	uint64_t *vp = data;
+	uint64_t version = xlate64(vp[0]);
+	uint64_t unconverted_version = vp[1];
+	struct clog_request *rq = data;
+
+	switch (version) {
+	case 5: /* Upstream */
+		if (version == unconverted_version)
+			return 0;
+		break;
+	case 4: /* RHEL 5.[45] */
+	case 3: /* RHEL 5.3 */
+	case 2: /* RHEL 5.2 */
+		/* FIXME: still need to account for payload */
+		if (data_len < (COMPAT_OFFSET + sizeof(*rq)))
+			return -ENOSPC;
+
+		rq = (struct clog_request *)((char *)data + COMPAT_OFFSET);
+		break;
+	default:
+		LOG_ERROR("Unable to process cluster message: "
+			  "Incompatible version");
+		return -EINVAL;
+	}
+
+	v5_endian_from_network(rq);
+	return 0;
+}

daemons/cmirrord/compat.h

@@ -0,0 +1,25 @@
+/*
+ * Copyright (C) 2010 Red Hat, Inc. All rights reserved.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ */
+#ifndef _LVM_CLOG_COMPAT_H
+#define _LVM_CLOG_COMPAT_H
+
+/*
+ * The intermachine communication structure version are:
+ *	0: Unused
+ *	1: Never in the wild
+ *	2: RHEL 5.2
+ *	3: RHEL 5.3
+ *	4: RHEL 5.4, RHEL 5.5
+ *	5: RHEL 6, Current Upstream Format
+ */
+#define CLOG_TFR_VERSION 5
+
+int clog_request_to_network(struct clog_request *rq);
+int clog_request_from_network(void *data, size_t data_len);
+
+#endif /* _LVM_CLOG_COMPAT_H */

daemons/cmirrord/functions.h

@@ -0,0 +1,34 @@
+/*
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+#ifndef _LVM_CLOG_FUNCTIONS_H
+#define _LVM_CLOG_FUNCTIONS_H
+
+#include "dm-log-userspace.h"
+#include "cluster.h"
+
+#define LOG_RESUMED   1
+#define LOG_SUSPENDED 2
+
+int local_resume(struct dm_ulog_request *rq);
+int cluster_postsuspend(char *, uint64_t);
+
+int do_request(struct clog_request *rq, int server);
+int push_state(const char *uuid, uint64_t luid,
+	       const char *which, char **buf, uint32_t debug_who);
+int pull_state(const char *uuid, uint64_t luid,
+	       const char *which, char *buf, int size);
+
+int log_get_state(struct dm_ulog_request *rq);
+int log_status(void);
+void log_debug(void);
+
+#endif /* _LVM_CLOG_FUNCTIONS_H */

daemons/cmirrord/link_mon.c

@@ -0,0 +1,151 @@
+/*
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+#include "logging.h"
+#include "link_mon.h"
+
+#include <errno.h>
+#include <poll.h>
+#include <stdlib.h>
+
+struct link_callback {
+	int fd;
+	const char *name;
+	void *data;
+	int (*callback)(void *data);
+
+	struct link_callback *next;
+};
+
+static unsigned used_pfds = 0;
+static unsigned free_pfds = 0;
+static struct pollfd *pfds = NULL;
+static struct link_callback *callbacks = NULL;
+
+int links_register(int fd, const char *name, int (*callback)(void *data), void *data)
+{
+	unsigned i;
+	struct link_callback *lc;
+
+	for (i = 0; i < used_pfds; i++) {
+		if (fd == pfds[i].fd) {
+			LOG_ERROR("links_register: Duplicate file descriptor");
+			return -EINVAL;
+		}
+	}
+
+	lc = malloc(sizeof(*lc));
+	if (!lc)
+		return -ENOMEM;
+
+	lc->fd = fd;
+	lc->name = name;
+	lc->data = data;
+	lc->callback = callback;
+
+	if (!free_pfds) {
+		struct pollfd *tmp;
+		tmp = realloc(pfds, sizeof(struct pollfd) * ((used_pfds*2) + 1));
+		if (!tmp) {
+			free(lc);
+			return -ENOMEM;
+		}
+
+		pfds = tmp;
+		free_pfds = used_pfds + 1;
+	}
+
+	free_pfds--;
+	pfds[used_pfds].fd = fd;
+	pfds[used_pfds].events = POLLIN;
+	pfds[used_pfds].revents = 0;
+	used_pfds++;
+
+	lc->next = callbacks;
+	callbacks = lc;
+	LOG_DBG("Adding %s/%d", lc->name, lc->fd);
+	LOG_DBG(" used_pfds = %u, free_pfds = %u",
+		used_pfds, free_pfds);
+
+	return 0;
+}
+
+int links_unregister(int fd)
+{
+	unsigned i;
+	struct link_callback *p, *c;
+
+	for (i = 0; i < used_pfds; i++)
+		if (fd == pfds[i].fd) {
+			/* entire struct is copied (overwritten) */
+			pfds[i] = pfds[used_pfds - 1];
+			used_pfds--;
+			free_pfds++;
+		}
+
+	for (p = NULL, c = callbacks; c; p = c, c = c->next)
+		if (fd == c->fd) {
+			LOG_DBG("Freeing up %s/%d", c->name, c->fd);
+			LOG_DBG(" used_pfds = %u, free_pfds = %u",
+				used_pfds, free_pfds);
+			if (p)
+				p->next = c->next;
+			else
+				callbacks = c->next;
+			free(c);
+			break;
+		}
+
+	return 0;
+}
+
+int links_monitor(void)
+{
+	unsigned i;
+	int r;
+
+	for (i = 0; i < used_pfds; i++) {
+		pfds[i].revents = 0;
+	}
+
+	r = poll(pfds, used_pfds, -1);
+	if (r <= 0)
+		return r;
+
+	r = 0;
+	/* FIXME: handle POLLHUP */
+	for (i = 0; i < used_pfds; i++)
+		if (pfds[i].revents & POLLIN) {
+			LOG_DBG("Data ready on %d", pfds[i].fd);
+
+			/* FIXME: Add this back return 1;*/
+			r++;
+		}
+
+	return r;
+}
+
+int links_issue_callbacks(void)
+{
+	unsigned i;
+	struct link_callback *lc;
+
+	for (i = 0; i < used_pfds; i++)
+		if (pfds[i].revents & POLLIN)
+			for (lc = callbacks; lc; lc = lc->next)
+				if (pfds[i].fd == lc->fd) {
+					LOG_DBG("Issuing callback on %s/%d",
+						lc->name, lc->fd);
+					lc->callback(lc->data);
+					break;
+				}
+	return 0;
+}

daemons/cmirrord/link_mon.h

@@ -0,0 +1,20 @@
+/*
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+#ifndef _LVM_CLOG_LINK_MON_H
+#define _LVM_CLOG_LINK_MON_H
+
+int links_register(int fd, const char *name, int (*callback)(void *data), void *data);
+int links_unregister(int fd);
+int links_monitor(void);
+int links_issue_callbacks(void);
+
+#endif /* _LVM_CLOG_LINK_MON_H */

daemons/cmirrord/local.c

@@ -0,0 +1,424 @@
+/*
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+#include "logging.h"
+#include "common.h"
+#include "functions.h"
+#include "link_mon.h"
+#include "local.h"
+
+#include <errno.h>
+#include <sys/socket.h>
+#include <linux/connector.h>
+#include <linux/netlink.h>
+#include <unistd.h>
+
+#ifndef CN_IDX_DM
+/* Kernel 2.6.31 is required to run this code */
+#define CN_IDX_DM                       0x7     /* Device Mapper */
+#define CN_VAL_DM_USERSPACE_LOG         0x1
+#endif
+
+static int cn_fd = -1;  /* Connector (netlink) socket fd */
+static char recv_buf[2048];
+static char send_buf[2048];
+
+
+/* FIXME: merge this function with kernel_send_helper */
+static int kernel_ack(uint32_t seq, int error)
+{
+	int r;
+	struct nlmsghdr *nlh = (struct nlmsghdr *)send_buf;
+	struct cn_msg *msg = NLMSG_DATA(nlh);
+
+	if (error < 0) {
+		LOG_ERROR("Programmer error: error codes must be positive");
+		return -EINVAL;
+	}
+
+	memset(send_buf, 0, sizeof(send_buf));
+
+	nlh->nlmsg_seq = 0;
+	nlh->nlmsg_pid = getpid();
+	nlh->nlmsg_type = NLMSG_DONE;
+	nlh->nlmsg_len = NLMSG_LENGTH(sizeof(struct cn_msg));
+	nlh->nlmsg_flags = 0;
+
+	msg->len = 0;
+	msg->id.idx = CN_IDX_DM;
+	msg->id.val = CN_VAL_DM_USERSPACE_LOG;
+	msg->seq = seq;
+	msg->ack = error;
+
+	r = send(cn_fd, nlh, NLMSG_LENGTH(sizeof(struct cn_msg)), 0);
+	/* FIXME: do better error processing */
+	if (r <= 0)
+		return -EBADE;
+
+	return 0;
+}
+
+
+/*
+ * kernel_recv
+ * @rq: the newly allocated request from kernel
+ *
+ * Read requests from the kernel and allocate space for the new request.
+ * If there is no request from the kernel, *rq is NULL.
+ *
+ * This function is not thread safe due to returned stack pointer.  In fact,
+ * the returned pointer must not be in-use when this function is called again.
+ *
+ * Returns: 0 on success, -EXXX on error
+ */
+static int kernel_recv(struct clog_request **rq)
+{
+	int r = 0;
+	ssize_t len;
+	char *foo;
+	struct cn_msg *msg;
+	struct dm_ulog_request *u_rq;
+	struct nlmsghdr *nlmsg_h;
+
+	*rq = NULL;
+	memset(recv_buf, 0, sizeof(recv_buf));
+
+	len = recv(cn_fd, recv_buf, sizeof(recv_buf), 0);
+	if (len < 0) {
+		LOG_ERROR("Failed to recv message from kernel");
+		r = -errno;
+		goto fail;
+	}
+
+	nlmsg_h = (struct nlmsghdr *)recv_buf;
+	switch (nlmsg_h->nlmsg_type) {
+	case NLMSG_ERROR:
+		LOG_ERROR("Unable to recv message from kernel: NLMSG_ERROR");
+		r = -EBADE;
+		goto fail;
+	case NLMSG_DONE:
+		msg = (struct cn_msg *)NLMSG_DATA((struct nlmsghdr *)recv_buf);
+		len -= (ssize_t)sizeof(struct nlmsghdr);
+
+		if (len < (ssize_t)sizeof(struct cn_msg)) {
+			LOG_ERROR("Incomplete request from kernel received");
+			r = -EBADE;
+			goto fail;
+		}
+
+		if (msg->len > DM_ULOG_REQUEST_SIZE) {
+			LOG_ERROR("Not enough space to receive kernel request (%d/%d)",
+				  msg->len, DM_ULOG_REQUEST_SIZE);
+			r = -EBADE;
+			goto fail;
+		}
+
+		if (!msg->len)
+			LOG_ERROR("Zero length message received");
+
+		len -= (ssize_t)sizeof(struct cn_msg);
+
+		if (len < msg->len)
+			LOG_ERROR("len = %zd, msg->len = %" PRIu16, len, msg->len);
+
+		msg->data[msg->len] = '\0'; /* Cleaner way to ensure this? */
+		u_rq = (struct dm_ulog_request *)msg->data;
+
+		if (!u_rq->request_type) {
+			LOG_DBG("Bad transmission, requesting resend [%u]",
+				msg->seq);
+			r = -EAGAIN;
+
+			if (kernel_ack(msg->seq, EAGAIN)) {
+				LOG_ERROR("Failed to NACK kernel transmission [%u]",
+					  msg->seq);
+				r = -EBADE;
+			}
+		}
+
+		/*
+		 * Now we've got sizeof(struct cn_msg) + sizeof(struct nlmsghdr)
+		 * worth of space that precede the request structure from the
+		 * kernel.  Since that space isn't going to be used again, we
+		 * can take it for our purposes; rather than allocating a whole
+		 * new structure and doing a memcpy.
+		 *
+		 * We should really make sure 'clog_request' doesn't grow
+		 * beyond what is available to us, but we need only check it
+		 * once... perhaps at compile time?
+		 */
+		foo = (char *)u_rq;
+		foo -= (sizeof(struct clog_request) - sizeof(struct dm_ulog_request));
+		*rq = (struct clog_request *) foo;
+
+		/* Clear the wrapper container fields */
+		memset(*rq, 0, (size_t)((char *)u_rq - (char *)(*rq)));
+		break;
+	default:
+		LOG_ERROR("Unknown nlmsg_type");
+		r = -EBADE;
+	}
+
+fail:
+	if (r)
+		*rq = NULL;
+
+	return (r == -EAGAIN) ? 0 : r;
+}
+
+static int kernel_send_helper(void *data, uint16_t out_size)
+{
+	int r;
+	struct nlmsghdr *nlh;
+	struct cn_msg *msg;
+
+	memset(send_buf, 0, sizeof(send_buf));
+
+	nlh = (struct nlmsghdr *)send_buf;
+	nlh->nlmsg_seq = 0;  /* FIXME: Is this used? */
+	nlh->nlmsg_pid = getpid();
+	nlh->nlmsg_type = NLMSG_DONE;
+	nlh->nlmsg_len = NLMSG_LENGTH(out_size + sizeof(struct cn_msg));
+	nlh->nlmsg_flags = 0;
+
+	msg = NLMSG_DATA(nlh);
+	memcpy(msg->data, data, out_size);
+	msg->len = out_size;
+	msg->id.idx = CN_IDX_DM;
+	msg->id.val = CN_VAL_DM_USERSPACE_LOG;
+	msg->seq = 0;
+
+	r = send(cn_fd, nlh, NLMSG_LENGTH(out_size + sizeof(struct cn_msg)), 0);
+	/* FIXME: do better error processing */
+	if (r <= 0)
+		return -EBADE;
+
+	return 0;
+}
+
+/*
+ * do_local_work
+ *
+ * Any processing errors are placed in the 'rq'
+ * structure to be reported back to the kernel.
+ * It may be pointless for this function to
+ * return an int.
+ *
+ * Returns: 0 on success, -EXXX on failure
+ */
+static int do_local_work(void *data __attribute__((unused)))
+{
+	int r;
+	struct clog_request *rq;
+	struct dm_ulog_request *u_rq = NULL;
+
+	r = kernel_recv(&rq);
+	if (r)
+		return r;
+
+	if (!rq)
+		return 0;
+
+	u_rq = &rq->u_rq;
+	LOG_DBG("[%s]  Request from kernel received: [%s/%u]",
+		SHORT_UUID(u_rq->uuid), RQ_TYPE(u_rq->request_type),
+		u_rq->seq);
+	switch (u_rq->request_type) {
+	case DM_ULOG_CTR:
+	case DM_ULOG_DTR:
+	case DM_ULOG_GET_REGION_SIZE:
+	case DM_ULOG_IN_SYNC:
+	case DM_ULOG_GET_SYNC_COUNT:
+	case DM_ULOG_STATUS_TABLE:
+	case DM_ULOG_PRESUSPEND:
+		/* We do not specify ourselves as server here */
+		r = do_request(rq, 0);
+		if (r)
+			LOG_DBG("Returning failed request to kernel [%s]",
+				RQ_TYPE(u_rq->request_type));
+		r = kernel_send(u_rq);
+		if (r)
+			LOG_ERROR("Failed to respond to kernel [%s]",
+				  RQ_TYPE(u_rq->request_type));
+
+		break;
+	case DM_ULOG_RESUME:
+		/*
+		 * Resume is a special case that requires a local
+		 * component to join the CPG, and a cluster component
+		 * to handle the request.
+		 */
+		r = local_resume(u_rq);
+		if (r) {
+			LOG_DBG("Returning failed request to kernel [%s]",
+				RQ_TYPE(u_rq->request_type));
+			r = kernel_send(u_rq);
+			if (r)
+				LOG_ERROR("Failed to respond to kernel [%s]",
+					  RQ_TYPE(u_rq->request_type));
+			break;
+		}
+		/* ELSE, fall through */
+	case DM_ULOG_IS_CLEAN:
+	case DM_ULOG_FLUSH:
+	case DM_ULOG_MARK_REGION:
+	case DM_ULOG_GET_RESYNC_WORK:
+	case DM_ULOG_SET_REGION_SYNC:
+	case DM_ULOG_STATUS_INFO:
+	case DM_ULOG_IS_REMOTE_RECOVERING:
+	case DM_ULOG_POSTSUSPEND:
+		r = cluster_send(rq);
+		if (r) {
+			u_rq->data_size = 0;
+			u_rq->error = r;
+			if (kernel_send(u_rq))
+				LOG_ERROR("Failed to respond to kernel [%s]",
+					  RQ_TYPE(u_rq->request_type));
+		}
+
+		break;
+	case DM_ULOG_CLEAR_REGION:
+		r = kernel_ack(u_rq->seq, 0);
+
+		r = cluster_send(rq);
+		if (r) {
+			/*
+			 * FIXME: store error for delivery on flush
+			 *        This would allow us to optimize MARK_REGION
+			 *        too.
+			 */
+		}
+
+		break;
+	default:
+		LOG_ERROR("Invalid log request received (%u), ignoring.",
+			  u_rq->request_type);
+
+		return 0;
+	}
+
+	if (r && !u_rq->error)
+		u_rq->error = r;
+
+	return r;
+}
+
+/*
+ * kernel_send
+ * @u_rq: result to pass back to kernel
+ *
+ * This function returns the u_rq structure
+ * (containing the results) to the kernel.
+ * It then frees the structure.
+ *
+ * WARNING: should the structure be freed if
+ * there is an error?  I vote 'yes'.  If the
+ * kernel doesn't get the response, it should
+ * resend the request.
+ *
+ * Returns: 0 on success, -EXXX on failure
+ */
+int kernel_send(struct dm_ulog_request *u_rq)
+{
+	int r;
+	uint16_t size;
+
+	if (!u_rq)
+		return -EINVAL;
+
+	size = (uint16_t)(sizeof(struct dm_ulog_request) + u_rq->data_size);
+
+	if (!u_rq->data_size && !u_rq->error) {
+		/* An ACK is all that is needed */
+
+		/* FIXME: add ACK code */
+	} else if (size > DM_ULOG_REQUEST_SIZE) {
+		/*
+		 * If we gotten here, we've already overrun
+		 * our allotted space somewhere.
+		 *
+		 * We must do something, because the kernel
+		 * is waiting for a response.
+		 */
+		LOG_ERROR("Not enough space to respond to server");
+		u_rq->error = -ENOSPC;
+		size = sizeof(struct dm_ulog_request);
+	}
+
+	r = kernel_send_helper(u_rq, size);
+	if (r)
+		LOG_ERROR("Failed to send msg to kernel.");
+
+	return r;
+}
+
+/*
+ * init_local
+ *
+ * Initialize kernel communication socket (netlink)
+ *
+ * Returns: 0 on success, values from common.h on failure
+ */
+int init_local(void)
+{
+	int r = 0;
+	unsigned opt;
+	struct sockaddr_nl addr;
+
+	cn_fd = socket(PF_NETLINK, SOCK_DGRAM, NETLINK_CONNECTOR);
+	if (cn_fd < 0)
+		return EXIT_KERNEL_SOCKET;
+
+	/* memset to fix valgrind complaint */
+	memset(&addr, 0, sizeof(struct sockaddr_nl));
+
+	addr.nl_family = AF_NETLINK;
+	addr.nl_groups = CN_IDX_DM;
+	addr.nl_pid = 0;
+
+	r = bind(cn_fd, (struct sockaddr *) &addr, sizeof(addr));
+	if (r < 0) {
+		if (close(cn_fd))
+			LOG_ERROR("Failed to close socket: %s",
+				  strerror(errno));
+		return EXIT_KERNEL_BIND;
+	}
+
+	opt = addr.nl_groups;
+	r = setsockopt(cn_fd, 270, NETLINK_ADD_MEMBERSHIP, &opt, sizeof(opt));
+	if (r) {
+		if (close(cn_fd))
+			LOG_ERROR("Failed to close socket: %s",
+				  strerror(errno));
+		return EXIT_KERNEL_SETSOCKOPT;
+	}
+
+	/*
+	r = fcntl(cn_fd, F_SETFL, FNDELAY);
+	*/
+
+	links_register(cn_fd, "local", do_local_work, NULL);
+
+	return 0;
+}
+
+/*
+ * cleanup_local
+ *
+ * Clean up before exiting
+ */
+void cleanup_local(void)
+{
+	links_unregister(cn_fd);
+	if (cn_fd >= 0 && close(cn_fd))
+		LOG_ERROR("Failed to close socket: %s",
+			  strerror(errno));
+}

daemons/cmirrord/local.h

@@ -0,0 +1,20 @@
+/*
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+#ifndef _LVM_CLOG_LOCAL_H
+#define _LVM_CLOG_LOCAL_H
+
+int init_local(void);
+void cleanup_local(void);
+
+int kernel_send(struct dm_ulog_request *rq);
+
+#endif /* _LVM_CLOG_LOCAL_H */

daemons/cmirrord/logging.c

@@ -0,0 +1,57 @@
+/*
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+#include "logging.h"
+
+const char *__rq_types_off_by_one[] = {
+	"DM_ULOG_CTR",
+	"DM_ULOG_DTR",
+	"DM_ULOG_PRESUSPEND",
+	"DM_ULOG_POSTSUSPEND",
+	"DM_ULOG_RESUME",
+	"DM_ULOG_GET_REGION_SIZE",
+	"DM_ULOG_IS_CLEAN",
+	"DM_ULOG_IN_SYNC",
+	"DM_ULOG_FLUSH",
+	"DM_ULOG_MARK_REGION",
+	"DM_ULOG_CLEAR_REGION",
+	"DM_ULOG_GET_RESYNC_WORK",
+	"DM_ULOG_SET_REGION_SYNC",
+	"DM_ULOG_GET_SYNC_COUNT",
+	"DM_ULOG_STATUS_INFO",
+	"DM_ULOG_STATUS_TABLE",
+	"DM_ULOG_IS_REMOTE_RECOVERING",
+	NULL
+};
+
+int log_tabbing = 0;
+int log_is_open = 0;
+
+/*
+ * Variables for various conditional logging
+ */
+#ifdef MEMB
+int log_membership_change = 1;
+#else
+int log_membership_change = 0;
+#endif
+
+#ifdef CKPT
+int log_checkpoint = 1;
+#else
+int log_checkpoint = 0;
+#endif
+
+#ifdef RESEND
+int log_resend_requests = 1;
+#else
+int log_resend_requests = 0;
+#endif

daemons/cmirrord/logging.h

@@ -0,0 +1,77 @@
+/*
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#ifndef _LVM_CLOG_LOGGING_H
+#define _LVM_CLOG_LOGGING_H
+
+#define _GNU_SOURCE
+#define _FILE_OFFSET_BITS 64
+
+#include "configure.h"
+#include <stdio.h>
+#include <stdint.h>
+#include <syslog.h>
+
+/* SHORT_UUID - print last 8 chars of a string */
+#define SHORT_UUID(x) (strlen(x) > 8) ? ((x) + (strlen(x) - 8)) : (x)
+
+extern const char *__rq_types_off_by_one[];
+#define RQ_TYPE(x) __rq_types_off_by_one[(x) - 1]
+
+extern int log_tabbing;
+extern int log_is_open;
+extern int log_membership_change;
+extern int log_checkpoint;
+extern int log_resend_requests;
+
+#define LOG_OPEN(ident, option, facility) do { \
+		openlog(ident, option, facility); \
+		log_is_open = 1;		  \
+	} while (0)
+
+#define LOG_CLOSE(void) do { \
+		log_is_open = 0; \
+		closelog();	 \
+	} while (0)
+
+#define LOG_OUTPUT(level, f, arg...) do {				\
+		int __i;						\
+		char __buffer[16];					\
+		FILE *fp = (level > LOG_NOTICE) ? stderr : stdout;	\
+		if (log_is_open) {					\
+			for (__i = 0; (__i < log_tabbing) && (__i < 15); __i++) \
+				__buffer[__i] = '\t';			\
+			__buffer[__i] = '\0';				\
+			syslog(level, "%s" f "\n", __buffer, ## arg);	\
+		} else {						\
+			for (__i = 0; __i < log_tabbing; __i++)		\
+				fprintf(fp, "\t");			\
+			fprintf(fp, f "\n", ## arg);			\
+		}							\
+	} while (0)
+
+
+#ifdef DEBUG
+#define LOG_DBG(f, arg...) LOG_OUTPUT(LOG_DEBUG, f, ## arg)
+#else /* DEBUG */
+#define LOG_DBG(f, arg...)
+#endif /* DEBUG */
+
+#define LOG_COND(__X, f, arg...) do {\
+		if (__X) { 	     \
+			LOG_OUTPUT(LOG_NOTICE, f, ## arg); \
+		} \
+	} while (0)
+#define LOG_PRINT(f, arg...) LOG_OUTPUT(LOG_NOTICE, f, ## arg)
+#define LOG_ERROR(f, arg...) LOG_OUTPUT(LOG_ERR, f, ## arg)
+
+#endif /* _LVM_CLOG_LOGGING_H */

daemons/dmeventd/.exported_symbols

@@ -0,0 +1,4 @@
+init_fifos
+fini_fifos
+daemon_talk
+dm_event_get_version

daemons/dmeventd/Makefile.in

@@ -0,0 +1,108 @@
+#
+# Copyright (C) 2005-2011 Red Hat, Inc. All rights reserved.
+#
+# This file is part of the device-mapper userspace tools.
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions
+# of the GNU Lesser General Public License v.2.1.
+#
+# You should have received a copy of the GNU Lesser General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+srcdir = @srcdir@
+top_srcdir = @top_srcdir@
+top_builddir = @top_builddir@
+
+SOURCES = libdevmapper-event.c
+SOURCES2 = dmeventd.c
+
+TARGETS = dmeventd
+
+.PHONY: install_lib_dynamic install_lib_static install_include \
+	install_pkgconfig install_dmeventd_dynamic install_dmeventd_static \
+	install_lib install_dmeventd
+
+INSTALL_DMEVENTD_TARGETS = install_dmeventd_dynamic
+INSTALL_LIB_TARGETS = install_lib_dynamic
+
+LIB_NAME = libdevmapper-event
+ifeq ("@STATIC_LINK@", "yes")
+  LIB_STATIC = $(LIB_NAME).a
+  TARGETS += $(LIB_STATIC) dmeventd.static
+  INSTALL_DMEVENTD_TARGETS += install_dmeventd_static
+  INSTALL_LIB_TARGETS += install_lib_static
+endif
+
+LIB_VERSION = $(LIB_VERSION_DM)
+LIB_SHARED = $(LIB_NAME).$(LIB_SUFFIX)
+
+CLEAN_TARGETS = dmeventd.static $(LIB_NAME).a
+
+ifneq ($(MAKECMDGOALS),device-mapper)
+  SUBDIRS+=plugins
+endif
+
+CFLOW_LIST = $(SOURCES)
+CFLOW_LIST_TARGET = $(LIB_NAME).cflow
+CFLOW_TARGET = dmeventd
+
+EXPORTED_HEADER = $(srcdir)/libdevmapper-event.h
+EXPORTED_FN_PREFIX = dm_event
+
+include $(top_builddir)/make.tmpl
+
+all: device-mapper
+device-mapper: $(TARGETS)
+
+LIBS += -ldevmapper
+LVMLIBS += -ldevmapper-event $(PTHREAD_LIBS)
+
+dmeventd: $(LIB_SHARED) dmeventd.o
+	$(CC) $(CFLAGS) $(LDFLAGS) $(ELDFLAGS) -L. -o $@ dmeventd.o \
+	$(DL_LIBS) $(LVMLIBS) $(LIBS) -rdynamic
+
+dmeventd.static: $(LIB_STATIC) dmeventd.o $(interfacebuilddir)/libdevmapper.a
+	$(CC) $(CFLAGS) $(LDFLAGS) $(ELDFLAGS) -static -L. -L$(interfacebuilddir) -o $@ \
+	dmeventd.o $(DL_LIBS) $(LVMLIBS) $(LIBS) $(STATIC_LIBS)
+
+ifeq ("@PKGCONFIG@", "yes")
+  INSTALL_LIB_TARGETS += install_pkgconfig
+endif
+
+ifneq ("$(CFLOW_CMD)", "")
+CFLOW_SOURCES = $(addprefix $(srcdir)/, $(SOURCES))
+-include $(top_builddir)/libdm/libdevmapper.cflow
+-include $(top_builddir)/lib/liblvm-internal.cflow
+-include $(top_builddir)/lib/liblvm2cmd.cflow
+-include $(top_builddir)/daemons/dmeventd/$(LIB_NAME).cflow
+-include $(top_builddir)/daemons/dmeventd/plugins/mirror/$(LIB_NAME)-lvm2mirror.cflow
+endif
+
+install_include: $(srcdir)/libdevmapper-event.h
+	$(INSTALL_DATA) -D $< $(includedir)/$(<F)
+
+install_pkgconfig: libdevmapper-event.pc
+	$(INSTALL_DATA) -D $< $(pkgconfigdir)/devmapper-event.pc
+
+install_lib_dynamic: install_lib_shared
+
+install_lib_static: $(LIB_STATIC)
+	$(INSTALL_DATA) -D $< $(usrlibdir)/$(<F)
+
+install_lib: $(INSTALL_LIB_TARGETS)
+
+install_dmeventd_dynamic: dmeventd
+	$(INSTALL_PROGRAM) -D $< $(sbindir)/$(<F)
+
+install_dmeventd_static: dmeventd.static
+	$(INSTALL_PROGRAM) -D $< $(staticdir)/$(<F)
+
+install_dmeventd: $(INSTALL_DMEVENTD_TARGETS)
+
+install: install_include install_lib install_dmeventd
+
+install_device-mapper: install_include install_lib install_dmeventd
+
+DISTCLEAN_TARGETS += libdevmapper-event.pc

daemons/dmeventd/dmeventd.h

@@ -0,0 +1,75 @@
+/*
+ * Copyright (C) 2005-2007 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of the device-mapper userspace tools.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#ifndef __DMEVENTD_DOT_H__
+#define __DMEVENTD_DOT_H__
+
+/* FIXME This stuff must be configurable. */
+
+#define	DM_EVENT_FIFO_CLIENT	DEFAULT_DM_RUN_DIR "/dmeventd-client"
+#define	DM_EVENT_FIFO_SERVER	DEFAULT_DM_RUN_DIR "/dmeventd-server"
+
+#define DM_EVENT_DEFAULT_TIMEOUT 10
+
+/* Commands for the daemon passed in the message below. */
+enum dm_event_command {
+	DM_EVENT_CMD_ACTIVE = 1,
+	DM_EVENT_CMD_REGISTER_FOR_EVENT,
+	DM_EVENT_CMD_UNREGISTER_FOR_EVENT,
+	DM_EVENT_CMD_GET_REGISTERED_DEVICE,
+	DM_EVENT_CMD_GET_NEXT_REGISTERED_DEVICE,
+	DM_EVENT_CMD_SET_TIMEOUT,
+	DM_EVENT_CMD_GET_TIMEOUT,
+	DM_EVENT_CMD_HELLO,
+	DM_EVENT_CMD_DIE,
+	DM_EVENT_CMD_GET_STATUS,
+};
+
+/* Message passed between client and daemon. */
+struct dm_event_daemon_message {
+	uint32_t cmd;
+	uint32_t size;
+	char *data;
+};
+
+/* FIXME Is this meant to be exported?  I can't see where the
+   interface uses it. */
+/* Fifos for client/daemon communication. */
+struct dm_event_fifos {
+	int client;
+	int server;
+	const char *client_path;
+	const char *server_path;
+};
+
+/*      EXIT_SUCCESS             0 -- stdlib.h */
+/*      EXIT_FAILURE             1 -- stdlib.h */
+/*      EXIT_LOCKFILE_INUSE      2 -- obsoleted */
+#define EXIT_DESC_CLOSE_FAILURE  3
+#define EXIT_DESC_OPEN_FAILURE   4
+/*      EXIT_OPEN_PID_FAILURE    5 -- obsoleted */
+#define EXIT_FIFO_FAILURE        6
+#define EXIT_CHDIR_FAILURE       7
+
+/* Implemented in libdevmapper-event.c, but not part of public API. */
+// FIXME  misuse of bitmask as enum
+int daemon_talk(struct dm_event_fifos *fifos,
+		struct dm_event_daemon_message *msg, int cmd,
+		const char *dso_name, const char *dev_name,
+		enum dm_event_mask evmask, uint32_t timeout);
+int init_fifos(struct dm_event_fifos *fifos);
+void fini_fifos(struct dm_event_fifos *fifos);
+int dm_event_get_version(struct dm_event_fifos *fifos, int *version);
+
+#endif /* __DMEVENTD_DOT_H__ */

daemons/dmeventd/libdevmapper-event.c

@@ -0,0 +1,873 @@
+/*
+ * Copyright (C) 2005-2007 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of the device-mapper userspace tools.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include "dmlib.h"
+#include "libdevmapper-event.h"
+//#include "libmultilog.h"
+#include "dmeventd.h"
+
+#include <errno.h>
+#include <fcntl.h>
+#include <stdio.h>
+#include <stdint.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/file.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
+#include <sys/wait.h>
+#include <arpa/inet.h>		/* for htonl, ntohl */
+
+static int _sequence_nr = 0;
+
+struct dm_event_handler {
+	char *dso;
+
+	char *dmeventd_path;
+
+	char *dev_name;
+
+	char *uuid;
+	int major;
+	int minor;
+	uint32_t timeout;
+
+	enum dm_event_mask mask;
+};
+
+static void _dm_event_handler_clear_dev_info(struct dm_event_handler *dmevh)
+{
+	dm_free(dmevh->dev_name);
+	dm_free(dmevh->uuid);
+	dmevh->dev_name = dmevh->uuid = NULL;
+	dmevh->major = dmevh->minor = 0;
+}
+
+struct dm_event_handler *dm_event_handler_create(void)
+{
+	struct dm_event_handler *dmevh = NULL;
+
+	if (!(dmevh = dm_zalloc(sizeof(*dmevh)))) {
+		log_error("Failed to allocate event handler.");
+		return NULL;
+	}
+
+	return dmevh;
+}
+
+void dm_event_handler_destroy(struct dm_event_handler *dmevh)
+{
+	_dm_event_handler_clear_dev_info(dmevh);
+	dm_free(dmevh->dso);
+	dm_free(dmevh->dmeventd_path);
+	dm_free(dmevh);
+}
+
+int dm_event_handler_set_dmeventd_path(struct dm_event_handler *dmevh, const char *dmeventd_path)
+{
+	if (!dmeventd_path) /* noop */
+		return 0;
+
+	dm_free(dmevh->dmeventd_path);
+
+	dmevh->dmeventd_path = dm_strdup(dmeventd_path);
+	if (!dmevh->dmeventd_path)
+		return -ENOMEM;
+
+	return 0;
+}
+
+int dm_event_handler_set_dso(struct dm_event_handler *dmevh, const char *path)
+{
+	if (!path) /* noop */
+		return 0;
+	dm_free(dmevh->dso);
+
+	dmevh->dso = dm_strdup(path);
+	if (!dmevh->dso)
+		return -ENOMEM;
+
+	return 0;
+}
+
+int dm_event_handler_set_dev_name(struct dm_event_handler *dmevh, const char *dev_name)
+{
+	if (!dev_name)
+		return 0;
+
+	_dm_event_handler_clear_dev_info(dmevh);
+
+	dmevh->dev_name = dm_strdup(dev_name);
+	if (!dmevh->dev_name)
+		return -ENOMEM;
+	return 0;
+}
+
+int dm_event_handler_set_uuid(struct dm_event_handler *dmevh, const char *uuid)
+{
+	if (!uuid)
+		return 0;
+
+	_dm_event_handler_clear_dev_info(dmevh);
+
+	dmevh->uuid = dm_strdup(uuid);
+	if (!dmevh->uuid)
+		return -ENOMEM;
+	return 0;
+}
+
+void dm_event_handler_set_major(struct dm_event_handler *dmevh, int major)
+{
+	int minor = dmevh->minor;
+
+	_dm_event_handler_clear_dev_info(dmevh);
+
+	dmevh->major = major;
+	dmevh->minor = minor;
+}
+
+void dm_event_handler_set_minor(struct dm_event_handler *dmevh, int minor)
+{
+	int major = dmevh->major;
+
+	_dm_event_handler_clear_dev_info(dmevh);
+
+	dmevh->major = major;
+	dmevh->minor = minor;
+}
+
+void dm_event_handler_set_event_mask(struct dm_event_handler *dmevh,
+				     enum dm_event_mask evmask)
+{
+	dmevh->mask = evmask;
+}
+
+void dm_event_handler_set_timeout(struct dm_event_handler *dmevh, int timeout)
+{
+	dmevh->timeout = timeout;
+}
+
+const char *dm_event_handler_get_dso(const struct dm_event_handler *dmevh)
+{
+	return dmevh->dso;
+}
+
+const char *dm_event_handler_get_dev_name(const struct dm_event_handler *dmevh)
+{
+	return dmevh->dev_name;
+}
+
+const char *dm_event_handler_get_uuid(const struct dm_event_handler *dmevh)
+{
+	return dmevh->uuid;
+}
+
+int dm_event_handler_get_major(const struct dm_event_handler *dmevh)
+{
+	return dmevh->major;
+}
+
+int dm_event_handler_get_minor(const struct dm_event_handler *dmevh)
+{
+	return dmevh->minor;
+}
+
+int dm_event_handler_get_timeout(const struct dm_event_handler *dmevh)
+{
+	return dmevh->timeout;
+}
+
+enum dm_event_mask dm_event_handler_get_event_mask(const struct dm_event_handler *dmevh)
+{
+	return dmevh->mask;
+}
+
+static int _check_message_id(struct dm_event_daemon_message *msg)
+{
+	int pid, seq_nr;
+
+	if ((sscanf(msg->data, "%d:%d", &pid, &seq_nr) != 2) ||
+	    (pid != getpid()) || (seq_nr != _sequence_nr)) {
+		log_error("Ignoring out-of-sequence reply from dmeventd. "
+			  "Expected %d:%d but received %s", getpid(),
+			  _sequence_nr, msg->data);
+		return 0;
+	}
+
+	return 1;
+}
+
+/*
+ * daemon_read
+ * @fifos
+ * @msg
+ *
+ * Read message from daemon.
+ *
+ * Returns: 0 on failure, 1 on success
+ */
+static int _daemon_read(struct dm_event_fifos *fifos,
+			struct dm_event_daemon_message *msg)
+{
+	unsigned bytes = 0;
+	int ret, i;
+	fd_set fds;
+	struct timeval tval = { 0, 0 };
+	size_t size = 2 * sizeof(uint32_t);	/* status + size */
+	uint32_t *header = alloca(size);
+	char *buf = (char *)header;
+
+	while (bytes < size) {
+		for (i = 0, ret = 0; (i < 20) && (ret < 1); i++) {
+			/* Watch daemon read FIFO for input. */
+			FD_ZERO(&fds);
+			FD_SET(fifos->server, &fds);
+			tval.tv_sec = 1;
+			ret = select(fifos->server + 1, &fds, NULL, NULL,
+				     &tval);
+			if (ret < 0 && errno != EINTR) {
+				log_error("Unable to read from event server");
+				return 0;
+			}
+			if ((ret == 0) && (i > 4) && !bytes) {
+				log_error("No input from event server.");
+				return 0;
+			}
+		}
+		if (ret < 1) {
+			log_error("Unable to read from event server.");
+			return 0;
+		}
+
+		ret = read(fifos->server, buf + bytes, size);
+		if (ret < 0) {
+			if ((errno == EINTR) || (errno == EAGAIN))
+				continue;
+			else {
+				log_error("Unable to read from event server.");
+				return 0;
+			}
+		}
+
+		bytes += ret;
+		if (header && (bytes == 2 * sizeof(uint32_t))) {
+			msg->cmd = ntohl(header[0]);
+			msg->size = ntohl(header[1]);
+			buf = msg->data = dm_malloc(msg->size);
+			size = msg->size;
+			bytes = 0;
+			header = 0;
+		}
+	}
+
+	if (bytes != size) {
+		dm_free(msg->data);
+		msg->data = NULL;
+	}
+	return bytes == size;
+}
+
+/* Write message to daemon. */
+static int _daemon_write(struct dm_event_fifos *fifos,
+			 struct dm_event_daemon_message *msg)
+{
+	unsigned bytes = 0;
+	int ret = 0;
+	fd_set fds;
+
+	size_t size = 2 * sizeof(uint32_t) + msg->size;
+	uint32_t *header = alloca(size);
+	char *buf = (char *)header;
+	char drainbuf[128];
+	struct timeval tval = { 0, 0 };
+
+	header[0] = htonl(msg->cmd);
+	header[1] = htonl(msg->size);
+	memcpy(buf + 2 * sizeof(uint32_t), msg->data, msg->size);
+
+	/* drain the answer fifo */
+	while (1) {
+		FD_ZERO(&fds);
+		FD_SET(fifos->server, &fds);
+		tval.tv_usec = 100;
+		ret = select(fifos->server + 1, &fds, NULL, NULL, &tval);
+		if ((ret < 0) && (errno != EINTR)) {
+			log_error("Unable to talk to event daemon");
+			return 0;
+		}
+		if (ret == 0)
+			break;
+		ret = read(fifos->server, drainbuf, 127);
+	}
+
+	while (bytes < size) {
+		do {
+			/* Watch daemon write FIFO to be ready for output. */
+			FD_ZERO(&fds);
+			FD_SET(fifos->client, &fds);
+			ret = select(fifos->client + 1, NULL, &fds, NULL, NULL);
+			if ((ret < 0) && (errno != EINTR)) {
+				log_error("Unable to talk to event daemon");
+				return 0;
+			}
+		} while (ret < 1);
+
+		ret = write(fifos->client, buf + bytes, size - bytes);
+		if (ret < 0) {
+			if ((errno == EINTR) || (errno == EAGAIN))
+				continue;
+			else {
+				log_error("Unable to talk to event daemon");
+				return 0;
+			}
+		}
+
+		bytes += ret;
+	}
+
+	return bytes == size;
+}
+
+int daemon_talk(struct dm_event_fifos *fifos,
+		struct dm_event_daemon_message *msg, int cmd,
+		const char *dso_name, const char *dev_name,
+		enum dm_event_mask evmask, uint32_t timeout)
+{
+	const char *dso = dso_name ? dso_name : "-";
+	const char *dev = dev_name ? dev_name : "-";
+	const char *fmt = "%d:%d %s %s %u %" PRIu32;
+	int msg_size;
+	memset(msg, 0, sizeof(*msg));
+
+	/*
+	 * Set command and pack the arguments
+	 * into ASCII message string.
+	 */
+	msg->cmd = cmd;
+	if (cmd == DM_EVENT_CMD_HELLO)
+		fmt = "%d:%d HELLO";
+	if ((msg_size = dm_asprintf(&(msg->data), fmt, getpid(), _sequence_nr,
+				    dso, dev, evmask, timeout)) < 0) {
+		log_error("_daemon_talk: message allocation failed");
+		return -ENOMEM;
+	}
+	msg->size = msg_size;
+
+	/*
+	 * Write command and message to and
+	 * read status return code from daemon.
+	 */
+	if (!_daemon_write(fifos, msg)) {
+		stack;
+		dm_free(msg->data);
+		msg->data = 0;
+		return -EIO;
+	}
+
+	do {
+
+		dm_free(msg->data);
+		msg->data = 0;
+
+		if (!_daemon_read(fifos, msg)) {
+			stack;
+			return -EIO;
+		}
+	} while (!_check_message_id(msg));
+
+	_sequence_nr++;
+
+	return (int32_t) msg->cmd;
+}
+
+/*
+ * start_daemon
+ *
+ * This function forks off a process (dmeventd) that will handle
+ * the events.  I am currently test opening one of the fifos to
+ * ensure that the daemon is running and listening...  I thought
+ * this would be less expensive than fork/exec'ing every time.
+ * Perhaps there is an even quicker/better way (no, checking the
+ * lock file is _not_ a better way).
+ *
+ * Returns: 1 on success, 0 otherwise
+ */
+static int _start_daemon(char *dmeventd_path, struct dm_event_fifos *fifos)
+{
+	int pid, ret = 0;
+	int status;
+	struct stat statbuf;
+	char default_dmeventd_path[] = DMEVENTD_PATH;
+	char *args[] = { dmeventd_path ? : default_dmeventd_path, NULL };
+
+	if (stat(fifos->client_path, &statbuf))
+		goto start_server;
+
+	if (!S_ISFIFO(statbuf.st_mode)) {
+		log_error("%s is not a fifo.", fifos->client_path);
+		return 0;
+	}
+
+	/* Anyone listening?  If not, errno will be ENXIO */
+	fifos->client = open(fifos->client_path, O_WRONLY | O_NONBLOCK);
+	if (fifos->client >= 0) {
+		/* server is running and listening */
+		if (close(fifos->client))
+			log_sys_error("close", fifos->client_path);
+		return 1;
+	} else if (errno != ENXIO) {
+		/* problem */
+		log_sys_error("open", fifos->client_path);
+		return 0;
+	}
+
+      start_server:
+	/* server is not running */
+
+	if ((args[0][0] == '/') && stat(args[0], &statbuf)) {
+		log_sys_error("stat", args[0]);
+		return 0;
+	}
+
+	pid = fork();
+
+	if (pid < 0)
+		log_sys_error("fork", "");
+
+	else if (!pid) {
+		execvp(args[0], args);
+		log_error("Unable to exec dmeventd: %s", strerror(errno));
+		_exit(EXIT_FAILURE);
+	} else {
+		if (waitpid(pid, &status, 0) < 0)
+			log_error("Unable to start dmeventd: %s",
+				  strerror(errno));
+		else if (WEXITSTATUS(status))
+			log_error("Unable to start dmeventd.");
+		else
+			ret = 1;
+	}
+
+	return ret;
+}
+
+int init_fifos(struct dm_event_fifos *fifos)
+{
+	/* FIXME? Is fifo the most suitable method? Why not share
+	   comms/daemon code with something else e.g. multipath? */
+
+	/* FIXME Make these either configurable or depend directly on dmeventd_path */
+	fifos->client_path = DM_EVENT_FIFO_CLIENT;
+	fifos->server_path = DM_EVENT_FIFO_SERVER;
+
+	/* Open the fifo used to read from the daemon. */
+	if ((fifos->server = open(fifos->server_path, O_RDWR)) < 0) {
+		log_sys_error("open", fifos->server_path);
+		return 0;
+	}
+
+	/* Lock out anyone else trying to do communication with the daemon. */
+	if (flock(fifos->server, LOCK_EX) < 0) {
+		log_sys_error("flock", fifos->server_path);
+		if (close(fifos->server))
+			log_sys_error("close", fifos->server_path);
+		return 0;
+	}
+
+/*	if ((fifos->client = open(fifos->client_path, O_WRONLY | O_NONBLOCK)) < 0) {*/
+	if ((fifos->client = open(fifos->client_path, O_RDWR | O_NONBLOCK)) < 0) {
+		log_sys_error("open", fifos->client_path);
+		if (close(fifos->server))
+			log_sys_error("close", fifos->server_path);
+		return 0;
+	}
+
+	return 1;
+}
+
+/* Initialize client. */
+static int _init_client(char *dmeventd_path, struct dm_event_fifos *fifos)
+{
+	/* init fifos */
+	memset(fifos, 0, sizeof(*fifos));
+
+	/* FIXME Make these either configurable or depend directly on dmeventd_path */
+	fifos->client_path = DM_EVENT_FIFO_CLIENT;
+	fifos->server_path = DM_EVENT_FIFO_SERVER;
+
+	if (!_start_daemon(dmeventd_path, fifos))
+		return_0;
+
+	return init_fifos(fifos);
+}
+
+void fini_fifos(struct dm_event_fifos *fifos)
+{
+	if (flock(fifos->server, LOCK_UN))
+		log_error("flock unlock %s", fifos->server_path);
+
+	if (close(fifos->client))
+		log_sys_error("close", fifos->client_path);
+	if (close(fifos->server))
+		log_sys_error("close", fifos->server_path);
+}
+
+/* Get uuid of a device */
+static struct dm_task *_get_device_info(const struct dm_event_handler *dmevh)
+{
+	struct dm_task *dmt;
+	struct dm_info info;
+
+	if (!(dmt = dm_task_create(DM_DEVICE_INFO))) {
+		log_error("_get_device_info: dm_task creation for info failed");
+		return NULL;
+	}
+
+	if (dmevh->uuid) {
+		if (!dm_task_set_uuid(dmt, dmevh->uuid))
+			goto_bad;
+	} else if (dmevh->dev_name) {
+		if (!dm_task_set_name(dmt, dmevh->dev_name))
+			goto_bad;
+	} else if (dmevh->major && dmevh->minor) {
+		if (!dm_task_set_major(dmt, dmevh->major) ||
+		    !dm_task_set_minor(dmt, dmevh->minor))
+			goto_bad;
+	}
+
+	/* FIXME Add name or uuid or devno to messages */
+	if (!dm_task_run(dmt)) {
+		log_error("_get_device_info: dm_task_run() failed");
+		goto bad;
+	}
+
+	if (!dm_task_get_info(dmt, &info)) {
+		log_error("_get_device_info: failed to get info for device");
+		goto bad;
+	}
+
+	if (!info.exists) {
+		log_error("_get_device_info: %s%s%s%.0d%s%.0d%s%s: device not found",
+			  dmevh->uuid ? : "", 
+			  (!dmevh->uuid && dmevh->dev_name) ? dmevh->dev_name : "", 
+			  (!dmevh->uuid && !dmevh->dev_name && dmevh->major > 0) ? "(" : "",
+			  (!dmevh->uuid && !dmevh->dev_name && dmevh->major > 0) ? dmevh->major : 0,
+			  (!dmevh->uuid && !dmevh->dev_name && dmevh->major > 0) ? ":" : "",
+			  (!dmevh->uuid && !dmevh->dev_name && dmevh->minor > 0) ? dmevh->minor : 0,
+			  (!dmevh->uuid && !dmevh->dev_name && dmevh->major > 0) && dmevh->minor == 0 ? "0" : "",
+			  (!dmevh->uuid && !dmevh->dev_name && dmevh->major > 0) ? ") " : "");
+		goto bad;
+	}
+
+		  
+	return dmt;
+
+      bad:
+	dm_task_destroy(dmt);
+	return NULL;
+}
+
+/* Handle the event (de)registration call and return negative error codes. */
+static int _do_event(int cmd, char *dmeventd_path, struct dm_event_daemon_message *msg,
+		     const char *dso_name, const char *dev_name,
+		     enum dm_event_mask evmask, uint32_t timeout)
+{
+	int ret;
+	struct dm_event_fifos fifos;
+
+	if (!_init_client(dmeventd_path, &fifos)) {
+		stack;
+		return -ESRCH;
+	}
+
+	ret = daemon_talk(&fifos, msg, DM_EVENT_CMD_HELLO, NULL, NULL, 0, 0);
+
+	dm_free(msg->data);
+	msg->data = 0;
+
+	if (!ret)
+		ret = daemon_talk(&fifos, msg, cmd, dso_name, dev_name, evmask, timeout);
+
+	/* what is the opposite of init? */
+	fini_fifos(&fifos);
+
+	return ret;
+}
+
+/* External library interface. */
+int dm_event_register_handler(const struct dm_event_handler *dmevh)
+{
+	int ret = 1, err;
+	const char *uuid;
+	struct dm_task *dmt;
+	struct dm_event_daemon_message msg = { 0, 0, NULL };
+
+	if (!(dmt = _get_device_info(dmevh))) {
+		stack;
+		return 0;
+	}
+
+	uuid = dm_task_get_uuid(dmt);
+
+	if ((err = _do_event(DM_EVENT_CMD_REGISTER_FOR_EVENT, dmevh->dmeventd_path, &msg,
+			     dmevh->dso, uuid, dmevh->mask, dmevh->timeout)) < 0) {
+		log_error("%s: event registration failed: %s",
+			  dm_task_get_name(dmt),
+			  msg.data ? msg.data : strerror(-err));
+		ret = 0;
+	}
+
+	dm_free(msg.data);
+
+	dm_task_destroy(dmt);
+
+	return ret;
+}
+
+int dm_event_unregister_handler(const struct dm_event_handler *dmevh)
+{
+	int ret = 1, err;
+	const char *uuid;
+	struct dm_task *dmt;
+	struct dm_event_daemon_message msg = { 0, 0, NULL };
+
+	if (!(dmt = _get_device_info(dmevh))) {
+		stack;
+		return 0;
+	}
+
+	uuid = dm_task_get_uuid(dmt);
+
+	if ((err = _do_event(DM_EVENT_CMD_UNREGISTER_FOR_EVENT, dmevh->dmeventd_path, &msg,
+			    dmevh->dso, uuid, dmevh->mask, dmevh->timeout)) < 0) {
+		log_error("%s: event deregistration failed: %s",
+			  dm_task_get_name(dmt),
+			  msg.data ? msg.data : strerror(-err));
+		ret = 0;
+	}
+
+	dm_free(msg.data);
+
+	dm_task_destroy(dmt);
+
+	return ret;
+}
+
+/* Fetch a string off src and duplicate it into *dest. */
+/* FIXME: move to separate module to share with the daemon. */
+static char *_fetch_string(char **src, const int delimiter)
+{
+	char *p, *ret;
+
+	if ((p = strchr(*src, delimiter)))
+		*p = 0;
+
+	if ((ret = dm_strdup(*src)))
+		*src += strlen(ret) + 1;
+
+	if (p)
+		*p = delimiter;
+
+	return ret;
+}
+
+/* Parse a device message from the daemon. */
+static int _parse_message(struct dm_event_daemon_message *msg, char **dso_name,
+			 char **uuid, enum dm_event_mask *evmask)
+{
+	char *id = NULL;
+	char *p = msg->data;
+
+	if ((id = _fetch_string(&p, ' ')) &&
+	    (*dso_name = _fetch_string(&p, ' ')) &&
+	    (*uuid = _fetch_string(&p, ' '))) {
+		*evmask = atoi(p);
+
+		dm_free(id);
+		return 0;
+	}
+
+	if (id)
+		dm_free(id);
+	return -ENOMEM;
+}
+
+/*
+ * Returns 0 if handler found; error (-ENOMEM, -ENOENT) otherwise.
+ */
+int dm_event_get_registered_device(struct dm_event_handler *dmevh, int next)
+{
+	int ret = 0;
+	const char *uuid = NULL;
+	char *reply_dso = NULL, *reply_uuid = NULL;
+	enum dm_event_mask reply_mask = 0;
+	struct dm_task *dmt = NULL;
+	struct dm_event_daemon_message msg = { 0 };
+	struct dm_info info;
+
+	if (!(dmt = _get_device_info(dmevh))) {
+		log_debug("Device does not exists (uuid=%s, name=%s, %d:%d).",
+			  dmevh->uuid, dmevh->dev_name,
+			  dmevh->major, dmevh->minor);
+		ret = -ENODEV;
+		goto fail;
+	}
+
+	uuid = dm_task_get_uuid(dmt);
+
+	if (_do_event(next ? DM_EVENT_CMD_GET_NEXT_REGISTERED_DEVICE :
+		      DM_EVENT_CMD_GET_REGISTERED_DEVICE, dmevh->dmeventd_path,
+		      &msg, dmevh->dso, uuid, dmevh->mask, 0)) {
+		log_debug("%s: device not registered.", dm_task_get_name(dmt));
+		ret = -ENOENT;
+		goto fail;
+	}
+
+	/* FIXME this will probably horribly break if we get
+	   ill-formatted reply */
+	ret = _parse_message(&msg, &reply_dso, &reply_uuid, &reply_mask);
+
+	dm_task_destroy(dmt);
+	dmt = NULL;
+
+	dm_free(msg.data);
+	msg.data = NULL;
+
+	_dm_event_handler_clear_dev_info(dmevh);
+	if (!reply_uuid) {
+		ret = -ENXIO; /* dmeventd probably gave us bogus uuid back */
+		goto fail;
+	}
+	dmevh->uuid = dm_strdup(reply_uuid);
+	if (!dmevh->uuid) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	if (!(dmt = _get_device_info(dmevh))) {
+		ret = -ENXIO; /* dmeventd probably gave us bogus uuid back */
+		goto fail;
+	}
+
+	dm_event_handler_set_dso(dmevh, reply_dso);
+	dm_event_handler_set_event_mask(dmevh, reply_mask);
+
+	dm_free(reply_dso);
+	reply_dso = NULL;
+
+	dm_free(reply_uuid);
+	reply_uuid = NULL;
+
+	dmevh->dev_name = dm_strdup(dm_task_get_name(dmt));
+	if (!dmevh->dev_name) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	if (!dm_task_get_info(dmt, &info)) {
+		ret = -1;
+		goto fail;
+	}
+
+	dmevh->major = info.major;
+	dmevh->minor = info.minor;
+
+	dm_task_destroy(dmt);
+
+	return ret;
+
+ fail:
+	dm_free(msg.data);
+	dm_free(reply_dso);
+	dm_free(reply_uuid);
+	_dm_event_handler_clear_dev_info(dmevh);
+	if (dmt)
+		dm_task_destroy(dmt);
+	return ret;
+}
+
+/*
+ * You can (and have to) call this at the stage of the protocol where
+ *     daemon_talk(fifos, &msg, DM_EVENT_CMD_HELLO, NULL, NULL, 0, 0)
+ *
+ * would be normally sent. This call will parse the version reply from
+ * dmeventd, in addition to above call. It is not safe to call this at any
+ * other place in the protocol.
+ *
+ * This is an internal function, not exposed in the public API.
+ */
+
+int dm_event_get_version(struct dm_event_fifos *fifos, int *version) {
+	char *p;
+	struct dm_event_daemon_message msg = { 0, 0, NULL };
+
+	if (daemon_talk(fifos, &msg, DM_EVENT_CMD_HELLO, NULL, NULL, 0, 0))
+		return 0;
+	p = msg.data;
+	*version = 0;
+
+	if (!p || !(p = strchr(p, ' '))) /* Message ID */
+		return 0;
+	if (!(p = strchr(p + 1, ' '))) /* HELLO */
+		return 0;
+	if ((p = strchr(p + 1, ' '))) /* HELLO, once more */
+		*version = atoi(p);
+
+	return 1;
+}
+
+#if 0				/* left out for now */
+
+static char *_skip_string(char *src, const int delimiter)
+{
+	src = srtchr(src, delimiter);
+	if (src && *(src + 1))
+		return src + 1;
+	return NULL;
+}
+
+int dm_event_set_timeout(const char *device_path, uint32_t timeout)
+{
+	struct dm_event_daemon_message msg = { 0, 0, NULL };
+
+	if (!device_exists(device_path))
+		return -ENODEV;
+
+	return _do_event(DM_EVENT_CMD_SET_TIMEOUT, &msg,
+			 NULL, device_path, 0, timeout);
+}
+
+int dm_event_get_timeout(const char *device_path, uint32_t *timeout)
+{
+	int ret;
+	struct dm_event_daemon_message msg = { 0, 0, NULL };
+
+	if (!device_exists(device_path))
+		return -ENODEV;
+	if (!(ret = _do_event(DM_EVENT_CMD_GET_TIMEOUT, &msg, NULL, device_path,
+			     0, 0))) {
+		char *p = _skip_string(msg.data, ' ');
+		if (!p) {
+			log_error("malformed reply from dmeventd '%s'\n",
+				  msg.data);
+			return -EIO;
+		}
+		*timeout = atoi(p);
+	}
+	if (msg.data)
+		dm_free(msg.data);
+	return ret;
+}
+#endif

daemons/dmeventd/libdevmapper-event.h

@@ -0,0 +1,115 @@
+/*
+ * Copyright (C) 2005-2007 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of the device-mapper userspace tools.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+/*
+ * Note that this file is released only as part of a technology preview
+ * and its contents may change in future updates in ways that do not
+ * preserve compatibility.
+ */
+
+#ifndef LIB_DMEVENT_H
+#define LIB_DMEVENT_H
+
+#include <stdint.h>
+
+/*
+ * Event library interface.
+ */
+
+enum dm_event_mask {
+	DM_EVENT_SETTINGS_MASK  = 0x0000FF,
+	DM_EVENT_SINGLE		= 0x000001, /* Report multiple errors just once. */
+	DM_EVENT_MULTI		= 0x000002, /* Report all of them. */
+
+	DM_EVENT_ERROR_MASK     = 0x00FF00,
+	DM_EVENT_SECTOR_ERROR	= 0x000100, /* Failure on a particular sector. */
+	DM_EVENT_DEVICE_ERROR	= 0x000200, /* Device failure. */
+	DM_EVENT_PATH_ERROR	= 0x000400, /* Failure on an io path. */
+	DM_EVENT_ADAPTOR_ERROR	= 0x000800, /* Failure of a host adaptor. */
+
+	DM_EVENT_STATUS_MASK    = 0xFF0000,
+	DM_EVENT_SYNC_STATUS	= 0x010000, /* Mirror synchronization completed/failed. */
+	DM_EVENT_TIMEOUT	= 0x020000, /* Timeout has occured */
+
+	DM_EVENT_REGISTRATION_PENDING = 0x1000000, /* Monitor thread is setting-up/shutting-down */
+};
+
+#define DM_EVENT_ALL_ERRORS DM_EVENT_ERROR_MASK
+#define DM_EVENT_PROTOCOL_VERSION 1
+
+struct dm_event_handler;
+
+struct dm_event_handler *dm_event_handler_create(void);
+void dm_event_handler_destroy(struct dm_event_handler *dmevh);
+
+/*
+ * Path of shared library to handle events.
+ *
+ * All of dmeventd, dso, device_name and uuid strings are duplicated so
+ * you do not need to keep the pointers valid after the call succeeds.
+ * They may return -ENOMEM though.
+ */
+int dm_event_handler_set_dso(struct dm_event_handler *dmevh, const char *path);
+
+/*
+ * Path of dmeventd binary.
+ */
+int dm_event_handler_set_dmeventd_path(struct dm_event_handler *dmevh, const char *dmeventd_path);
+
+/*
+ * Identify the device to monitor by exactly one of device_name, uuid or
+ * device number. String arguments are duplicated, see above.
+ */
+int dm_event_handler_set_dev_name(struct dm_event_handler *dmevh, const char *device_name);
+
+int dm_event_handler_set_uuid(struct dm_event_handler *dmevh, const char *uuid);
+
+void dm_event_handler_set_major(struct dm_event_handler *dmevh, int major);
+void dm_event_handler_set_minor(struct dm_event_handler *dmevh, int minor);
+void dm_event_handler_set_timeout(struct dm_event_handler *dmevh, int timeout);
+
+/*
+ * Specify mask for events to monitor.
+ */
+// FIXME  misuse of bitmask as enum
+void dm_event_handler_set_event_mask(struct dm_event_handler *dmevh,
+				     enum dm_event_mask evmask);
+
+const char *dm_event_handler_get_dso(const struct dm_event_handler *dmevh);
+const char *dm_event_handler_get_dev_name(const struct dm_event_handler *dmevh);
+const char *dm_event_handler_get_uuid(const struct dm_event_handler *dmevh);
+int dm_event_handler_get_major(const struct dm_event_handler *dmevh);
+int dm_event_handler_get_minor(const struct dm_event_handler *dmevh);
+int dm_event_handler_get_timeout(const struct dm_event_handler *dmevh);
+// FIXME  misuse of bitmask as enum
+enum dm_event_mask dm_event_handler_get_event_mask(const struct dm_event_handler *dmevh);
+
+/* FIXME Review interface (what about this next thing?) */
+int dm_event_get_registered_device(struct dm_event_handler *dmevh, int next);
+
+/*
+ * Initiate monitoring using dmeventd.
+ */
+int dm_event_register_handler(const struct dm_event_handler *dmevh);
+int dm_event_unregister_handler(const struct dm_event_handler *dmevh);
+
+/* Prototypes for DSO interface, see dmeventd.c, struct dso_data for
+   detailed descriptions. */
+// FIXME  misuse of bitmask as enum
+void process_event(struct dm_task *dmt, enum dm_event_mask evmask, void **user);
+int register_device(const char *device_name, const char *uuid, int major, int minor, void **user);
+int unregister_device(const char *device_name, const char *uuid, int major,
+		      int minor, void **user);
+
+#endif

daemons/dmeventd/libdevmapper-event.pc.in

@@ -0,0 +1,11 @@
+prefix=@prefix@
+exec_prefix=@exec_prefix@
+libdir=@libdir@
+includedir=@includedir@
+
+Name: devmapper-event
+Description: device-mapper event library
+Version: @DM_LIB_PATCHLEVEL@
+Cflags: -I${includedir}
+Libs: -L${libdir} -ldevmapper-event
+Requires.private: devmapper

daemons/dmeventd/plugins/Makefile.in

@@ -0,0 +1,46 @@
+#
+# Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
+# Copyright (C) 2004-2005, 2011 Red Hat, Inc. All rights reserved.
+#
+# This file is part of LVM2.
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions
+# of the GNU General Public License v.2.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+srcdir = @srcdir@
+top_srcdir = @top_srcdir@
+top_builddir = @top_builddir@
+
+SUBDIRS += lvm2
+
+ifneq ("@MIRRORS@", "none")
+  SUBDIRS += mirror
+endif
+
+ifneq ("@SNAPSHOTS@", "none")
+  SUBDIRS += snapshot
+endif
+
+ifneq ("@RAID@", "none")
+  SUBDIRS += raid
+endif
+
+ifneq ("@THIN@", "none")
+  SUBDIRS += thin
+endif
+
+ifeq ($(MAKECMDGOALS),distclean)
+  SUBDIRS = lvm2 mirror snapshot raid thin
+endif
+
+include $(top_builddir)/make.tmpl
+
+snapshot: lvm2
+mirror: lvm2
+raid: lvm2
+thin: lvm2

daemons/dmeventd/plugins/lvm2/.exported_symbols

@@ -0,0 +1,7 @@
+dmeventd_lvm2_init
+dmeventd_lvm2_exit
+dmeventd_lvm2_lock
+dmeventd_lvm2_unlock
+dmeventd_lvm2_pool
+dmeventd_lvm2_run
+dmeventd_lvm2_command

daemons/dmeventd/plugins/lvm2/Makefile.in

@@ -0,0 +1,31 @@
+#
+# Copyright (C) 2010-2011 Red Hat, Inc. All rights reserved.
+#
+# This file is part of LVM2.
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions
+# of the GNU General Public License v.2.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+srcdir = @srcdir@
+top_srcdir = @top_srcdir@
+top_builddir = @top_builddir@
+
+CLDFLAGS += -L$(top_builddir)/tools
+
+SOURCES = dmeventd_lvm.c
+
+LIB_SHARED = libdevmapper-event-lvm2.$(LIB_SUFFIX)
+LIB_VERSION = $(LIB_VERSION_LVM)
+
+include $(top_builddir)/make.tmpl
+
+LIBS += @LVM2CMD_LIB@ -ldevmapper $(PTHREAD_LIBS) $(DAEMON_LIBS)
+
+install_lvm2: install_lib_shared
+
+install: install_lvm2

daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c

@@ -0,0 +1,176 @@
+/*
+ * Copyright (C) 2010 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include "lib.h"
+#include "log.h"
+
+#include "lvm2cmd.h"
+#include "dmeventd_lvm.h"
+
+#include <pthread.h>
+#include <syslog.h>
+
+extern int dmeventd_debug;
+
+/*
+ * register_device() is called first and performs initialisation.
+ * Only one device may be registered or unregistered at a time.
+ */
+static pthread_mutex_t _register_mutex = PTHREAD_MUTEX_INITIALIZER;
+
+/*
+ * Number of active registrations.
+ */
+static int _register_count = 0;
+static struct dm_pool *_mem_pool = NULL;
+static void *_lvm_handle = NULL;
+
+/*
+ * Currently only one event can be processed at a time.
+ */
+static pthread_mutex_t _event_mutex = PTHREAD_MUTEX_INITIALIZER;
+
+/*
+ * FIXME Do not pass things directly to syslog, rather use the existing logging
+ * facilities to sort logging ... however that mechanism needs to be somehow
+ * configurable and we don't have that option yet
+ */
+static void _temporary_log_fn(int level,
+			      const char *file __attribute__((unused)),
+			      int line __attribute__((unused)),
+			      int dm_errno __attribute__((unused)),
+			      const char *message)
+{
+	level &= ~(_LOG_STDERR | _LOG_ONCE);
+
+	switch (level) {
+	case _LOG_DEBUG:
+		if (dmeventd_debug >= 3)
+			syslog(LOG_DEBUG, "%s", message);
+		break;
+	case _LOG_INFO:
+		if (dmeventd_debug >= 2)
+			syslog(LOG_INFO, "%s", message);
+		break;
+	case _LOG_NOTICE:
+		if (dmeventd_debug >= 1)
+			syslog(LOG_NOTICE, "%s", message);
+		break;
+	case _LOG_WARN:
+		syslog(LOG_WARNING, "%s", message);
+		break;
+	case _LOG_ERR:
+		syslog(LOG_ERR, "%s", message);
+		break;
+	default:
+		syslog(LOG_CRIT, "%s", message);
+	}
+}
+
+void dmeventd_lvm2_lock(void)
+{
+	pthread_mutex_lock(&_event_mutex);
+}
+
+void dmeventd_lvm2_unlock(void)
+{
+	pthread_mutex_unlock(&_event_mutex);
+}
+
+int dmeventd_lvm2_init(void)
+{
+	int r = 0;
+
+	pthread_mutex_lock(&_register_mutex);
+
+	/*
+	 * Need some space for allocations.  1024 should be more
+	 * than enough for what we need (device mapper name splitting)
+	 */
+	if (!_mem_pool && !(_mem_pool = dm_pool_create("mirror_dso", 1024)))
+		goto out;
+
+	if (!_lvm_handle) {
+		lvm2_log_fn(_temporary_log_fn);
+		if (!(_lvm_handle = lvm2_init())) {
+			dm_pool_destroy(_mem_pool);
+			_mem_pool = NULL;
+			goto out;
+		}
+		lvm2_disable_dmeventd_monitoring(_lvm_handle);
+		/* FIXME Temporary: move to dmeventd core */
+		lvm2_run(_lvm_handle, "_memlock_inc");
+	}
+
+	_register_count++;
+	r = 1;
+
+out:
+	pthread_mutex_unlock(&_register_mutex);
+	return r;
+}
+
+void dmeventd_lvm2_exit(void)
+{
+	pthread_mutex_lock(&_register_mutex);
+
+	if (!--_register_count) {
+		lvm2_run(_lvm_handle, "_memlock_dec");
+		dm_pool_destroy(_mem_pool);
+		_mem_pool = NULL;
+		lvm2_exit(_lvm_handle);
+		_lvm_handle = NULL;
+	}
+
+	pthread_mutex_unlock(&_register_mutex);
+}
+
+struct dm_pool *dmeventd_lvm2_pool(void)
+{
+	return _mem_pool;
+}
+
+int dmeventd_lvm2_run(const char *cmdline)
+{
+	return lvm2_run(_lvm_handle, cmdline);
+}
+
+int dmeventd_lvm2_command(struct dm_pool *mem, char *buffer, size_t size,
+			  const char *cmd, const char *device)
+{
+	char *vg = NULL, *lv = NULL, *layer;
+	int r;
+
+	if (!dm_split_lvm_name(mem, device, &vg, &lv, &layer)) {
+		syslog(LOG_ERR, "Unable to determine VG name from %s.\n",
+		       device);
+		return 0;
+	}
+
+	/* strip off the mirror component designations */
+	layer = strstr(lv, "_mlog");
+	if (layer)
+		*layer = '\0';
+
+	r = dm_snprintf(buffer, size, "%s %s/%s", cmd, vg, lv);
+
+	dm_pool_free(mem, vg);
+
+	if (r < 0) {
+		syslog(LOG_ERR, "Unable to form LVM command. (too long).\n");
+		return 0;
+	}
+
+	return 1;
+}

daemons/dmeventd/plugins/lvm2/dmeventd_lvm.h

@@ -0,0 +1,42 @@
+/*
+ * Copyright (C) 2010 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+/*
+ * Wrappers around liblvm2cmd functions for dmeventd plug-ins.
+ *
+ * liblvm2cmd is not thread-safe so the locking in this library helps dmeventd
+ * threads to co-operate in sharing a single instance.
+ *
+ * FIXME Either support this properly as a generic liblvm2cmd wrapper or make
+ * liblvm2cmd thread-safe so this can go away.
+ */
+
+#include "libdevmapper.h"
+
+#ifndef _DMEVENTD_LVMWRAP_H
+#define _DMEVENTD_LVMWRAP_H
+
+int dmeventd_lvm2_init(void);
+void dmeventd_lvm2_exit(void);
+int dmeventd_lvm2_run(const char *cmdline);
+
+void dmeventd_lvm2_lock(void);
+void dmeventd_lvm2_unlock(void);
+
+struct dm_pool *dmeventd_lvm2_pool(void);
+
+int dmeventd_lvm2_command(struct dm_pool *mem, char *buffer, size_t size,
+			  const char *cmd, const char *device);
+
+#endif /* _DMEVENTD_LVMWRAP_H */

daemons/dmeventd/plugins/mirror/.exported_symbols

@@ -0,0 +1,3 @@
+process_event
+register_device
+unregister_device

daemons/dmeventd/plugins/mirror/Makefile.in

@@ -0,0 +1,37 @@
+#
+# Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
+# Copyright (C) 2004-2005, 2008-2011 Red Hat, Inc. All rights reserved.
+#
+# This file is part of LVM2.
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions
+# of the GNU General Public License v.2.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+srcdir = @srcdir@
+top_srcdir = @top_srcdir@
+top_builddir = @top_builddir@
+
+INCLUDES += -I$(top_srcdir)/tools -I$(top_srcdir)/daemons/dmeventd/plugins/lvm2
+CLDFLAGS += -L$(top_builddir)/tools -L$(top_builddir)/daemons/dmeventd/plugins/lvm2
+
+SOURCES = dmeventd_mirror.c
+
+LIB_NAME = libdevmapper-event-lvm2mirror
+LIB_SHARED = $(LIB_NAME).$(LIB_SUFFIX)
+LIB_VERSION = $(LIB_VERSION_LVM)
+
+CFLOW_LIST = $(SOURCES)
+CFLOW_LIST_TARGET = $(LIB_NAME).cflow
+
+include $(top_builddir)/make.tmpl
+
+LIBS += -ldevmapper-event-lvm2 -ldevmapper $(DAEMON_LIBS)
+
+install_lvm2: install_dm_plugin
+
+install: install_lvm2

daemons/dmeventd/plugins/mirror/dmeventd_mirror.c

@@ -0,0 +1,235 @@
+/*
+ * Copyright (C) 2005-2012 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include "lib.h"
+
+#include "lvm2cmd.h"
+#include "errors.h"
+#include "libdevmapper-event.h"
+#include "dmeventd_lvm.h"
+#include "defaults.h"
+
+#include <syslog.h> /* FIXME Replace syslog with multilog */
+/* FIXME Missing openlog? */
+/* FIXME Replace most syslogs with log_error() style messages and add complete context. */
+/* FIXME Reformat to 80 char lines. */
+
+#define ME_IGNORE    0
+#define ME_INSYNC    1
+#define ME_FAILURE   2
+
+static int _process_status_code(const char status_code, const char *dev_name,
+				const char *dev_type, int r)
+{
+	/*
+	 *    A => Alive - No failures
+	 *    D => Dead - A write failure occurred leaving mirror out-of-sync
+	 *    F => Flush failed.
+	 *    S => Sync - A sychronization failure occurred, mirror out-of-sync
+	 *    R => Read - A read failure occurred, mirror data unaffected
+	 *    U => Unclassified failure (bug)
+	 */ 
+	if (status_code == 'F') {
+		syslog(LOG_ERR, "%s device %s flush failed.",
+		       dev_type, dev_name);
+		r = ME_FAILURE;
+	} else if (status_code == 'S')
+		syslog(LOG_ERR, "%s device %s sync failed.",
+		       dev_type, dev_name);
+	else if (status_code == 'R')
+		syslog(LOG_ERR, "%s device %s read failed.",
+		       dev_type, dev_name);
+	else if (status_code != 'A') {
+		syslog(LOG_ERR, "%s device %s has failed (%c).",
+		       dev_type, dev_name, status_code);
+		r = ME_FAILURE;
+	}
+
+	return r;
+}
+
+static int _get_mirror_event(char *params)
+{
+	int i, r = ME_INSYNC;
+	char **args = NULL;
+	char *dev_status_str;
+	char *log_status_str;
+	char *sync_str;
+	char *p = NULL;
+	int log_argc, num_devs;
+
+	/*
+	 * dm core parms:	     0 409600 mirror
+	 * Mirror core parms:	     2 253:4 253:5 400/400
+	 * New-style failure params: 1 AA
+	 * New-style log params:     3 cluster 253:3 A
+	 *			 or  3 disk 253:3 A
+	 *			 or  1 core
+	 */
+
+	/* number of devices */
+	if (!dm_split_words(params, 1, 0, &p))
+		goto out_parse;
+
+	if (!(num_devs = atoi(p)) ||
+	    (num_devs > DEFAULT_MIRROR_MAX_IMAGES) || (num_devs < 0))
+		goto out_parse;
+	p += strlen(p) + 1;
+
+	/* devices names + "400/400" + "1 AA" + 1 or 3 log parms + NULL */
+	args = dm_malloc((num_devs + 7) * sizeof(char *));
+	if (!args || dm_split_words(p, num_devs + 7, 0, args) < num_devs + 5)
+		goto out_parse;
+
+	/* FIXME: Code differs from lib/mirror/mirrored.c */
+	dev_status_str = args[2 + num_devs];
+	log_argc = atoi(args[3 + num_devs]);
+	log_status_str = args[3 + num_devs + log_argc];
+	sync_str = args[num_devs];
+
+	/* Check for bad mirror devices */
+	for (i = 0; i < num_devs; i++)
+		r = _process_status_code(dev_status_str[i], args[i],
+			i ? "Secondary mirror" : "Primary mirror", r);
+
+	/* Check for bad disk log device */
+	if (log_argc > 1)
+		r = _process_status_code(log_status_str[0],
+					 args[2 + num_devs + log_argc],
+					 "Log", r);
+
+	if (r == ME_FAILURE)
+		goto out;
+
+	p = strstr(sync_str, "/");
+	if (p) {
+		p[0] = '\0';
+		if (strcmp(sync_str, p+1))
+			r = ME_IGNORE;
+		p[0] = '/';
+	} else
+		goto out_parse;
+
+out:
+	dm_free(args);
+	return r;
+
+out_parse:
+	dm_free(args);
+	syslog(LOG_ERR, "Unable to parse mirror status string.");
+	return ME_IGNORE;
+}
+
+static int _remove_failed_devices(const char *device)
+{
+	int r;
+#define CMD_SIZE 256	/* FIXME Use system restriction */
+	char cmd_str[CMD_SIZE];
+
+	if (!dmeventd_lvm2_command(dmeventd_lvm2_pool(), cmd_str, sizeof(cmd_str),
+				  "lvconvert --config devices{ignore_suspended_devices=1} "
+				  "--repair --use-policies", device))
+		return -ENAMETOOLONG; /* FIXME Replace with generic error return - reason for failure has already got logged */
+
+	r = dmeventd_lvm2_run(cmd_str);
+
+	syslog(LOG_INFO, "Repair of mirrored device %s %s.", device,
+	       (r == ECMD_PROCESSED) ? "finished successfully" : "failed");
+
+	return (r == ECMD_PROCESSED) ? 0 : -1;
+}
+
+void process_event(struct dm_task *dmt,
+		   enum dm_event_mask event __attribute__((unused)),
+		   void **unused __attribute__((unused)))
+{
+	void *next = NULL;
+	uint64_t start, length;
+	char *target_type = NULL;
+	char *params;
+	const char *device = dm_task_get_name(dmt);
+
+	dmeventd_lvm2_lock();
+
+	do {
+		next = dm_get_next_target(dmt, next, &start, &length,
+					  &target_type, &params);
+
+		if (!target_type) {
+			syslog(LOG_INFO, "%s mapping lost.", device);
+			continue;
+		}
+
+		if (strcmp(target_type, "mirror")) {
+			syslog(LOG_INFO, "%s has unmirrored portion.", device);
+			continue;
+		}
+
+		switch(_get_mirror_event(params)) {
+		case ME_INSYNC:
+			/* FIXME: all we really know is that this
+			   _part_ of the device is in sync
+			   Also, this is not an error
+			*/
+			syslog(LOG_NOTICE, "%s is now in-sync.", device);
+			break;
+		case ME_FAILURE:
+			syslog(LOG_ERR, "Device failure in %s.", device);
+			if (_remove_failed_devices(device))
+				/* FIXME Why are all the error return codes unused? Get rid of them? */
+				syslog(LOG_ERR, "Failed to remove faulty devices in %s.",
+				       device);
+			/* Should check before warning user that device is now linear
+			else
+				syslog(LOG_NOTICE, "%s is now a linear device.\n",
+					device);
+			*/
+			break;
+		case ME_IGNORE:
+			break;
+		default:
+			/* FIXME Provide value then! */
+			syslog(LOG_INFO, "Unknown event received.");
+		}
+	} while (next);
+
+	dmeventd_lvm2_unlock();
+}
+
+int register_device(const char *device,
+		    const char *uuid __attribute__((unused)),
+		    int major __attribute__((unused)),
+		    int minor __attribute__((unused)),
+		    void **unused __attribute__((unused)))
+{
+	if (!dmeventd_lvm2_init())
+		return 0;
+
+	syslog(LOG_INFO, "Monitoring mirror device %s for events.", device);
+
+	return 1;
+}
+
+int unregister_device(const char *device,
+		      const char *uuid __attribute__((unused)),
+		      int major __attribute__((unused)),
+		      int minor __attribute__((unused)),
+		      void **unused __attribute__((unused)))
+{
+	syslog(LOG_INFO, "No longer monitoring mirror device %s for events.",
+	       device);
+	dmeventd_lvm2_exit();
+
+	return 1;
+}

daemons/dmeventd/plugins/raid/.exported_symbols

@@ -0,0 +1,3 @@
+process_event
+register_device
+unregister_device

daemons/dmeventd/plugins/raid/Makefile.in

@@ -0,0 +1,36 @@
+#
+# Copyright (C) 2011 Red Hat, Inc. All rights reserved.
+#
+# This file is part of LVM2.
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions
+# of the GNU General Public License v.2.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+srcdir = @srcdir@
+top_srcdir = @top_srcdir@
+top_builddir = @top_builddir@
+
+INCLUDES += -I$(top_srcdir)/tools -I$(top_srcdir)/daemons/dmeventd/plugins/lvm2
+CLDFLAGS += -L$(top_builddir)/tools -L$(top_builddir)/daemons/dmeventd/plugins/lvm2
+
+SOURCES = dmeventd_raid.c
+
+LIB_NAME = libdevmapper-event-lvm2raid
+LIB_SHARED = $(LIB_NAME).$(LIB_SUFFIX)
+LIB_VERSION = $(LIB_VERSION_LVM)
+
+CFLOW_LIST = $(SOURCES)
+CFLOW_LIST_TARGET = $(LIB_NAME).cflow
+
+include $(top_builddir)/make.tmpl
+
+LIBS += -ldevmapper-event-lvm2 -ldevmapper
+
+install_lvm2: install_dm_plugin
+
+install: install_lvm2

daemons/dmeventd/plugins/raid/dmeventd_raid.c

@@ -0,0 +1,172 @@
+/*
+ * Copyright (C) 2005-2011 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include "lib.h"
+
+#include "lvm2cmd.h"
+#include "errors.h"
+#include "libdevmapper-event.h"
+#include "dmeventd_lvm.h"
+
+#include <syslog.h> /* FIXME Replace syslog with multilog */
+/* FIXME Missing openlog? */
+/* FIXME Replace most syslogs with log_error() style messages and add complete context. */
+/* FIXME Reformat to 80 char lines. */
+
+/*
+ * run_repair is a close copy to
+ * plugins/mirror/dmeventd_mirror.c:_remove_failed_devices()
+ */
+static int run_repair(const char *device)
+{
+	int r;
+#define CMD_SIZE 256	/* FIXME Use system restriction */
+	char cmd_str[CMD_SIZE];
+
+	if (!dmeventd_lvm2_command(dmeventd_lvm2_pool(), cmd_str, sizeof(cmd_str),
+				  "lvconvert --config devices{ignore_suspended_devices=1} "
+				  "--repair --use-policies", device))
+		return -1;
+
+	r = dmeventd_lvm2_run(cmd_str);
+
+	if (r != ECMD_PROCESSED)
+		syslog(LOG_INFO, "Repair of RAID device %s failed.", device);
+
+	return (r == ECMD_PROCESSED) ? 0 : -1;
+}
+
+static int _process_raid_event(char *params, const char *device)
+{
+	int i, n, failure = 0;
+	char *p, *a[4];
+	char *raid_type;
+	char *num_devices;
+	char *health_chars;
+	char *resync_ratio;
+
+	/*
+	 * RAID parms:     <raid_type> <#raid_disks> \
+	 *                 <health chars> <resync ratio>
+	 */
+	if (!dm_split_words(params, 4, 0, a)) {
+		syslog(LOG_ERR, "Failed to process status line for %s\n",
+		       device);
+		return -EINVAL;
+	}
+	raid_type = a[0];
+	num_devices = a[1];
+	health_chars = a[2];
+	resync_ratio = a[3];
+
+	if (!(n = atoi(num_devices))) {
+		syslog(LOG_ERR, "Failed to parse number of devices for %s: %s",
+		       device, num_devices);
+		return -EINVAL;
+	}
+
+	for (i = 0; i < n; i++) {
+		switch (health_chars[i]) {
+		case 'A':
+			/* Device is 'A'live and well */
+		case 'a':
+			/* Device is 'a'live, but not yet in-sync */
+			break;
+		case 'D':
+			syslog(LOG_ERR,
+			       "Device #%d of %s array, %s, has failed.",
+			       i, raid_type, device);
+			failure++;
+			break;
+		default:
+			/* Unhandled character returned from kernel */
+			break;
+		}
+		if (failure)
+			return run_repair(device);
+	}
+
+	p = strstr(resync_ratio, "/");
+	if (!p) {
+		syslog(LOG_ERR, "Failed to parse resync_ratio for %s: %s",
+		       device, resync_ratio);
+		return -EINVAL;
+	}
+	p[0] = '\0';
+	syslog(LOG_INFO, "%s array, %s, is %s in-sync.",
+	       raid_type, device, strcmp(resync_ratio, p+1) ? "not" : "now");
+
+	return 0;
+}
+
+void process_event(struct dm_task *dmt,
+		   enum dm_event_mask event __attribute__((unused)),
+		   void **unused __attribute__((unused)))
+{
+	void *next = NULL;
+	uint64_t start, length;
+	char *target_type = NULL;
+	char *params;
+	const char *device = dm_task_get_name(dmt);
+
+	dmeventd_lvm2_lock();
+
+	do {
+		next = dm_get_next_target(dmt, next, &start, &length,
+					  &target_type, &params);
+
+		if (!target_type) {
+			syslog(LOG_INFO, "%s mapping lost.", device);
+			continue;
+		}
+
+		if (strcmp(target_type, "raid")) {
+			syslog(LOG_INFO, "%s has non-raid portion.", device);
+			continue;
+		}
+
+		if (_process_raid_event(params, device))
+			syslog(LOG_ERR, "Failed to process event for %s",
+			       device);
+	} while (next);
+
+	dmeventd_lvm2_unlock();
+}
+
+int register_device(const char *device,
+		    const char *uuid __attribute__((unused)),
+		    int major __attribute__((unused)),
+		    int minor __attribute__((unused)),
+		    void **unused __attribute__((unused)))
+{
+	if (!dmeventd_lvm2_init())
+		return 0;
+
+	syslog(LOG_INFO, "Monitoring RAID device %s for events.", device);
+
+	return 1;
+}
+
+int unregister_device(const char *device,
+		      const char *uuid __attribute__((unused)),
+		      int major __attribute__((unused)),
+		      int minor __attribute__((unused)),
+		      void **unused __attribute__((unused)))
+{
+	syslog(LOG_INFO, "No longer monitoring RAID device %s for events.",
+	       device);
+	dmeventd_lvm2_exit();
+
+	return 1;
+}

daemons/dmeventd/plugins/snapshot/.exported_symbols

@@ -0,0 +1,3 @@
+process_event
+register_device
+unregister_device

daemons/dmeventd/plugins/snapshot/Makefile.in

@@ -0,0 +1,33 @@
+#
+# Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
+# Copyright (C) 2004-2011 Red Hat, Inc. All rights reserved.
+#
+# This file is part of the LVM2.
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions
+# of the GNU General Public License v.2.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+srcdir = @srcdir@
+top_srcdir = @top_srcdir@
+top_builddir = @top_builddir@
+
+INCLUDES += -I$(top_srcdir)/tools -I$(top_srcdir)/daemons/dmeventd/plugins/lvm2
+CLDFLAGS += -L$(top_builddir)/tools -L$(top_builddir)/daemons/dmeventd/plugins/lvm2
+
+SOURCES = dmeventd_snapshot.c
+
+LIB_SHARED = libdevmapper-event-lvm2snapshot.$(LIB_SUFFIX)
+LIB_VERSION = $(LIB_VERSION_LVM)
+
+include $(top_builddir)/make.tmpl
+
+LIBS += -ldevmapper-event-lvm2 -ldevmapper $(DAEMON_LIBS)
+
+install_lvm2: install_dm_plugin
+
+install: install_lvm2

daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c

@@ -0,0 +1,248 @@
+/*
+ * Copyright (C) 2007-2011 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include "lib.h"
+
+#include "lvm2cmd.h"
+#include "errors.h"
+#include "libdevmapper-event.h"
+#include "dmeventd_lvm.h"
+
+#include <sys/wait.h>
+#include <syslog.h> /* FIXME Replace syslog with multilog */
+/* FIXME Missing openlog? */
+
+/* First warning when snapshot is 80% full. */
+#define WARNING_THRESH 80
+/* Run a check every 5%. */
+#define CHECK_STEP 5
+/* Do not bother checking snapshots less than 50% full. */
+#define CHECK_MINIMUM 50
+
+#define UMOUNT_COMMAND "/bin/umount"
+
+struct dso_state {
+	struct dm_pool *mem;
+	int percent_check;
+	uint64_t known_size;
+	char cmd_str[1024];
+};
+
+static int _run(const char *cmd, ...)
+{
+        va_list ap;
+        int argc = 1; /* for argv[0], i.e. cmd */
+        int i = 0;
+        const char **argv;
+        pid_t pid = fork();
+        int status;
+
+        if (pid == 0) { /* child */
+                va_start(ap, cmd);
+                while (va_arg(ap, const char *))
+                        ++ argc;
+                va_end(ap);
+
+                /* + 1 for the terminating NULL */
+                argv = alloca(sizeof(const char *) * (argc + 1));
+
+                argv[0] = cmd;
+                va_start(ap, cmd);
+                while ((argv[++i] = va_arg(ap, const char *)));
+                va_end(ap);
+
+                execvp(cmd, (char **)argv);
+                syslog(LOG_ERR, "Failed to execute %s: %s.\n", cmd, strerror(errno));
+                exit(127);
+        }
+
+        if (pid > 0) { /* parent */
+                if (waitpid(pid, &status, 0) != pid)
+                        return 0; /* waitpid failed */
+                if (!WIFEXITED(status) || WEXITSTATUS(status))
+                        return 0; /* the child failed */
+        }
+
+        if (pid < 0)
+                return 0; /* fork failed */
+
+        return 1; /* all good */
+}
+
+static int _extend(const char *cmd)
+{
+	return dmeventd_lvm2_run(cmd) == ECMD_PROCESSED;
+}
+
+static void _umount(const char *device, int major, int minor)
+{
+	FILE *mounts;
+	char buffer[4096];
+	char *words[3];
+	struct stat st;
+
+	if (!(mounts = fopen("/proc/mounts", "r"))) {
+		syslog(LOG_ERR, "Could not read /proc/mounts. Not umounting %s.\n", device);
+		return;
+	}
+
+	while (!feof(mounts)) {
+		/* read a line of /proc/mounts */
+		if (!fgets(buffer, sizeof(buffer), mounts))
+			break; /* eof, likely */
+
+		/* words[0] is the mount point and words[1] is the device path */
+		if (dm_split_words(buffer, 3, 0, words) < 2)
+			continue;
+
+		/* find the major/minor of the device */
+		if (stat(words[0], &st))
+			continue; /* can't stat, skip this one */
+
+		if (S_ISBLK(st.st_mode) &&
+		    major(st.st_rdev) == major &&
+		    minor(st.st_rdev) == minor) {
+			syslog(LOG_ERR, "Unmounting invalid snapshot %s from %s.\n", device, words[1]);
+                        if (!_run(UMOUNT_COMMAND, "-fl", words[1], NULL))
+                                syslog(LOG_ERR, "Failed to umount snapshot %s from %s: %s.\n",
+                                       device, words[1], strerror(errno));
+		}
+	}
+
+	if (fclose(mounts))
+		syslog(LOG_ERR, "Failed to close /proc/mounts.\n");
+}
+
+void process_event(struct dm_task *dmt,
+		   enum dm_event_mask event __attribute__((unused)),
+		   void **private)
+{
+	void *next = NULL;
+	uint64_t start, length;
+	char *target_type = NULL;
+	char *params;
+	struct dm_status_snapshot *status = NULL;
+	const char *device = dm_task_get_name(dmt);
+	int percent;
+	struct dso_state *state = *private;
+
+	/* No longer monitoring, waiting for remove */
+	if (!state->percent_check)
+		return;
+
+	dmeventd_lvm2_lock();
+
+	dm_get_next_target(dmt, next, &start, &length, &target_type, &params);
+	if (!target_type)
+		goto out;
+
+	if (!dm_get_status_snapshot(state->mem, params, &status))
+		goto out;
+
+	if (status->invalid) {
+		struct dm_info info;
+		if (dm_task_get_info(dmt, &info)) {
+			dmeventd_lvm2_unlock();
+			_umount(device, info.major, info.minor);
+			return;
+		} /* else; too bad, but this is best-effort thing... */
+	}
+
+	/* Snapshot size had changed. Clear the threshold. */
+	if (state->known_size != status->total_sectors) {
+		state->percent_check = CHECK_MINIMUM;
+		state->known_size = status->total_sectors;
+	}
+
+	/*
+	 * If the snapshot has been invalidated or we failed to parse
+	 * the status string. Report the full status string to syslog.
+	 */
+	if (status->invalid || !status->total_sectors) {
+		syslog(LOG_ERR, "Snapshot %s changed state to: %s\n", device, params);
+		state->percent_check = 0;
+		goto out;
+	}
+
+	percent = (int) (100 * status->used_sectors / status->total_sectors);
+	if (percent >= state->percent_check) {
+		/* Usage has raised more than CHECK_STEP since the last
+		   time. Run actions. */
+		state->percent_check = (percent / CHECK_STEP) * CHECK_STEP + CHECK_STEP;
+
+		if (percent >= WARNING_THRESH) /* Print a warning to syslog. */
+			syslog(LOG_WARNING, "Snapshot %s is now %i%% full.\n", device, percent);
+		/* Try to extend the snapshot, in accord with user-set policies */
+		if (!_extend(state->cmd_str))
+			syslog(LOG_ERR, "Failed to extend snapshot %s.\n", device);
+	}
+
+out:
+	if (status)
+		dm_pool_free(state->mem, status);
+	dmeventd_lvm2_unlock();
+}
+
+int register_device(const char *device,
+		    const char *uuid __attribute__((unused)),
+		    int major __attribute__((unused)),
+		    int minor __attribute__((unused)),
+		    void **private)
+{
+	struct dm_pool *statemem = NULL;
+	struct dso_state *state;
+
+	if (!dmeventd_lvm2_init())
+		goto out;
+
+	if (!(statemem = dm_pool_create("snapshot_state", 512)) ||
+	    !(state = dm_pool_zalloc(statemem, sizeof(*state))))
+		goto bad;
+
+	if (!dmeventd_lvm2_command(statemem, state->cmd_str,
+				   sizeof(state->cmd_str),
+				   "lvextend --use-policies", device))
+		goto bad;
+
+	state->mem = statemem;
+	state->percent_check = CHECK_MINIMUM;
+	*private = state;
+
+	syslog(LOG_INFO, "Monitoring snapshot %s\n", device);
+
+	return 1;
+bad:
+	if (statemem)
+		dm_pool_destroy(statemem);
+	dmeventd_lvm2_exit();
+out:
+	syslog(LOG_ERR, "Failed to monitor snapshot %s.\n", device);
+
+	return 0;
+}
+
+int unregister_device(const char *device,
+		      const char *uuid __attribute__((unused)),
+		      int major __attribute__((unused)),
+		      int minor __attribute__((unused)),
+		      void **private)
+{
+	struct dso_state *state = *private;
+
+	syslog(LOG_INFO, "No longer monitoring snapshot %s\n", device);
+	dm_pool_destroy(state->mem);
+	dmeventd_lvm2_exit();
+
+	return 1;
+}

daemons/dmeventd/plugins/thin/.exported_symbols

@@ -0,0 +1,3 @@
+process_event
+register_device
+unregister_device

daemons/dmeventd/plugins/thin/Makefile.in

@@ -0,0 +1,36 @@
+#
+# Copyright (C) 2011 Red Hat, Inc. All rights reserved.
+#
+# This file is part of LVM2.
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions
+# of the GNU General Public License v.2.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+srcdir = @srcdir@
+top_srcdir = @top_srcdir@
+top_builddir = @top_builddir@
+
+INCLUDES += -I$(top_srcdir)/tools -I$(top_srcdir)/daemons/dmeventd/plugins/lvm2
+CLDFLAGS += -L$(top_builddir)/tools -L$(top_builddir)/daemons/dmeventd/plugins/lvm2
+
+SOURCES = dmeventd_thin.c
+
+LIB_NAME = libdevmapper-event-lvm2thin
+LIB_SHARED = $(LIB_NAME).$(LIB_SUFFIX)
+LIB_VERSION = $(LIB_VERSION_LVM)
+
+CFLOW_LIST = $(SOURCES)
+CFLOW_LIST_TARGET = $(LIB_NAME).cflow
+
+include $(top_builddir)/make.tmpl
+
+LIBS += -ldevmapper-event-lvm2 -ldevmapper
+
+install_lvm2: install_dm_plugin
+
+install: install_lvm2

daemons/dmeventd/plugins/thin/dmeventd_thin.c

@@ -0,0 +1,401 @@
+/*
+ * Copyright (C) 2011-2013 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include "lib.h"
+
+#include "lvm2cmd.h"
+#include "errors.h"
+#include "libdevmapper-event.h"
+#include "dmeventd_lvm.h"
+
+#include <sys/wait.h>
+#include <syslog.h> /* FIXME Replace syslog with multilog */
+/* FIXME Missing openlog? */
+
+/* First warning when thin is 80% full. */
+#define WARNING_THRESH 80
+/* Run a check every 5%. */
+#define CHECK_STEP 5
+/* Do not bother checking thins less than 50% full. */
+#define CHECK_MINIMUM 50
+
+#define UMOUNT_COMMAND "/bin/umount"
+
+#define THIN_DEBUG 0
+
+struct dso_state {
+	struct dm_pool *mem;
+	int metadata_percent_check;
+	int data_percent_check;
+	uint64_t known_metadata_size;
+	uint64_t known_data_size;
+	char cmd_str[1024];
+};
+
+
+/* TODO - move this mountinfo code into library to be reusable */
+#ifdef linux
+#  include "kdev_t.h"
+#else
+#  define MAJOR(x) major((x))
+#  define MINOR(x) minor((x))
+#  define MKDEV(x,y) makedev((x),(y))
+#endif
+
+/* Get dependencies for device, and try to find matching device */
+static int _has_deps(const char *name, int tp_major, int tp_minor, int *dev_minor)
+{
+	struct dm_task *dmt;
+	const struct dm_deps *deps;
+	struct dm_info info;
+	int major, minor;
+	int r = 0;
+
+	if (!(dmt = dm_task_create(DM_DEVICE_DEPS)))
+		return 0;
+
+	if (!dm_task_set_name(dmt, name))
+		goto out;
+
+	if (!dm_task_no_open_count(dmt))
+		goto out;
+
+	if (!dm_task_run(dmt))
+		goto out;
+
+	if (!dm_task_get_info(dmt, &info))
+		goto out;
+
+	if (!(deps = dm_task_get_deps(dmt)))
+		goto out;
+
+	if (!info.exists || deps->count != 1)
+		goto out;
+
+	major = (int) MAJOR(deps->device[0]);
+	minor = (int) MINOR(deps->device[0]);
+	if ((major != tp_major) || (minor != tp_minor))
+		goto out;
+
+	*dev_minor = info.minor;
+
+#if THIN_DEBUG
+	{
+		char dev_name[PATH_MAX];
+		if (dm_device_get_name(major, minor, 0, dev_name, sizeof(dev_name)))
+			syslog(LOG_DEBUG, "Found %s (%u:%u) depends on %s",
+			       name, major, *dev_minor, dev_name);
+	}
+#endif
+	r = 1;
+out:
+	dm_task_destroy(dmt);
+
+	return r;
+}
+
+/* Get all active devices */
+static int _find_all_devs(dm_bitset_t bs, int tp_major, int tp_minor)
+{
+	struct dm_task *dmt;
+	struct dm_names *names;
+	unsigned next = 0;
+	int minor, r = 1;
+
+	if (!(dmt = dm_task_create(DM_DEVICE_LIST)))
+		return 0;
+
+	if (!dm_task_run(dmt)) {
+		r = 0;
+		goto out;
+	}
+
+	if (!(names = dm_task_get_names(dmt))) {
+		r = 0;
+		goto out;
+	}
+
+	if (!names->dev)
+		goto out;
+
+	do {
+		names = (struct dm_names *)((char *) names + next);
+		if (_has_deps(names->name, tp_major, tp_minor, &minor))
+			dm_bit_set(bs, minor);
+		next = names->next;
+	} while (next);
+
+out:
+	dm_task_destroy(dmt);
+
+	return r;
+}
+
+static int _extend(struct dso_state *state)
+{
+#if THIN_DEBUG
+	syslog(LOG_INFO, "dmeventd executes: %s.\n", state->cmd_str);
+#endif
+	return (dmeventd_lvm2_run(state->cmd_str) == ECMD_PROCESSED);
+}
+
+static int _run(const char *cmd, ...)
+{
+	va_list ap;
+	int argc = 1; /* for argv[0], i.e. cmd */
+	int i = 0;
+	const char **argv;
+	pid_t pid = fork();
+	int status;
+
+	if (pid == 0) { /* child */
+		va_start(ap, cmd);
+		while (va_arg(ap, const char *))
+			++argc;
+		va_end(ap);
+
+		/* + 1 for the terminating NULL */
+		argv = alloca(sizeof(const char *) * (argc + 1));
+
+		argv[0] = cmd;
+                va_start(ap, cmd);
+		while ((argv[++i] = va_arg(ap, const char *)));
+		va_end(ap);
+
+		execvp(cmd, (char **)argv);
+		syslog(LOG_ERR, "Failed to execute %s: %s.\n", cmd, strerror(errno));
+		exit(127);
+	}
+
+	if (pid > 0) { /* parent */
+		if (waitpid(pid, &status, 0) != pid)
+			return 0; /* waitpid failed */
+		if (!WIFEXITED(status) || WEXITSTATUS(status))
+			return 0; /* the child failed */
+	}
+
+	if (pid < 0)
+		return 0; /* fork failed */
+
+	return 1; /* all good */
+}
+
+struct mountinfo_s {
+	struct dm_info info;
+	dm_bitset_t minors; /* Bitset for active thin pool minors */
+	const char *device;
+};
+
+static int _umount_device(char *buffer, unsigned major, unsigned minor,
+			  char *target, void *cb_data)
+{
+	struct mountinfo_s *data = cb_data;
+
+	if ((major == data->info.major) && dm_bit(data->minors, minor)) {
+		syslog(LOG_INFO, "Unmounting thin volume %s from %s.\n",
+		       data->device, target);
+		if (!_run(UMOUNT_COMMAND, "-fl", target, NULL))
+			syslog(LOG_ERR, "Failed to umount thin %s from %s: %s.\n",
+			       data->device, target, strerror(errno));
+	}
+
+	return 1;
+}
+
+/*
+ * Find all thin pool users and try to umount them.
+ * TODO: work with read-only thin pool support
+ */
+static void _umount(struct dm_task *dmt, const char *device)
+{
+	static const size_t MINORS = 4096;
+	struct mountinfo_s data = {
+		.device = device,
+	};
+
+	if (!dm_task_get_info(dmt, &data.info))
+		return;
+
+	dmeventd_lvm2_unlock();
+
+	if (!(data.minors = dm_bitset_create(NULL, MINORS))) {
+		syslog(LOG_ERR, "Failed to allocate bitset. Not unmounting %s.\n", device);
+		goto out;
+	}
+
+	if (!_find_all_devs(data.minors, data.info.major, data.info.minor)) {
+		syslog(LOG_ERR, "Failed to detect mounted volumes for %s.\n", device);
+		goto out;
+	}
+
+	if (!dm_mountinfo_read(_umount_device, &data)) {
+		syslog(LOG_ERR, "Could not parse mountinfo file.\n");
+		goto out;
+	}
+
+out:
+	if (data.minors)
+		dm_bitset_destroy(data.minors);
+	dmeventd_lvm2_lock();
+}
+
+void process_event(struct dm_task *dmt,
+		   enum dm_event_mask event __attribute__((unused)),
+		   void **private)
+{
+	const char *device = dm_task_get_name(dmt);
+	int percent;
+	struct dso_state *state = *private;
+	struct dm_status_thin_pool *tps = NULL;
+	void *next = NULL;
+	uint64_t start, length;
+	char *target_type = NULL;
+	char *params;
+
+#if 0
+	/* No longer monitoring, waiting for remove */
+	if (!state->meta_percent_check && !state->data_percent_check)
+		return;
+#endif
+	dmeventd_lvm2_lock();
+
+	dm_get_next_target(dmt, next, &start, &length, &target_type, &params);
+
+	if (!target_type || (strcmp(target_type, "thin-pool") != 0)) {
+		syslog(LOG_ERR, "Invalid target type.\n");
+		goto out;
+	}
+
+	if (!dm_get_status_thin_pool(state->mem, params, &tps)) {
+		syslog(LOG_ERR, "Failed to parse status.\n");
+		_umount(dmt, device);
+		goto out;
+	}
+
+#if THIN_DEBUG
+	syslog(LOG_INFO, "%p: Got status %" PRIu64 " / %" PRIu64
+	       " %" PRIu64  " / %" PRIu64 ".\n", state,
+	       tps->used_metadata_blocks, tps->total_metadata_blocks,
+	       tps->used_data_blocks, tps->total_data_blocks);
+#endif
+
+	/* Thin pool size had changed. Clear the threshold. */
+	if (state->known_metadata_size != tps->total_metadata_blocks) {
+		state->metadata_percent_check = CHECK_MINIMUM;
+		state->known_metadata_size = tps->total_metadata_blocks;
+	}
+
+	if (state->known_data_size != tps->total_data_blocks) {
+		state->data_percent_check = CHECK_MINIMUM;
+		state->known_data_size = tps->total_data_blocks;
+	}
+
+	percent = 100 * tps->used_metadata_blocks / tps->total_metadata_blocks;
+	if (percent >= state->metadata_percent_check) {
+		/*
+		 * Usage has raised more than CHECK_STEP since the last
+		 * time. Run actions.
+		 */
+		state->metadata_percent_check = (percent / CHECK_STEP) * CHECK_STEP + CHECK_STEP;
+
+		/* FIXME: extension of metadata needs to be written! */
+		if (percent >= WARNING_THRESH) /* Print a warning to syslog. */
+			syslog(LOG_WARNING, "Thin metadata %s is now %i%% full.\n",
+			       device, percent);
+		 /* Try to extend the metadata, in accord with user-set policies */
+		if (!_extend(state)) {
+			syslog(LOG_ERR, "Failed to extend thin metadata %s.\n",
+			       device);
+			_umount(dmt, device);
+		}
+		/* FIXME: hmm READ-ONLY switch should happen in error path */
+	}
+
+	percent = 100 * tps->used_data_blocks / tps->total_data_blocks;
+	if (percent >= state->data_percent_check) {
+		/*
+		 * Usage has raised more than CHECK_STEP since
+		 * the last time. Run actions.
+		 */
+		state->data_percent_check = (percent / CHECK_STEP) * CHECK_STEP + CHECK_STEP;
+
+		if (percent >= WARNING_THRESH) /* Print a warning to syslog. */
+			syslog(LOG_WARNING, "Thin %s is now %i%% full.\n", device, percent);
+		/* Try to extend the thin data, in accord with user-set policies */
+		if (!_extend(state)) {
+			syslog(LOG_ERR, "Failed to extend thin %s.\n", device);
+			state->data_percent_check = 0;
+			_umount(dmt, device);
+		}
+		/* FIXME: hmm READ-ONLY switch should happen in error path */
+	}
+out:
+	if (tps)
+		dm_pool_free(state->mem, tps);
+
+	dmeventd_lvm2_unlock();
+}
+
+int register_device(const char *device,
+		    const char *uuid __attribute__((unused)),
+		    int major __attribute__((unused)),
+		    int minor __attribute__((unused)),
+		    void **private)
+{
+	struct dm_pool *statemem = NULL;
+	struct dso_state *state;
+
+	if (!dmeventd_lvm2_init())
+		goto bad;
+
+	if (!(statemem = dm_pool_create("thin_pool_state", 2048)) ||
+	    !(state = dm_pool_zalloc(statemem, sizeof(*state))) ||
+	    !dmeventd_lvm2_command(statemem, state->cmd_str,
+				   sizeof(state->cmd_str),
+				   "lvextend --use-policies",
+				   device)) {
+		if (statemem)
+			dm_pool_destroy(statemem);
+		dmeventd_lvm2_exit();
+		goto bad;
+	}
+
+	state->mem = statemem;
+	state->metadata_percent_check = CHECK_MINIMUM;
+	state->data_percent_check = CHECK_MINIMUM;
+	*private = state;
+
+	syslog(LOG_INFO, "Monitoring thin %s.\n", device);
+
+	return 1;
+bad:
+	syslog(LOG_ERR, "Failed to monitor thin %s.\n", device);
+
+	return 0;
+}
+
+int unregister_device(const char *device,
+		      const char *uuid __attribute__((unused)),
+		      int major __attribute__((unused)),
+		      int minor __attribute__((unused)),
+		      void **private)
+{
+	struct dso_state *state = *private;
+
+	syslog(LOG_INFO, "No longer monitoring thin %s.\n", device);
+	dm_pool_destroy(state->mem);
+	dmeventd_lvm2_exit();
+
+	return 1;
+}

daemons/lvmetad/Makefile.in

@@ -0,0 +1,59 @@
+#
+# Copyright (C) 2011-2012 Red Hat, Inc.
+#
+# This file is part of LVM2.
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions
+# of the GNU Lesser General Public License v.2.1.
+#
+# You should have received a copy of the GNU Lesser General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+srcdir = @srcdir@
+top_srcdir = @top_srcdir@
+top_builddir = @top_builddir@
+
+SOURCES = lvmetad-core.c
+SOURCES2 = testclient.c
+
+TARGETS = lvmetad lvmetad-testclient
+
+.PHONY: install_lvmetad
+
+CFLOW_LIST = $(SOURCES)
+CFLOW_LIST_TARGET = $(LIB_NAME).cflow
+CFLOW_TARGET = lvmetad
+
+include $(top_builddir)/make.tmpl
+
+INCLUDES += -I$(top_srcdir)/libdaemon/server
+LVMLIBS = -ldaemonserver $(LVMINTERNAL_LIBS) -ldevmapper
+
+LIBS += $(PTHREAD_LIBS)
+
+LDFLAGS += -L$(top_builddir)/libdaemon/server
+CLDFLAGS += -L$(top_builddir)/libdaemon/server
+
+lvmetad: $(OBJECTS) $(top_builddir)/libdaemon/client/libdaemonclient.a \
+		    $(top_builddir)/libdaemon/server/libdaemonserver.a
+	$(CC) $(CFLAGS) $(LDFLAGS) -o $@ $(OBJECTS) \
+	$(DL_LIBS) $(LVMLIBS) $(LIBS) -rdynamic
+
+# TODO: No idea. No idea how to test either.
+#ifneq ("$(CFLOW_CMD)", "")
+#CFLOW_SOURCES = $(addprefix $(srcdir)/, $(SOURCES))
+#-include $(top_builddir)/libdm/libdevmapper.cflow
+#-include $(top_builddir)/lib/liblvm-internal.cflow
+#-include $(top_builddir)/lib/liblvm2cmd.cflow
+#-include $(top_builddir)/daemons/dmeventd/$(LIB_NAME).cflow
+#-include $(top_builddir)/daemons/dmeventd/plugins/mirror/$(LIB_NAME)-lvm2mirror.cflow
+#endif
+
+install_lvmetad: lvmetad
+	$(INSTALL_PROGRAM) -D $< $(sbindir)/$(<F)
+
+install_lvm2: install_lvmetad
+
+install: install_lvm2

daemons/lvmetad/lvmetad-client.h

@@ -0,0 +1,81 @@
+/*
+ * Copyright (C) 2011-2012 Red Hat, Inc.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#ifndef _LVM_LVMETAD_CLIENT_H
+#define _LVM_LVMETAD_CLIENT_H
+
+#include "daemon-client.h"
+
+struct volume_group;
+
+/* Different types of replies we may get from lvmetad. */
+
+typedef struct {
+	daemon_reply r;
+	const char **uuids; /* NULL terminated array */
+} lvmetad_uuidlist;
+
+typedef struct {
+	daemon_reply r;
+	struct dm_config_tree *cft;
+} lvmetad_vg;
+
+/* Get a list of VG UUIDs that match a given VG name. */
+lvmetad_uuidlist lvmetad_lookup_vgname(daemon_handle h, const char *name);
+
+/* Get the metadata of a single VG, identified by UUID. */
+lvmetad_vg lvmetad_get_vg(daemon_handle h, const char *uuid);
+
+/*
+ * Add and remove PVs on demand. Udev-driven systems will use this interface
+ * instead of scanning.
+ */
+daemon_reply lvmetad_add_pv(daemon_handle h, const char *pv_uuid, const char *mda_content);
+daemon_reply lvmetad_remove_pv(daemon_handle h, const char *pv_uuid);
+
+/* Trigger a full disk scan, throwing away all caches. XXX do we eventually want
+ * this? Probably not yet, anyway.
+ *     daemon_reply lvmetad_rescan(daemon_handle h);
+ */
+
+/*
+ * Update the version of metadata of a volume group. The VG has to be locked for
+ * writing for this, and the VG metadata here has to match whatever has been
+ * written to the disk (under this lock). This initially avoids the requirement
+ * for lvmetad to write to disk (in later revisions, lvmetad_supersede_vg may
+ * also do the writing, or we probably add another function to do that).
+ */
+daemon_reply lvmetad_supersede_vg(daemon_handle h, struct volume_group *vg);
+
+/* Wrappers to open/close connection */
+
+static inline daemon_handle lvmetad_open(const char *socket)
+{
+	daemon_info lvmetad_info = {
+		.path = "lvmetad",
+		.socket = socket ?: DEFAULT_RUN_DIR "/lvmetad.socket",
+		.protocol = "lvmetad",
+		.protocol_version = 1,
+		.autostart = 0
+	};
+
+	return daemon_open(lvmetad_info);
+}
+
+static inline void lvmetad_close(daemon_handle h)
+{
+	return daemon_close(h);
+}
+
+#endif

daemons/lvmetad/test.sh

@@ -0,0 +1,16 @@
+#!/bin/bash
+
+export LD_LIBRARY_PATH="$1"
+
+test -n "$2" && {
+    rm -f /var/run/lvmetad.{socket,pid}
+    chmod +rx lvmetad
+    valgrind ./lvmetad -f &
+    PID=$!
+    sleep 1
+    ./testclient
+    kill $PID
+    exit 0
+}
+
+sudo ./test.sh "$1" .

daemons/lvmetad/testclient.c

@@ -0,0 +1,127 @@
+#include "lvmetad-client.h"
+#include "label.h"
+#include "lvmcache.h"
+#include "metadata.h"
+
+const char *uuid1 = "abcd-efgh";
+const char *uuid2 = "bbcd-efgh";
+const char *vgid = "yada-yada";
+const char *uuid3 = "cbcd-efgh";
+
+const char *metadata2 = "{\n"
+	"id = \"yada-yada\"\n"
+	"seqno = 15\n"
+	"status = [\"READ\", \"WRITE\"]\n"
+	"flags = []\n"
+	"extent_size = 8192\n"
+	"physical_volumes {\n"
+	"    pv0 {\n"
+	"        id = \"abcd-efgh\"\n"
+	"    }\n"
+	"    pv1 {\n"
+	"        id = \"bbcd-efgh\"\n"
+	"    }\n"
+	"    pv2 {\n"
+	"        id = \"cbcd-efgh\"\n"
+	"    }\n"
+	"}\n"
+	"}\n";
+
+void _handle_reply(daemon_reply reply) {
+	const char *repl = daemon_reply_str(reply, "response", NULL);
+	const char *status = daemon_reply_str(reply, "status", NULL);
+	const char *vgid = daemon_reply_str(reply, "vgid", NULL);
+
+	fprintf(stderr, "[C] REPLY: %s\n", repl);
+	if (!strcmp(repl, "failed"))
+		fprintf(stderr, "[C] REASON: %s\n", daemon_reply_str(reply, "reason", "unknown"));
+	if (vgid)
+		fprintf(stderr, "[C] VGID: %s\n", vgid);
+	if (status)
+		fprintf(stderr, "[C] STATUS: %s\n", status);
+	daemon_reply_destroy(reply);
+}
+
+void _pv_add(daemon_handle h, const char *uuid, const char *metadata)
+{
+	daemon_reply reply = daemon_send_simple(h, "pv_add", "uuid = %s", uuid,
+						             "metadata = %b", metadata,
+						             NULL);
+	_handle_reply(reply);
+}
+
+int scan(daemon_handle h, char *fn) {
+	struct device *dev = dev_cache_get(fn, NULL);
+
+	struct label *label;
+	if (!label_read(dev, &label, 0)) {
+		fprintf(stderr, "[C] no label found on %s\n", fn);
+		return;
+	}
+
+	char uuid[64];
+	id_write_format(dev->pvid, uuid, 64);
+	fprintf(stderr, "[C] found PV: %s\n", uuid);
+	struct lvmcache_info *info = (struct lvmcache_info *) label->info;
+	struct physical_volume pv = { 0, };
+
+	if (!(info->fmt->ops->pv_read(info->fmt, dev_name(dev), &pv, 0))) {
+		fprintf(stderr, "[C] Failed to read PV %s", dev_name(dev));
+		return;
+	}
+
+	struct format_instance_ctx fic;
+	struct format_instance *fid = info->fmt->ops->create_instance(info->fmt, &fic);
+	struct metadata_area *mda;
+	struct volume_group *vg = NULL;
+	dm_list_iterate_items(mda, &info->mdas) {
+		struct volume_group *this = mda->ops->vg_read(fid, "", mda);
+		if (this && !vg || this->seqno > vg->seqno)
+			vg = this;
+	}
+	if (vg) {
+		char *buf = NULL;
+		/* TODO. This is not entirely correct, since export_vg_to_buffer
+		 * adds trailing garbage to the buffer. We may need to use
+		 * export_vg_to_config_tree and format the buffer ourselves. It
+		 * does, however, work for now, since the garbage is well
+		 * formatted and has no conflicting keys with the rest of the
+		 * request.  */
+		export_vg_to_buffer(vg, &buf);
+		daemon_reply reply =
+			daemon_send_simple(h, "pv_add", "uuid = %s", uuid,
+					      "metadata = %b", strchr(buf, '{'),
+					      NULL);
+		_handle_reply(reply);
+	}
+}
+
+void _dump_vg(daemon_handle h, const char *uuid)
+{
+	daemon_reply reply = daemon_send_simple(h, "vg_by_uuid", "uuid = %s", uuid, NULL);
+	fprintf(stderr, "[C] reply buffer: %s\n", reply.buffer);
+	daemon_reply_destroy(reply);
+}
+
+int main(int argc, char **argv) {
+	daemon_handle h = lvmetad_open();
+
+	if (argc > 1) {
+		int i;
+		struct cmd_context *cmd = create_toolcontext(0, NULL, 0, 0);
+		for (i = 1; i < argc; ++i) {
+			const char *uuid = NULL;
+			scan(h, argv[i]);
+		}
+		destroy_toolcontext(cmd);
+		return 0;
+	}
+
+	_pv_add(h, uuid1, NULL);
+	_pv_add(h, uuid2, metadata2);
+	_dump_vg(h, vgid);
+	_pv_add(h, uuid3, NULL);
+
+	daemon_close(h);
+	return 0;
+}

doc/.gitignore

@@ -0,0 +1 @@
+example.conf

doc/Makefile.in

@@ -0,0 +1,31 @@
+#
+# Copyright (C) 2004-2010 Red Hat, Inc. All rights reserved.
+#
+# This file is part of LVM2.
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions
+# of the GNU General Public License v.2.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+
+srcdir = @srcdir@
+top_srcdir = @top_srcdir@
+top_builddir = @top_builddir@
+
+CONFSRC=example.conf
+CONFDEST=lvm.conf
+
+include $(top_builddir)/make.tmpl
+
+install_lvm2: $(CONFSRC)
+	@if [ ! -e $(confdir)/$(CONFDEST) ]; then \
+		echo "$(INSTALL_WDATA) -D $< $(confdir)/$(CONFDEST)"; \
+		$(INSTALL_WDATA) -D $< $(confdir)/$(CONFDEST); \
+	fi
+
+install: install_lvm2
+
+DISTCLEAN_TARGETS += $(CONFSRC)

doc/README

@@ -1 +0,0 @@
-Wow!  This is really incredible documentation!

doc/example.conf.in

@@ -0,0 +1,893 @@
+# This is an example configuration file for the LVM2 system.
+# It contains the default settings that would be used if there was no
+# @DEFAULT_SYS_DIR@/lvm.conf file.
+#
+# Refer to 'man lvm.conf' for further information including the file layout.
+#
+# To put this file in a different directory and override @DEFAULT_SYS_DIR@ set
+# the environment variable LVM_SYSTEM_DIR before running the tools.
+#
+# N.B. Take care that each setting only appears once if uncommenting
+# example settings in this file.
+
+# This section allows you to set the way the configuration settings are handled.
+config {
+
+    # If enabled, any LVM2 configuration mismatch is reported.
+    # This implies checking that the configuration key is understood
+    # by LVM2 and that the value of the key is of a proper type.
+    # If disabled, any configuration mismatch is ignored and default
+    # value is used instead without any warning (a message about the
+    # configuration key not being found is issued in verbose mode only).
+    checks = 1
+
+    # If enabled, any configuration mismatch aborts the LVM2 process.
+    abort_on_errors = 0
+}
+
+# This section allows you to configure which block devices should
+# be used by the LVM system.
+devices {
+
+    # Where do you want your volume groups to appear ?
+    dir = "/dev"
+
+    # An array of directories that contain the device nodes you wish
+    # to use with LVM2.
+    scan = [ "/dev" ]
+
+    # If set, the cache of block device nodes with all associated symlinks
+    # will be constructed out of the existing udev database content.
+    # This avoids using and opening any inapplicable non-block devices or
+    # subdirectories found in the device directory. This setting is applied
+    # to udev-managed device directory only, other directories will be scanned
+    # fully. LVM2 needs to be compiled with udev support for this setting to
+    # take effect. N.B. Any device node or symlink not managed by udev in
+    # udev directory will be ignored with this setting on.
+    obtain_device_list_from_udev = 1
+
+    # If several entries in the scanned directories correspond to the
+    # same block device and the tools need to display a name for device,
+    # all the pathnames are matched against each item in the following
+    # list of regular expressions in turn and the first match is used.
+    preferred_names = [ ]
+
+    # Try to avoid using undescriptive /dev/dm-N names, if present.
+    # preferred_names = [ "^/dev/mpath/", "^/dev/mapper/mpath", "^/dev/[hs]d" ]
+
+    # A filter that tells LVM2 to only use a restricted set of devices.
+    # The filter consists of an array of regular expressions.  These
+    # expressions can be delimited by a character of your choice, and
+    # prefixed with either an 'a' (for accept) or 'r' (for reject).
+    # The first expression found to match a device name determines if
+    # the device will be accepted or rejected (ignored).  Devices that
+    # don't match any patterns are accepted.
+
+    # Be careful if there there are symbolic links or multiple filesystem 
+    # entries for the same device as each name is checked separately against
+    # the list of patterns.  The effect is that if the first pattern in the 
+    # list to match a name is an 'a' pattern for any of the names, the device
+    # is accepted; otherwise if the first pattern in the list to match a name
+    # is an 'r' pattern for any of the names it is rejected; otherwise it is
+    # accepted.
+
+    # Don't have more than one filter line active at once: only one gets used.
+
+    # Run vgscan after you change this parameter to ensure that
+    # the cache file gets regenerated (see below).
+    # If it doesn't do what you expect, check the output of 'vgscan -vvvv'.
+
+
+    # By default we accept every block device:
+    filter = [ "a/.*/" ]
+
+    # Exclude the cdrom drive
+    # filter = [ "r|/dev/cdrom|" ]
+
+    # When testing I like to work with just loopback devices:
+    # filter = [ "a/loop/", "r/.*/" ]
+
+    # Or maybe all loops and ide drives except hdc:
+    # filter =[ "a|loop|", "r|/dev/hdc|", "a|/dev/ide|", "r|.*|" ]
+
+    # Use anchors if you want to be really specific
+    # filter = [ "a|^/dev/hda8$|", "r/.*/" ]
+
+    # Since "filter" is often overriden from command line, it is not suitable
+    # for system-wide device filtering (udev rules, lvmetad). To hide devices
+    # from LVM-specific udev processing and/or from lvmetad, you need to set
+    # global_filter. The syntax is the same as for normal "filter"
+    # above. Devices that fail the global_filter are not even opened by LVM.
+
+    # global_filter = []
+
+    # The results of the filtering are cached on disk to avoid
+    # rescanning dud devices (which can take a very long time).
+    # By default this cache is stored in the @DEFAULT_SYS_DIR@/@DEFAULT_CACHE_SUBDIR@ directory
+    # in a file called '.cache'.
+    # It is safe to delete the contents: the tools regenerate it.
+    # (The old setting 'cache' is still respected if neither of
+    # these new ones is present.)
+    # N.B. If obtain_device_list_from_udev is set to 1 the list of
+    # devices is instead obtained from udev and any existing .cache
+    # file is removed.
+    cache_dir = "@DEFAULT_SYS_DIR@/@DEFAULT_CACHE_SUBDIR@"
+    cache_file_prefix = ""
+
+    # You can turn off writing this cache file by setting this to 0.
+    write_cache_state = 1
+
+    # Advanced settings.
+
+    # List of pairs of additional acceptable block device types found 
+    # in /proc/devices with maximum (non-zero) number of partitions.
+    # types = [ "fd", 16 ]
+
+    # If sysfs is mounted (2.6 kernels) restrict device scanning to 
+    # the block devices it believes are valid.
+    # 1 enables; 0 disables.
+    sysfs_scan = 1
+
+    # By default, LVM2 will ignore devices used as component paths
+    # of device-mapper multipath devices.
+    # 1 enables; 0 disables.
+    multipath_component_detection = 1
+
+    # By default, LVM2 will ignore devices used as components of
+    # software RAID (md) devices by looking for md superblocks.
+    # 1 enables; 0 disables.
+    md_component_detection = 1
+
+    # By default, if a PV is placed directly upon an md device, LVM2
+    # will align its data blocks with the md device's stripe-width.
+    # 1 enables; 0 disables.
+    md_chunk_alignment = 1
+
+    # Default alignment of the start of a data area in MB.  If set to 0,
+    # a value of 64KB will be used.  Set to 1 for 1MiB, 2 for 2MiB, etc.
+    # default_data_alignment = @DEFAULT_DATA_ALIGNMENT@
+
+    # By default, the start of a PV's data area will be a multiple of
+    # the 'minimum_io_size' or 'optimal_io_size' exposed in sysfs.
+    # - minimum_io_size - the smallest request the device can perform
+    #   w/o incurring a read-modify-write penalty (e.g. MD's chunk size)
+    # - optimal_io_size - the device's preferred unit of receiving I/O
+    #   (e.g. MD's stripe width)
+    # minimum_io_size is used if optimal_io_size is undefined (0).
+    # If md_chunk_alignment is enabled, that detects the optimal_io_size.
+    # This setting takes precedence over md_chunk_alignment.
+    # 1 enables; 0 disables.
+    data_alignment_detection = 1
+
+    # Alignment (in KB) of start of data area when creating a new PV.
+    # md_chunk_alignment and data_alignment_detection are disabled if set.
+    # Set to 0 for the default alignment (see: data_alignment_default)
+    # or page size, if larger.
+    data_alignment = 0
+
+    # By default, the start of the PV's aligned data area will be shifted by
+    # the 'alignment_offset' exposed in sysfs.  This offset is often 0 but
+    # may be non-zero; e.g.: certain 4KB sector drives that compensate for
+    # windows partitioning will have an alignment_offset of 3584 bytes
+    # (sector 7 is the lowest aligned logical block, the 4KB sectors start
+    # at LBA -1, and consequently sector 63 is aligned on a 4KB boundary).
+    # But note that pvcreate --dataalignmentoffset will skip this detection.
+    # 1 enables; 0 disables.
+    data_alignment_offset_detection = 1
+
+    # If, while scanning the system for PVs, LVM2 encounters a device-mapper
+    # device that has its I/O suspended, it waits for it to become accessible.
+    # Set this to 1 to skip such devices.  This should only be needed
+    # in recovery situations.
+    ignore_suspended_devices = 0
+
+    # During each LVM operation errors received from each device are counted.
+    # If the counter of a particular device exceeds the limit set here, no
+    # further I/O is sent to that device for the remainder of the respective
+    # operation. Setting the parameter to 0 disables the counters altogether.
+    disable_after_error_count = 0
+
+    # Allow use of pvcreate --uuid without requiring --restorefile.
+    require_restorefile_with_uuid = 1
+
+    # Minimum size (in KB) of block devices which can be used as PVs.
+    # In a clustered environment all nodes must use the same value.
+    # Any value smaller than 512KB is ignored.
+
+    # Ignore devices smaller than 2MB such as floppy drives.
+    pv_min_size = 2048
+
+    # The original built-in setting was 512 up to and including version 2.02.84.
+    # pv_min_size = 512
+
+    # Issue discards to a logical volumes's underlying physical volume(s) when
+    # the logical volume is no longer using the physical volumes' space (e.g.
+    # lvremove, lvreduce, etc).  Discards inform the storage that a region is
+    # no longer in use.  Storage that supports discards advertise the protocol
+    # specific way discards should be issued by the kernel (TRIM, UNMAP, or
+    # WRITE SAME with UNMAP bit set).  Not all storage will support or benefit
+    # from discards but SSDs and thinly provisioned LUNs generally do.  If set
+    # to 1, discards will only be issued if both the storage and kernel provide
+    # support.
+    # 1 enables; 0 disables.
+    issue_discards = 0
+}
+
+# This section allows you to configure the way in which LVM selects
+# free space for its Logical Volumes.
+allocation {
+
+    # When searching for free space to extend an LV, the "cling"
+    # allocation policy will choose space on the same PVs as the last
+    # segment of the existing LV.  If there is insufficient space and a
+    # list of tags is defined here, it will check whether any of them are
+    # attached to the PVs concerned and then seek to match those PV tags
+    # between existing extents and new extents.
+    # Use the special tag "@*" as a wildcard to match any PV tag.
+ 
+    # Example: LVs are mirrored between two sites within a single VG.
+    # PVs are tagged with either @site1 or @site2 to indicate where
+    # they are situated.
+
+    # cling_tag_list = [ "@site1", "@site2" ]
+    # cling_tag_list = [ "@*" ]
+
+    # Changes made in version 2.02.85 extended the reach of the 'cling'
+    # policies to detect more situations where data can be grouped
+    # onto the same disks.  Set this to 0 to revert to the previous
+    # algorithm.
+    maximise_cling = 1
+
+    # Set to 1 to guarantee that mirror logs will always be placed on 
+    # different PVs from the mirror images.  This was the default
+    # until version 2.02.85.
+    mirror_logs_require_separate_pvs = 0
+
+    # Set to 1 to guarantee that thin pool metadata will always
+    # be placed on different PVs from the pool data.
+    thin_pool_metadata_require_separate_pvs = 0
+
+    # Specify the minimal chunk size (in KB) for thin pool volumes.
+    # Use of the larger chunk size may improve perfomance for plain
+    # thin volumes, however using them for snapshot volumes is less efficient,
+    # as it consumes more space and takes extra time for copying.
+    # When unset, lvm tries to estimate chunk size starting from 64KB
+    # Supported values are in range from 64 to 1048576.
+    # thin_pool_chunk_size = 64
+
+    # Specify discards behavior of the thin pool volume.
+    # Select one of  "ignore", "nopassdown", "passdown"
+    # thin_pool_discards = "passdown"
+
+    # Set to 0, to disable zeroing of thin pool data chunks before their
+    # first use.
+    # N.B. zeroing larger thin pool chunk size degrades performance.
+    # thin_pool_zero = 1
+}
+
+# This section that allows you to configure the nature of the
+# information that LVM2 reports.
+log {
+
+    # Controls the messages sent to stdout or stderr.
+    # There are three levels of verbosity, 3 being the most verbose.
+    verbose = 0
+
+    # Set to 1 to suppress all non-essential messages from stdout.
+    # This has the same effect as -qq.
+    # When this is set, the following commands still produce output:
+    # dumpconfig, lvdisplay, lvmdiskscan, lvs, pvck, pvdisplay, 
+    # pvs, version, vgcfgrestore -l, vgdisplay, vgs.
+    # Non-essential messages are shifted from log level 4 to log level 5
+    # for syslog and lvm2_log_fn purposes.
+    # Any 'yes' or 'no' questions not overridden by other arguments
+    # are suppressed and default to 'no'.
+    silent = 0
+
+    # Should we send log messages through syslog?
+    # 1 is yes; 0 is no.
+    syslog = 1
+
+    # Should we log error and debug messages to a file?
+    # By default there is no log file.
+    #file = "/var/log/lvm2.log"
+
+    # Should we overwrite the log file each time the program is run?
+    # By default we append.
+    overwrite = 0
+
+    # What level of log messages should we send to the log file and/or syslog?
+    # There are 6 syslog-like log levels currently in use - 2 to 7 inclusive.
+    # 7 is the most verbose (LOG_DEBUG).
+    level = 0
+
+    # Format of output messages
+    # Whether or not (1 or 0) to indent messages according to their severity
+    indent = 1
+
+    # Whether or not (1 or 0) to display the command name on each line output
+    command_names = 0
+
+    # A prefix to use before the message text (but after the command name,
+    # if selected).  Default is two spaces, so you can see/grep the severity
+    # of each message.
+    prefix = "  "
+
+    # To make the messages look similar to the original LVM tools use:
+    #   indent = 0
+    #   command_names = 1
+    #   prefix = " -- "
+
+    # Set this if you want log messages during activation.
+    # Don't use this in low memory situations (can deadlock).
+    # activation = 0
+
+    # Some debugging messages are assigned to a class and only appear
+    # in debug output if the class is listed here.
+    # Classes currently available:
+    #   memory, devices, activation, allocation, lvmetad, metadata, cache,
+    #   locking
+    # Use "all" to see everything.
+    debug_classes = [ "memory", "devices", "activation", "allocation",
+		      "lvmetad", "metadata", "cache", "locking" ]
+}
+
+# Configuration of metadata backups and archiving.  In LVM2 when we
+# talk about a 'backup' we mean making a copy of the metadata for the
+# *current* system.  The 'archive' contains old metadata configurations.
+# Backups are stored in a human readeable text format.
+backup {
+
+    # Should we maintain a backup of the current metadata configuration ?
+    # Use 1 for Yes; 0 for No.
+    # Think very hard before turning this off!
+    backup = 1
+
+    # Where shall we keep it ?
+    # Remember to back up this directory regularly!
+    backup_dir = "@DEFAULT_SYS_DIR@/@DEFAULT_BACKUP_SUBDIR@"
+
+    # Should we maintain an archive of old metadata configurations.
+    # Use 1 for Yes; 0 for No.
+    # On by default.  Think very hard before turning this off.
+    archive = 1
+
+    # Where should archived files go ?
+    # Remember to back up this directory regularly!
+    archive_dir = "@DEFAULT_SYS_DIR@/@DEFAULT_ARCHIVE_SUBDIR@"
+
+    # What is the minimum number of archive files you wish to keep ?
+    retain_min = 10
+
+    # What is the minimum time you wish to keep an archive file for ?
+    retain_days = 30
+}
+
+# Settings for the running LVM2 in shell (readline) mode.
+shell {
+
+    # Number of lines of history to store in ~/.lvm_history
+    history_size = 100
+}
+
+
+# Miscellaneous global LVM2 settings
+global {
+    # The file creation mask for any files and directories created.
+    # Interpreted as octal if the first digit is zero.
+    umask = 077
+
+    # Allow other users to read the files
+    #umask = 022
+
+    # Enabling test mode means that no changes to the on disk metadata
+    # will be made.  Equivalent to having the -t option on every
+    # command.  Defaults to off.
+    test = 0
+
+    # Default value for --units argument
+    units = "h"
+
+    # Since version 2.02.54, the tools distinguish between powers of
+    # 1024 bytes (e.g. KiB, MiB, GiB) and powers of 1000 bytes (e.g.
+    # KB, MB, GB).
+    # If you have scripts that depend on the old behaviour, set this to 0
+    # temporarily until you update them.
+    si_unit_consistency = 1
+
+    # Whether or not to communicate with the kernel device-mapper.
+    # Set to 0 if you want to use the tools to manipulate LVM metadata 
+    # without activating any logical volumes.
+    # If the device-mapper kernel driver is not present in your kernel
+    # setting this to 0 should suppress the error messages.
+    activation = 1
+
+    # If we can't communicate with device-mapper, should we try running 
+    # the LVM1 tools?
+    # This option only applies to 2.4 kernels and is provided to help you
+    # switch between device-mapper kernels and LVM1 kernels.
+    # The LVM1 tools need to be installed with .lvm1 suffices
+    # e.g. vgscan.lvm1 and they will stop working after you start using
+    # the new lvm2 on-disk metadata format.
+    # The default value is set when the tools are built.
+    # fallback_to_lvm1 = 0
+
+    # The default metadata format that commands should use - "lvm1" or "lvm2".
+    # The command line override is -M1 or -M2.
+    # Defaults to "lvm2".
+    # format = "lvm2"
+
+    # Location of proc filesystem
+    proc = "/proc"
+
+    # Type of locking to use. Defaults to local file-based locking (1).
+    # Turn locking off by setting to 0 (dangerous: risks metadata corruption
+    # if LVM2 commands get run concurrently).
+    # Type 2 uses the external shared library locking_library.
+    # Type 3 uses built-in clustered locking.
+    # Type 4 uses read-only locking which forbids any operations that might 
+    # change metadata.
+    locking_type = 1
+
+    # Set to 0 to fail when a lock request cannot be satisfied immediately.
+    wait_for_locks = 1
+
+    # If using external locking (type 2) and initialisation fails,
+    # with this set to 1 an attempt will be made to use the built-in
+    # clustered locking.
+    # If you are using a customised locking_library you should set this to 0.
+    fallback_to_clustered_locking = 1
+
+    # If an attempt to initialise type 2 or type 3 locking failed, perhaps
+    # because cluster components such as clvmd are not running, with this set
+    # to 1 an attempt will be made to use local file-based locking (type 1).
+    # If this succeeds, only commands against local volume groups will proceed.
+    # Volume Groups marked as clustered will be ignored.
+    fallback_to_local_locking = 1
+
+    # Local non-LV directory that holds file-based locks while commands are
+    # in progress.  A directory like /tmp that may get wiped on reboot is OK.
+    locking_dir = "@DEFAULT_LOCK_DIR@"
+
+    # Whenever there are competing read-only and read-write access requests for
+    # a volume group's metadata, instead of always granting the read-only
+    # requests immediately, delay them to allow the read-write requests to be
+    # serviced.  Without this setting, write access may be stalled by a high
+    # volume of read-only requests.
+    # NB. This option only affects locking_type = 1 viz. local file-based
+    # locking.
+    prioritise_write_locks = 1
+
+    # Other entries can go here to allow you to load shared libraries
+    # e.g. if support for LVM1 metadata was compiled as a shared library use
+    #   format_libraries = "liblvm2format1.so" 
+    # Full pathnames can be given.
+
+    # Search this directory first for shared libraries.
+    #   library_dir = "/lib"
+
+    # The external locking library to load if locking_type is set to 2.
+    #   locking_library = "liblvm2clusterlock.so"
+
+    # Treat any internal errors as fatal errors, aborting the process that
+    # encountered the internal error. Please only enable for debugging.
+    abort_on_internal_errors = 0
+
+    # Check whether CRC is matching when parsed VG is used multiple times.
+    # This is useful to catch unexpected internal cached volume group
+    # structure modification. Please only enable for debugging.
+    detect_internal_vg_cache_corruption = 0
+
+    # If set to 1, no operations that change on-disk metadata will be permitted.
+    # Additionally, read-only commands that encounter metadata in need of repair
+    # will still be allowed to proceed exactly as if the repair had been 
+    # performed (except for the unchanged vg_seqno).
+    # Inappropriate use could mess up your system, so seek advice first!
+    metadata_read_only = 0
+
+    # 'mirror_segtype_default' defines which segtype will be used when the
+    # shorthand '-m' option is used for mirroring.  The possible options are:
+    #
+    # "mirror" - The original RAID1 implementation provided by LVM2/DM.  It is
+    # 	         characterized by a flexible log solution (core, disk, mirrored)
+    #		 and by the necessity to block I/O while reconfiguring in the
+    #		 event of a failure.
+    #
+    #		 There is an inherent race in the dmeventd failure handling
+    #		 logic with snapshots of devices using this type of RAID1 that
+    #		 in the worst case could cause a deadlock.
+    #		   Ref: https://bugzilla.redhat.com/show_bug.cgi?id=817130#c10
+    #
+    # "raid1"  - This implementation leverages MD's RAID1 personality through
+    # 	       	 device-mapper.  It is characterized by a lack of log options.
+    #		 (A log is always allocated for every device and they are placed
+    #		 on the same device as the image - no separate devices are
+    #		 required.)  This mirror implementation does not require I/O
+    #		 to be blocked in the kernel in the event of a failure.
+    #		 This mirror implementation is not cluster-aware and cannot be
+    #		 used in a shared (active/active) fashion in a cluster.
+    #
+    # Specify the '--type <mirror|raid1>' option to override this default
+    # setting.
+    mirror_segtype_default = "mirror"
+
+    # 'raid10_segtype_default' determines the segment types used by default
+    # when the '--stripes/-i' and '--mirrors/-m' arguments are both specified
+    # during the creation of a logical volume.
+    # Possible settings include:
+    #
+    # "raid10" - This implementation leverages MD's RAID10 personality through
+    #            device-mapper.
+    #
+    # "mirror" - LVM will layer the 'mirror' and 'stripe' segment types.  It
+    #            will do this by creating a mirror on top of striped sub-LVs;
+    #            effectively creating a RAID 0+1 array.  This is suboptimal
+    #            in terms of providing redunancy and performance.  Changing to
+    #            this setting is not advised.
+    # Specify the '--type <raid10|mirror>' option to override this default
+    # setting.
+    raid10_segtype_default = "mirror"
+
+    # The default format for displaying LV names in lvdisplay was changed 
+    # in version 2.02.89 to show the LV name and path separately.
+    # Previously this was always shown as /dev/vgname/lvname even when that
+    # was never a valid path in the /dev filesystem.
+    # Set to 1 to reinstate the previous format.
+    #
+    # lvdisplay_shows_full_device_path = 0
+
+    # Whether to use (trust) a running instance of lvmetad. If this is set to
+    # 0, all commands fall back to the usual scanning mechanisms. When set to 1
+    # *and* when lvmetad is running (it is not auto-started), the volume group
+    # metadata and PV state flags are obtained from the lvmetad instance and no
+    # scanning is done by the individual commands. In a setup with lvmetad,
+    # lvmetad udev rules *must* be set up for LVM to work correctly. Without
+    # proper udev rules, all changes in block device configuration will be
+    # *ignored* until a manual 'pvscan --cache' is performed.
+    #
+    # If lvmetad has been running while use_lvmetad was 0, it MUST be stopped
+    # before changing use_lvmetad to 1 and started again afterwards.
+    use_lvmetad = 0
+
+    # Full path of the utility called to check that a thin metadata device
+    # is in a state that allows it to be used.
+    # Each time a thin pool needs to be activated or after it is deactivated
+    # this utility is executed. The activation will only proceed if the utility
+    # has an exit status of 0.
+    # Set to "" to skip this check.  (Not recommended.)
+    # The thin tools are available as part of the device-mapper-persistent-data
+    # package from https://github.com/jthornber/thin-provisioning-tools.
+    #
+    thin_check_executable = "@THIN_CHECK_CMD@"
+
+    # String with options passed with thin_check command. By default,
+    # option '-q' is for quiet output.
+    thin_check_options = [ "-q" ]
+
+    # If set, given features are not used by thin driver.
+    # This can be helpful not just for testing, but i.e. allows to avoid
+    # using problematic implementation of some thin feature.
+    # Features:
+    #   block_size
+    #   discards
+    #   discards_non_power_2
+    #
+    # thin_disabled_features = [ "discards", "block_size" ]
+}
+
+activation {
+    # Set to 1 to perform internal checks on the operations issued to
+    # libdevmapper.  Useful for debugging problems with activation.
+    # Some of the checks may be expensive, so it's best to use this
+    # only when there seems to be a problem.
+    checks = 0
+
+    # Set to 0 to disable udev synchronisation (if compiled into the binaries).
+    # Processes will not wait for notification from udev.
+    # They will continue irrespective of any possible udev processing
+    # in the background.  You should only use this if udev is not running
+    # or has rules that ignore the devices LVM2 creates.
+    # The command line argument --nodevsync takes precedence over this setting.
+    # If set to 1 when udev is not running, and there are LVM2 processes
+    # waiting for udev, run 'dmsetup udevcomplete_all' manually to wake them up.
+    udev_sync = 1
+
+    # Set to 0 to disable the udev rules installed by LVM2 (if built with
+    # --enable-udev_rules). LVM2 will then manage the /dev nodes and symlinks
+    # for active logical volumes directly itself.
+    # N.B. Manual intervention may be required if this setting is changed
+    # while any logical volumes are active.
+    udev_rules = 1
+
+    # Set to 1 for LVM2 to verify operations performed by udev. This turns on
+    # additional checks (and if necessary, repairs) on entries in the device
+    # directory after udev has completed processing its events. 
+    # Useful for diagnosing problems with LVM2/udev interactions.
+    verify_udev_operations = 0
+
+    # If set to 1 and if deactivation of an LV fails, perhaps because
+    # a process run from a quick udev rule temporarily opened the device,
+    # retry the operation for a few seconds before failing.
+    retry_deactivation = 1
+
+    # How to fill in missing stripes if activating an incomplete volume.
+    # Using "error" will make inaccessible parts of the device return
+    # I/O errors on access.  You can instead use a device path, in which 
+    # case, that device will be used to in place of missing stripes.
+    # But note that using anything other than "error" with mirrored 
+    # or snapshotted volumes is likely to result in data corruption.
+    missing_stripe_filler = "error"
+
+    # The linear target is an optimised version of the striped target
+    # that only handles a single stripe.  Set this to 0 to disable this
+    # optimisation and always use the striped target.
+    use_linear_target = 1
+
+    # How much stack (in KB) to reserve for use while devices suspended
+    # Prior to version 2.02.89 this used to be set to 256KB
+    reserved_stack = 64
+
+    # How much memory (in KB) to reserve for use while devices suspended
+    reserved_memory = 8192
+
+    # Nice value used while devices suspended
+    process_priority = -18
+
+    # If volume_list is defined, each LV is only activated if there is a
+    # match against the list.
+    #
+    #   "vgname" and "vgname/lvname" are matched exactly.
+    #   "@tag" matches any tag set in the LV or VG.
+    #   "@*" matches if any tag defined on the host is also set in the LV or VG
+    #
+    # If any host tags exist but volume_list is not defined, a default
+    # single-entry list containing "@*" is assumed.
+    #
+    # volume_list = [ "vg1", "vg2/lvol1", "@tag1", "@*" ]
+
+    # If auto_activation_volume_list is defined, each LV that is to be
+    # activated with the autoactivation option (--activate ay/-a ay)
+    # is first checked against the list.  If it does not match, the LV
+    # is not activated.  This list is checked as well as volume_list.
+    #
+    #   "vgname" and "vgname/lvname" are matched exactly.
+    #   "@tag" matches any tag set in the LV or VG.
+    #   "@*" matches if any tag defined on the host is also set in the LV or VG
+    #
+    # auto_activation_volume_list = [ "vg1", "vg2/lvol1", "@tag1", "@*" ]
+
+    # If read_only_volume_list is defined, each LV that is to be activated 
+    # is checked against the list, and if it matches, it as activated
+    # in read-only mode.  (This overrides '--permission rw' stored in the
+    # metadata.)
+    #
+    #   "vgname" and "vgname/lvname" are matched exactly.
+    #   "@tag" matches any tag set in the LV or VG.
+    #   "@*" matches if any tag defined on the host is also set in the LV or VG
+    #
+    # read_only_volume_list = [ "vg1", "vg2/lvol1", "@tag1", "@*" ]
+
+    # For RAID or 'mirror' segment types, 'raid_region_size' is the
+    # size (in kiB) of each:
+    # - synchronization operation when initializing
+    # - each copy operation when performing a 'pvmove' (using 'mirror' segtype)
+    # This setting has replaced 'mirror_region_size' since version 2.02.99
+    raid_region_size = 512
+
+    # Setting to use when there is no readahead value stored in the metadata.
+    #
+    # "none" - Disable readahead.
+    # "auto" - Use default value chosen by kernel.
+    readahead = "auto"
+
+    # 'raid_fault_policy' defines how a device failure in a RAID logical
+    # volume is handled.  This includes logical volumes that have the following
+    # segment types: raid1, raid4, raid5*, and raid6*.
+    #
+    # In the event of a failure, the following policies will determine what
+    # actions are performed during the automated response to failures (when
+    # dmeventd is monitoring the RAID logical volume) and when 'lvconvert' is
+    # called manually with the options '--repair' and '--use-policies'.
+    #
+    # "warn"	- Use the system log to warn the user that a device in the RAID
+    # 		  logical volume has failed.  It is left to the user to run
+    #		  'lvconvert --repair' manually to remove or replace the failed
+    #		  device.  As long as the number of failed devices does not
+    #		  exceed the redundancy of the logical volume (1 device for
+    #		  raid4/5, 2 for raid6, etc) the logical volume will remain
+    #		  usable.
+    #
+    # "allocate" - Attempt to use any extra physical volumes in the volume
+    # 		  group as spares and replace faulty devices.
+    #
+    raid_fault_policy = "warn"
+
+    # 'mirror_image_fault_policy' and 'mirror_log_fault_policy' define
+    # how a device failure affecting a mirror (of "mirror" segment type) is
+    # handled.  A mirror is composed of mirror images (copies) and a log.
+    # A disk log ensures that a mirror does not need to be re-synced
+    # (all copies made the same) every time a machine reboots or crashes.
+    #
+    # In the event of a failure, the specified policy will be used to determine
+    # what happens. This applies to automatic repairs (when the mirror is being
+    # monitored by dmeventd) and to manual lvconvert --repair when
+    # --use-policies is given.
+    #
+    # "remove" - Simply remove the faulty device and run without it.  If
+    #            the log device fails, the mirror would convert to using
+    #            an in-memory log.  This means the mirror will not
+    #            remember its sync status across crashes/reboots and
+    #            the entire mirror will be re-synced.  If a
+    #            mirror image fails, the mirror will convert to a
+    #            non-mirrored device if there is only one remaining good
+    #            copy.
+    #
+    # "allocate" - Remove the faulty device and try to allocate space on
+    #            a new device to be a replacement for the failed device.
+    #            Using this policy for the log is fast and maintains the
+    #            ability to remember sync state through crashes/reboots.
+    #            Using this policy for a mirror device is slow, as it
+    #            requires the mirror to resynchronize the devices, but it
+    #            will preserve the mirror characteristic of the device.
+    #            This policy acts like "remove" if no suitable device and
+    #            space can be allocated for the replacement.
+    #
+    # "allocate_anywhere" - Not yet implemented. Useful to place the log device
+    #            temporarily on same physical volume as one of the mirror
+    #            images. This policy is not recommended for mirror devices
+    #            since it would break the redundant nature of the mirror. This
+    #            policy acts like "remove" if no suitable device and space can
+    #            be allocated for the replacement.
+
+    mirror_log_fault_policy = "allocate"
+    mirror_image_fault_policy = "remove"
+
+    # 'snapshot_autoextend_threshold' and 'snapshot_autoextend_percent' define
+    # how to handle automatic snapshot extension. The former defines when the
+    # snapshot should be extended: when its space usage exceeds this many
+    # percent. The latter defines how much extra space should be allocated for
+    # the snapshot, in percent of its current size.
+    #
+    # For example, if you set snapshot_autoextend_threshold to 70 and
+    # snapshot_autoextend_percent to 20, whenever a snapshot exceeds 70% usage,
+    # it will be extended by another 20%. For a 1G snapshot, using up 700M will
+    # trigger a resize to 1.2G. When the usage exceeds 840M, the snapshot will
+    # be extended to 1.44G, and so on.
+    #
+    # Setting snapshot_autoextend_threshold to 100 disables automatic
+    # extensions. The minimum value is 50 (A setting below 50 will be treated
+    # as 50).
+
+    snapshot_autoextend_threshold = 100
+    snapshot_autoextend_percent = 20
+
+    # 'thin_pool_autoextend_threshold' and 'thin_pool_autoextend_percent' define
+    # how to handle automatic pool extension. The former defines when the
+    # pool should be extended: when its space usage exceeds this many
+    # percent. The latter defines how much extra space should be allocated for
+    # the pool, in percent of its current size.
+    #
+    # For example, if you set thin_pool_autoextend_threshold to 70 and
+    # thin_pool_autoextend_percent to 20, whenever a pool exceeds 70% usage,
+    # it will be extended by another 20%. For a 1G pool, using up 700M will
+    # trigger a resize to 1.2G. When the usage exceeds 840M, the pool will
+    # be extended to 1.44G, and so on.
+    #
+    # Setting thin_pool_autoextend_threshold to 100 disables automatic
+    # extensions. The minimum value is 50 (A setting below 50 will be treated
+    # as 50).
+
+    thin_pool_autoextend_threshold = 100
+    thin_pool_autoextend_percent = 20
+
+    # While activating devices, I/O to devices being (re)configured is
+    # suspended, and as a precaution against deadlocks, LVM2 needs to pin
+    # any memory it is using so it is not paged out.  Groups of pages that
+    # are known not to be accessed during activation need not be pinned
+    # into memory.  Each string listed in this setting is compared against
+    # each line in /proc/self/maps, and the pages corresponding to any
+    # lines that match are not pinned.  On some systems locale-archive was
+    # found to make up over 80% of the memory used by the process.
+    # mlock_filter = [ "locale/locale-archive", "gconv/gconv-modules.cache" ]
+
+    # Set to 1 to revert to the default behaviour prior to version 2.02.62
+    # which used mlockall() to pin the whole process's memory while activating
+    # devices.
+    use_mlockall = 0
+
+    # Monitoring is enabled by default when activating logical volumes.
+    # Set to 0 to disable monitoring or use the --ignoremonitoring option.
+    monitoring = 1
+
+    # When pvmove or lvconvert must wait for the kernel to finish
+    # synchronising or merging data, they check and report progress
+    # at intervals of this number of seconds.  The default is 15 seconds.
+    # If this is set to 0 and there is only one thing to wait for, there
+    # are no progress reports, but the process is awoken immediately the
+    # operation is complete.
+    polling_interval = 15
+}
+
+
+####################
+# Advanced section #
+####################
+
+# Metadata settings
+#
+# metadata {
+    # Default number of copies of metadata to hold on each PV.  0, 1 or 2.
+    # You might want to override it from the command line with 0 
+    # when running pvcreate on new PVs which are to be added to large VGs.
+
+    # pvmetadatacopies = 1
+
+    # Default number of copies of metadata to maintain for each VG.
+    # If set to a non-zero value, LVM automatically chooses which of
+    # the available metadata areas to use to achieve the requested
+    # number of copies of the VG metadata.  If you set a value larger
+    # than the the total number of metadata areas available then
+    # metadata is stored in them all.
+    # The default value of 0 ("unmanaged") disables this automatic
+    # management and allows you to control which metadata areas
+    # are used at the individual PV level using 'pvchange
+    # --metadataignore y/n'.
+
+    # vgmetadatacopies = 0
+
+    # Approximate default size of on-disk metadata areas in sectors.
+    # You should increase this if you have large volume groups or
+    # you want to retain a large on-disk history of your metadata changes.
+
+    # pvmetadatasize = 255
+
+    # List of directories holding live copies of text format metadata.
+    # These directories must not be on logical volumes!
+    # It's possible to use LVM2 with a couple of directories here,
+    # preferably on different (non-LV) filesystems, and with no other 
+    # on-disk metadata (pvmetadatacopies = 0). Or this can be in
+    # addition to on-disk metadata areas.
+    # The feature was originally added to simplify testing and is not
+    # supported under low memory situations - the machine could lock up.
+    #
+    # Never edit any files in these directories by hand unless you
+    # you are absolutely sure you know what you are doing! Use
+    # the supplied toolset to make changes (e.g. vgcfgrestore).
+
+    # dirs = [ "/etc/lvm/metadata", "/mnt/disk2/lvm/metadata2" ]
+#}
+
+# Event daemon
+#
+dmeventd {
+    # mirror_library is the library used when monitoring a mirror device.
+    #
+    # "libdevmapper-event-lvm2mirror.so" attempts to recover from
+    # failures.  It removes failed devices from a volume group and
+    # reconfigures a mirror as necessary. If no mirror library is
+    # provided, mirrors are not monitored through dmeventd.
+
+    mirror_library = "libdevmapper-event-lvm2mirror.so"
+
+    # snapshot_library is the library used when monitoring a snapshot device.
+    #
+    # "libdevmapper-event-lvm2snapshot.so" monitors the filling of
+    # snapshots and emits a warning through syslog when the use of
+    # the snapshot exceeds 80%. The warning is repeated when 85%, 90% and
+    # 95% of the snapshot is filled.
+
+    snapshot_library = "libdevmapper-event-lvm2snapshot.so"
+
+    # thin_library is the library used when monitoring a thin device.
+    #
+    # "libdevmapper-event-lvm2thin.so" monitors the filling of
+    # pool and emits a warning through syslog when the use of
+    # the pool exceeds 80%. The warning is repeated when 85%, 90% and
+    # 95% of the pool is filled.
+
+    thin_library = "libdevmapper-event-lvm2thin.so"
+
+    # Full path of the dmeventd binary.
+    #
+    # executable = "@DMEVENTD_PATH@"
+}

doc/example_cmdlib.c

@@ -0,0 +1,49 @@
+/*
+ * Copyright (C) 2004 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU General Public License v.2.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include "lvm2cmd.h"
+#include <stdio.h>
+
+/* All output gets passed to this function line-by-line */
+void test_log_fn(int level, const char *file, int line,
+		 int dm_errno, const char *format)
+{
+	/* Extract and process output here rather than printing it */
+
+	if (level != 4)
+		return;
+
+	printf("%s\n", format);
+	return;
+}
+
+int main(int argc, char **argv)
+{
+	void *handle;
+	int r;
+
+	lvm2_log_fn(test_log_fn);
+
+	handle = lvm2_init();
+
+	lvm2_log_level(handle, 1);
+	r = lvm2_run(handle, "vgs --noheadings vg1");
+
+	/* More commands here */
+
+	lvm2_exit(handle);
+
+	return r;
+}
+

doc/kernel/cache-policies.txt

@@ -0,0 +1,77 @@
+Guidance for writing policies
+=============================
+
+Try to keep transactionality out of it.  The core is careful to
+avoid asking about anything that is migrating.  This is a pain, but
+makes it easier to write the policies.
+
+Mappings are loaded into the policy at construction time.
+
+Every bio that is mapped by the target is referred to the policy.
+The policy can return a simple HIT or MISS or issue a migration.
+
+Currently there's no way for the policy to issue background work,
+e.g. to start writing back dirty blocks that are going to be evicte
+soon.
+
+Because we map bios, rather than requests it's easy for the policy
+to get fooled by many small bios.  For this reason the core target
+issues periodic ticks to the policy.  It's suggested that the policy
+doesn't update states (eg, hit counts) for a block more than once
+for each tick.  The core ticks by watching bios complete, and so
+trying to see when the io scheduler has let the ios run.
+
+
+Overview of supplied cache replacement policies
+===============================================
+
+multiqueue
+----------
+
+This policy is the default.
+
+The multiqueue policy has two sets of 16 queues: one set for entries
+waiting for the cache and another one for those in the cache.
+Cache entries in the queues are aged based on logical time. Entry into
+the cache is based on variable thresholds and queue selection is based
+on hit count on entry. The policy aims to take different cache miss
+costs into account and to adjust to varying load patterns automatically.
+
+Message and constructor argument pairs are:
+	'sequential_threshold <#nr_sequential_ios>' and
+	'random_threshold <#nr_random_ios>'.
+
+The sequential threshold indicates the number of contiguous I/Os
+required before a stream is treated as sequential.  The random threshold
+is the number of intervening non-contiguous I/Os that must be seen
+before the stream is treated as random again.
+
+The sequential and random thresholds default to 512 and 4 respectively.
+
+Large, sequential ios are probably better left on the origin device
+since spindles tend to have good bandwidth. The io_tracker counts
+contiguous I/Os to try to spot when the io is in one of these sequential
+modes.
+
+cleaner
+-------
+
+The cleaner writes back all dirty blocks in a cache to decommission it.
+
+Examples
+========
+
+The syntax for a table is:
+	cache <metadata dev> <cache dev> <origin dev> <block size>
+	<#feature_args> [<feature arg>]*
+	<policy> <#policy_args> [<policy arg>]*
+
+The syntax to send a message using the dmsetup command is:
+	dmsetup message <mapped device> 0 sequential_threshold 1024
+	dmsetup message <mapped device> 0 random_threshold 8
+
+Using dmsetup:
+	dmsetup create blah --table "0 268435456 cache /dev/sdb /dev/sdc \
+	    /dev/sdd 512 0 mq 4 sequential_threshold 1024 random_threshold 8"
+	creates a 128GB large mapped device named 'blah' with the
+	sequential threshold set to 1024 and the random_threshold set to 8.

doc/kernel/cache.txt

@@ -0,0 +1,243 @@
+Introduction
+============
+
+dm-cache is a device mapper target written by Joe Thornber, Heinz
+Mauelshagen, and Mike Snitzer.
+
+It aims to improve performance of a block device (eg, a spindle) by
+dynamically migrating some of its data to a faster, smaller device
+(eg, an SSD).
+
+This device-mapper solution allows us to insert this caching at
+different levels of the dm stack, for instance above the data device for
+a thin-provisioning pool.  Caching solutions that are integrated more
+closely with the virtual memory system should give better performance.
+
+The target reuses the metadata library used in the thin-provisioning
+library.
+
+The decision as to what data to migrate and when is left to a plug-in
+policy module.  Several of these have been written as we experiment,
+and we hope other people will contribute others for specific io
+scenarios (eg. a vm image server).
+
+Glossary
+========
+
+  Migration -  Movement of the primary copy of a logical block from one
+	       device to the other.
+  Promotion -  Migration from slow device to fast device.
+  Demotion  -  Migration from fast device to slow device.
+
+The origin device always contains a copy of the logical block, which
+may be out of date or kept in sync with the copy on the cache device
+(depending on policy).
+
+Design
+======
+
+Sub-devices
+-----------
+
+The target is constructed by passing three devices to it (along with
+other parameters detailed later):
+
+1. An origin device - the big, slow one.
+
+2. A cache device - the small, fast one.
+
+3. A small metadata device - records which blocks are in the cache,
+   which are dirty, and extra hints for use by the policy object.
+   This information could be put on the cache device, but having it
+   separate allows the volume manager to configure it differently,
+   e.g. as a mirror for extra robustness.
+
+Fixed block size
+----------------
+
+The origin is divided up into blocks of a fixed size.  This block size
+is configurable when you first create the cache.  Typically we've been
+using block sizes of 256k - 1024k.
+
+Having a fixed block size simplifies the target a lot.  But it is
+something of a compromise.  For instance, a small part of a block may be
+getting hit a lot, yet the whole block will be promoted to the cache.
+So large block sizes are bad because they waste cache space.  And small
+block sizes are bad because they increase the amount of metadata (both
+in core and on disk).
+
+Writeback/writethrough
+----------------------
+
+The cache has two modes, writeback and writethrough.
+
+If writeback, the default, is selected then a write to a block that is
+cached will go only to the cache and the block will be marked dirty in
+the metadata.
+
+If writethrough is selected then a write to a cached block will not
+complete until it has hit both the origin and cache devices.  Clean
+blocks should remain clean.
+
+A simple cleaner policy is provided, which will clean (write back) all
+dirty blocks in a cache.  Useful for decommissioning a cache.
+
+Migration throttling
+--------------------
+
+Migrating data between the origin and cache device uses bandwidth.
+The user can set a throttle to prevent more than a certain amount of
+migration occuring at any one time.  Currently we're not taking any
+account of normal io traffic going to the devices.  More work needs
+doing here to avoid migrating during those peak io moments.
+
+For the time being, a message "migration_threshold <#sectors>"
+can be used to set the maximum number of sectors being migrated,
+the default being 204800 sectors (or 100MB).
+
+Updating on-disk metadata
+-------------------------
+
+On-disk metadata is committed every time a REQ_SYNC or REQ_FUA bio is
+written.  If no such requests are made then commits will occur every
+second.  This means the cache behaves like a physical disk that has a
+write cache (the same is true of the thin-provisioning target).  If
+power is lost you may lose some recent writes.  The metadata should
+always be consistent in spite of any crash.
+
+The 'dirty' state for a cache block changes far too frequently for us
+to keep updating it on the fly.  So we treat it as a hint.  In normal
+operation it will be written when the dm device is suspended.  If the
+system crashes all cache blocks will be assumed dirty when restarted.
+
+Per-block policy hints
+----------------------
+
+Policy plug-ins can store a chunk of data per cache block.  It's up to
+the policy how big this chunk is, but it should be kept small.  Like the
+dirty flags this data is lost if there's a crash so a safe fallback
+value should always be possible.
+
+For instance, the 'mq' policy, which is currently the default policy,
+uses this facility to store the hit count of the cache blocks.  If
+there's a crash this information will be lost, which means the cache
+may be less efficient until those hit counts are regenerated.
+
+Policy hints affect performance, not correctness.
+
+Policy messaging
+----------------
+
+Policies will have different tunables, specific to each one, so we
+need a generic way of getting and setting these.  Device-mapper
+messages are used.  Refer to cache-policies.txt.
+
+Discard bitset resolution
+-------------------------
+
+We can avoid copying data during migration if we know the block has
+been discarded.  A prime example of this is when mkfs discards the
+whole block device.  We store a bitset tracking the discard state of
+blocks.  However, we allow this bitset to have a different block size
+from the cache blocks.  This is because we need to track the discard
+state for all of the origin device (compare with the dirty bitset
+which is just for the smaller cache device).
+
+Target interface
+================
+
+Constructor
+-----------
+
+ cache <metadata dev> <cache dev> <origin dev> <block size>
+       <#feature args> [<feature arg>]*
+       <policy> <#policy args> [policy args]*
+
+ metadata dev    : fast device holding the persistent metadata
+ cache dev	 : fast device holding cached data blocks
+ origin dev	 : slow device holding original data blocks
+ block size      : cache unit size in sectors
+
+ #feature args   : number of feature arguments passed
+ feature args    : writethrough.  (The default is writeback.)
+
+ policy          : the replacement policy to use
+ #policy args    : an even number of arguments corresponding to
+                   key/value pairs passed to the policy
+ policy args     : key/value pairs passed to the policy
+		   E.g. 'sequential_threshold 1024'
+		   See cache-policies.txt for details.
+
+Optional feature arguments are:
+   writethrough  : write through caching that prohibits cache block
+		   content from being different from origin block content.
+		   Without this argument, the default behaviour is to write
+		   back cache block contents later for performance reasons,
+		   so they may differ from the corresponding origin blocks.
+
+A policy called 'default' is always registered.  This is an alias for
+the policy we currently think is giving best all round performance.
+
+As the default policy could vary between kernels, if you are relying on
+the characteristics of a specific policy, always request it by name.
+
+Status
+------
+
+<#used metadata blocks>/<#total metadata blocks> <#read hits> <#read misses>
+<#write hits> <#write misses> <#demotions> <#promotions> <#blocks in cache>
+<#dirty> <#features> <features>* <#core args> <core args>* <#policy args>
+<policy args>*
+
+#used metadata blocks    : Number of metadata blocks used
+#total metadata blocks   : Total number of metadata blocks
+#read hits               : Number of times a READ bio has been mapped
+			     to the cache
+#read misses             : Number of times a READ bio has been mapped
+			     to the origin
+#write hits              : Number of times a WRITE bio has been mapped
+			     to the cache
+#write misses            : Number of times a WRITE bio has been
+			     mapped to the origin
+#demotions               : Number of times a block has been removed
+			     from the cache
+#promotions              : Number of times a block has been moved to
+			     the cache
+#blocks in cache         : Number of blocks resident in the cache
+#dirty                   : Number of blocks in the cache that differ
+			     from the origin
+#feature args            : Number of feature args to follow
+feature args             : 'writethrough' (optional)
+#core args               : Number of core arguments (must be even)
+core args                : Key/value pairs for tuning the core
+			     e.g. migration_threshold
+#policy args             : Number of policy arguments to follow (must be even)
+policy args              : Key/value pairs
+			     e.g. 'sequential_threshold 1024
+
+Messages
+--------
+
+Policies will have different tunables, specific to each one, so we
+need a generic way of getting and setting these.  Device-mapper
+messages are used.  (A sysfs interface would also be possible.)
+
+The message format is:
+
+   <key> <value>
+
+E.g.
+   dmsetup message my_cache 0 sequential_threshold 1024
+
+Examples
+========
+
+The test suite can be found here:
+
+https://github.com/jthornber/thinp-test-suite
+
+dmsetup create my_cache --table '0 41943040 cache /dev/mapper/metadata \
+	/dev/mapper/ssd /dev/mapper/origin 512 1 writeback default 0'
+dmsetup create my_cache --table '0 41943040 cache /dev/mapper/metadata \
+	/dev/mapper/ssd /dev/mapper/origin 1024 1 writeback \
+	mq 4 sequential_threshold 1024 random_threshold 8'

doc/kernel/crypt.txt

@@ -0,0 +1,76 @@
+dm-crypt
+=========
+
+Device-Mapper's "crypt" target provides transparent encryption of block devices
+using the kernel crypto API.
+
+Parameters: <cipher> <key> <iv_offset> <device path> \
+	      <offset> [<#opt_params> <opt_params>]
+
+<cipher>
+    Encryption cipher and an optional IV generation mode.
+    (In format cipher[:keycount]-chainmode-ivopts:ivmode).
+    Examples:
+       des
+       aes-cbc-essiv:sha256
+       twofish-ecb
+
+    /proc/crypto contains supported crypto modes
+
+<key>
+    Key used for encryption. It is encoded as a hexadecimal number.
+    You can only use key sizes that are valid for the selected cipher.
+
+<keycount>
+    Multi-key compatibility mode. You can define <keycount> keys and
+    then sectors are encrypted according to their offsets (sector 0 uses key0;
+    sector 1 uses key1 etc.).  <keycount> must be a power of two.
+
+<iv_offset>
+    The IV offset is a sector count that is added to the sector number
+    before creating the IV.
+
+<device path>
+    This is the device that is going to be used as backend and contains the
+    encrypted data.  You can specify it as a path like /dev/xxx or a device
+    number <major>:<minor>.
+
+<offset>
+    Starting sector within the device where the encrypted data begins.
+
+<#opt_params>
+    Number of optional parameters. If there are no optional parameters,
+    the optional paramaters section can be skipped or #opt_params can be zero.
+    Otherwise #opt_params is the number of following arguments.
+
+    Example of optional parameters section:
+        1 allow_discards
+
+allow_discards
+    Block discard requests (a.k.a. TRIM) are passed through the crypt device.
+    The default is to ignore discard requests.
+
+    WARNING: Assess the specific security risks carefully before enabling this
+    option.  For example, allowing discards on encrypted devices may lead to
+    the leak of information about the ciphertext device (filesystem type,
+    used space etc.) if the discarded blocks can be located easily on the
+    device later.
+
+Example scripts
+===============
+LUKS (Linux Unified Key Setup) is now the preferred way to set up disk
+encryption with dm-crypt using the 'cryptsetup' utility, see
+http://code.google.com/p/cryptsetup/
+
+[[
+#!/bin/sh
+# Create a crypt device using dmsetup
+dmsetup create crypt1 --table "0 `blockdev --getsize $1` crypt aes-cbc-essiv:sha256 babebabebabebabebabebabebabebabe 0 $1 0"
+]]
+
+[[
+#!/bin/sh
+# Create a crypt device using cryptsetup and LUKS header with default cipher
+cryptsetup luksFormat $1
+cryptsetup luksOpen $1 crypt1
+]]

doc/kernel/delay.txt

@@ -0,0 +1,26 @@
+dm-delay
+========
+
+Device-Mapper's "delay" target delays reads and/or writes
+and maps them to different devices.
+
+Parameters:
+    <device> <offset> <delay> [<write_device> <write_offset> <write_delay>]
+
+With separate write parameters, the first set is only used for reads.
+Delays are specified in milliseconds.
+
+Example scripts
+===============
+[[
+#!/bin/sh
+# Create device delaying rw operation for 500ms
+echo "0 `blockdev --getsize $1` delay $1 0 500" | dmsetup create delayed
+]]
+
+[[
+#!/bin/sh
+# Create device delaying only write operation for 500ms and
+# splitting reads and writes to different devices $1 $2
+echo "0 `blockdev --getsize $1` delay $1 0 0 $2 0 500" | dmsetup create delayed
+]]

doc/kernel/flakey.txt

@@ -0,0 +1,53 @@
+dm-flakey
+=========
+
+This target is the same as the linear target except that it exhibits
+unreliable behaviour periodically.  It's been found useful in simulating
+failing devices for testing purposes.
+
+Starting from the time the table is loaded, the device is available for
+<up interval> seconds, then exhibits unreliable behaviour for <down
+interval> seconds, and then this cycle repeats.
+
+Also, consider using this in combination with the dm-delay target too,
+which can delay reads and writes and/or send them to different
+underlying devices.
+
+Table parameters
+----------------
+  <dev path> <offset> <up interval> <down interval> \
+    [<num_features> [<feature arguments>]]
+
+Mandatory parameters:
+    <dev path>: Full pathname to the underlying block-device, or a
+                "major:minor" device-number.
+    <offset>: Starting sector within the device.
+    <up interval>: Number of seconds device is available.
+    <down interval>: Number of seconds device returns errors.
+
+Optional feature parameters:
+  If no feature parameters are present, during the periods of
+  unreliability, all I/O returns errors.
+
+  drop_writes:
+	All write I/O is silently ignored.
+	Read I/O is handled correctly.
+
+  corrupt_bio_byte <Nth_byte> <direction> <value> <flags>:
+	During <down interval>, replace <Nth_byte> of the data of
+	each matching bio with <value>.
+
+    <Nth_byte>: The offset of the byte to replace.
+		Counting starts at 1, to replace the first byte.
+    <direction>: Either 'r' to corrupt reads or 'w' to corrupt writes.
+		 'w' is incompatible with drop_writes.
+    <value>: The value (from 0-255) to write.
+    <flags>: Perform the replacement only if bio->bi_rw has all the
+	     selected flags set.
+
+Examples:
+  corrupt_bio_byte 32 r 1 0
+	- replaces the 32nd byte of READ bios with the value 1
+
+  corrupt_bio_byte 224 w 0 32
+	- replaces the 224th byte of REQ_META (=32) bios with the value 0

doc/kernel/io.txt

@@ -0,0 +1,75 @@
+dm-io
+=====
+
+Dm-io provides synchronous and asynchronous I/O services. There are three
+types of I/O services available, and each type has a sync and an async
+version.
+
+The user must set up an io_region structure to describe the desired location
+of the I/O. Each io_region indicates a block-device along with the starting
+sector and size of the region.
+
+   struct io_region {
+      struct block_device *bdev;
+      sector_t sector;
+      sector_t count;
+   };
+
+Dm-io can read from one io_region or write to one or more io_regions. Writes
+to multiple regions are specified by an array of io_region structures.
+
+The first I/O service type takes a list of memory pages as the data buffer for
+the I/O, along with an offset into the first page.
+
+   struct page_list {
+      struct page_list *next;
+      struct page *page;
+   };
+
+   int dm_io_sync(unsigned int num_regions, struct io_region *where, int rw,
+                  struct page_list *pl, unsigned int offset,
+                  unsigned long *error_bits);
+   int dm_io_async(unsigned int num_regions, struct io_region *where, int rw,
+                   struct page_list *pl, unsigned int offset,
+                   io_notify_fn fn, void *context);
+
+The second I/O service type takes an array of bio vectors as the data buffer
+for the I/O. This service can be handy if the caller has a pre-assembled bio,
+but wants to direct different portions of the bio to different devices.
+
+   int dm_io_sync_bvec(unsigned int num_regions, struct io_region *where,
+                       int rw, struct bio_vec *bvec,
+                       unsigned long *error_bits);
+   int dm_io_async_bvec(unsigned int num_regions, struct io_region *where,
+                        int rw, struct bio_vec *bvec,
+                        io_notify_fn fn, void *context);
+
+The third I/O service type takes a pointer to a vmalloc'd memory buffer as the
+data buffer for the I/O. This service can be handy if the caller needs to do
+I/O to a large region but doesn't want to allocate a large number of individual
+memory pages.
+
+   int dm_io_sync_vm(unsigned int num_regions, struct io_region *where, int rw,
+                     void *data, unsigned long *error_bits);
+   int dm_io_async_vm(unsigned int num_regions, struct io_region *where, int rw,
+                      void *data, io_notify_fn fn, void *context);
+
+Callers of the asynchronous I/O services must include the name of a completion
+callback routine and a pointer to some context data for the I/O.
+
+   typedef void (*io_notify_fn)(unsigned long error, void *context);
+
+The "error" parameter in this callback, as well as the "*error" parameter in
+all of the synchronous versions, is a bitset (instead of a simple error value).
+In the case of an write-I/O to multiple regions, this bitset allows dm-io to
+indicate success or failure on each individual region.
+
+Before using any of the dm-io services, the user should call dm_io_get()
+and specify the number of pages they expect to perform I/O on concurrently.
+Dm-io will attempt to resize its mempool to make sure enough pages are
+always available in order to avoid unnecessary waiting while performing I/O.
+
+When the user is finished using the dm-io services, they should call
+dm_io_put() and specify the same number of pages that were given on the
+dm_io_get() call.
+

doc/kernel/kcopyd.txt

@@ -0,0 +1,47 @@
+kcopyd
+======
+
+Kcopyd provides the ability to copy a range of sectors from one block-device
+to one or more other block-devices, with an asynchronous completion
+notification. It is used by dm-snapshot and dm-mirror.
+
+Users of kcopyd must first create a client and indicate how many memory pages
+to set aside for their copy jobs. This is done with a call to
+kcopyd_client_create().
+
+   int kcopyd_client_create(unsigned int num_pages,
+                            struct kcopyd_client **result);
+
+To start a copy job, the user must set up io_region structures to describe
+the source and destinations of the copy. Each io_region indicates a
+block-device along with the starting sector and size of the region. The source
+of the copy is given as one io_region structure, and the destinations of the
+copy are given as an array of io_region structures.
+
+   struct io_region {
+      struct block_device *bdev;
+      sector_t sector;
+      sector_t count;
+   };
+
+To start the copy, the user calls kcopyd_copy(), passing in the client
+pointer, pointers to the source and destination io_regions, the name of a
+completion callback routine, and a pointer to some context data for the copy.
+
+   int kcopyd_copy(struct kcopyd_client *kc, struct io_region *from,
+                   unsigned int num_dests, struct io_region *dests,
+                   unsigned int flags, kcopyd_notify_fn fn, void *context);
+
+   typedef void (*kcopyd_notify_fn)(int read_err, unsigned int write_err,
+				    void *context);
+
+When the copy completes, kcopyd will call the user's completion routine,
+passing back the user's context pointer. It will also indicate if a read or
+write error occurred during the copy.
+
+When a user is done with all their copy jobs, they should call
+kcopyd_client_destroy() to delete the kcopyd client, which will release the
+associated memory pages.
+
+   void kcopyd_client_destroy(struct kcopyd_client *kc);
+

doc/kernel/linear.txt

@@ -0,0 +1,61 @@
+dm-linear
+=========
+
+Device-Mapper's "linear" target maps a linear range of the Device-Mapper
+device onto a linear range of another device.  This is the basic building
+block of logical volume managers.
+
+Parameters: <dev path> <offset>
+    <dev path>: Full pathname to the underlying block-device, or a
+                "major:minor" device-number.
+    <offset>: Starting sector within the device.
+
+
+Example scripts
+===============
+[[
+#!/bin/sh
+# Create an identity mapping for a device
+echo "0 `blockdev --getsize $1` linear $1 0" | dmsetup create identity
+]]
+
+
+[[
+#!/bin/sh
+# Join 2 devices together
+size1=`blockdev --getsize $1`
+size2=`blockdev --getsize $2`
+echo "0 $size1 linear $1 0
+$size1 $size2 linear $2 0" | dmsetup create joined
+]]
+
+
+[[
+#!/usr/bin/perl -w
+# Split a device into 4M chunks and then join them together in reverse order.
+
+my $name = "reverse";
+my $extent_size = 4 * 1024 * 2;
+my $dev = $ARGV[0];
+my $table = "";
+my $count = 0;
+
+if (!defined($dev)) {
+        die("Please specify a device.\n");
+}
+
+my $dev_size = `blockdev --getsize $dev`;
+my $extents = int($dev_size / $extent_size) -
+              (($dev_size % $extent_size) ? 1 : 0);
+
+while ($extents > 0) {
+        my $this_start = $count * $extent_size;
+        $extents--;
+        $count++;
+        my $this_offset = $extents * $extent_size;
+
+        $table .= "$this_start $extent_size linear $dev $this_offset\n";
+}
+
+`echo \"$table\" | dmsetup create $name`;
+]]

doc/kernel/log.txt

@@ -0,0 +1,54 @@
+Device-Mapper Logging
+=====================
+The device-mapper logging code is used by some of the device-mapper
+RAID targets to track regions of the disk that are not consistent.
+A region (or portion of the address space) of the disk may be
+inconsistent because a RAID stripe is currently being operated on or
+a machine died while the region was being altered.  In the case of
+mirrors, a region would be considered dirty/inconsistent while you
+are writing to it because the writes need to be replicated for all
+the legs of the mirror and may not reach the legs at the same time.
+Once all writes are complete, the region is considered clean again.
+
+There is a generic logging interface that the device-mapper RAID
+implementations use to perform logging operations (see
+dm_dirty_log_type in include/linux/dm-dirty-log.h).  Various different
+logging implementations are available and provide different
+capabilities.  The list includes:
+
+Type		Files
+====		=====
+disk		drivers/md/dm-log.c
+core		drivers/md/dm-log.c
+userspace	drivers/md/dm-log-userspace* include/linux/dm-log-userspace.h
+
+The "disk" log type
+-------------------
+This log implementation commits the log state to disk.  This way, the
+logging state survives reboots/crashes.
+
+The "core" log type
+-------------------
+This log implementation keeps the log state in memory.  The log state
+will not survive a reboot or crash, but there may be a small boost in
+performance.  This method can also be used if no storage device is
+available for storing log state.
+
+The "userspace" log type
+------------------------
+This log type simply provides a way to export the log API to userspace,
+so log implementations can be done there.  This is done by forwarding most
+logging requests to userspace, where a daemon receives and processes the
+request.
+
+The structure used for communication between kernel and userspace are
+located in include/linux/dm-log-userspace.h.  Due to the frequency,
+diversity, and 2-way communication nature of the exchanges between
+kernel and userspace, 'connector' is used as the interface for
+communication.
+
+There are currently two userspace log implementations that leverage this
+framework - "clustered-disk" and "clustered-core".  These implementations
+provide a cluster-coherent log for shared-storage.  Device-mapper mirroring
+can be used in a shared-storage environment when the cluster log implementations
+are employed.

doc/kernel/persistent-data.txt

@@ -0,0 +1,84 @@
+Introduction
+============
+
+The more-sophisticated device-mapper targets require complex metadata
+that is managed in kernel.  In late 2010 we were seeing that various
+different targets were rolling their own data structures, for example:
+
+- Mikulas Patocka's multisnap implementation
+- Heinz Mauelshagen's thin provisioning target
+- Another btree-based caching target posted to dm-devel
+- Another multi-snapshot target based on a design of Daniel Phillips
+
+Maintaining these data structures takes a lot of work, so if possible
+we'd like to reduce the number.
+
+The persistent-data library is an attempt to provide a re-usable
+framework for people who want to store metadata in device-mapper
+targets.  It's currently used by the thin-provisioning target and an
+upcoming hierarchical storage target.
+
+Overview
+========
+
+The main documentation is in the header files which can all be found
+under drivers/md/persistent-data.
+
+The block manager
+-----------------
+
+dm-block-manager.[hc]
+
+This provides access to the data on disk in fixed sized-blocks.  There
+is a read/write locking interface to prevent concurrent accesses, and
+keep data that is being used in the cache.
+
+Clients of persistent-data are unlikely to use this directly.
+
+The transaction manager
+-----------------------
+
+dm-transaction-manager.[hc]
+
+This restricts access to blocks and enforces copy-on-write semantics.
+The only way you can get hold of a writable block through the
+transaction manager is by shadowing an existing block (ie. doing
+copy-on-write) or allocating a fresh one.  Shadowing is elided within
+the same transaction so performance is reasonable.  The commit method
+ensures that all data is flushed before it writes the superblock.
+On power failure your metadata will be as it was when last committed.
+
+The Space Maps
+--------------
+
+dm-space-map.h
+dm-space-map-metadata.[hc]
+dm-space-map-disk.[hc]
+
+On-disk data structures that keep track of reference counts of blocks.
+Also acts as the allocator of new blocks.  Currently two
+implementations: a simpler one for managing blocks on a different
+device (eg. thinly-provisioned data blocks); and one for managing
+the metadata space.  The latter is complicated by the need to store
+its own data within the space it's managing.
+
+The data structures
+-------------------
+
+dm-btree.[hc]
+dm-btree-remove.c
+dm-btree-spine.c
+dm-btree-internal.h
+
+Currently there is only one data structure, a hierarchical btree.
+There are plans to add more.  For example, something with an
+array-like interface would see a lot of use.
+
+The btree is 'hierarchical' in that you can define it to be composed
+of nested btrees, and take multiple keys.  For example, the
+thin-provisioning target uses a btree with two levels of nesting.
+The first maps a device id to a mapping tree, and that in turn maps a
+virtual block to a physical block.
+
+Values stored in the btrees can have arbitrary size.  Keys are always
+64bits, although nesting allows you to use multiple keys.

doc/kernel/queue-length.txt

@@ -0,0 +1,39 @@
+dm-queue-length
+===============
+
+dm-queue-length is a path selector module for device-mapper targets,
+which selects a path with the least number of in-flight I/Os.
+The path selector name is 'queue-length'.
+
+Table parameters for each path: [<repeat_count>]
+	<repeat_count>: The number of I/Os to dispatch using the selected
+			path before switching to the next path.
+			If not given, internal default is used. To check
+			the default value, see the activated table.
+
+Status for each path: <status> <fail-count> <in-flight>
+	<status>: 'A' if the path is active, 'F' if the path is failed.
+	<fail-count>: The number of path failures.
+	<in-flight>: The number of in-flight I/Os on the path.
+
+
+Algorithm
+=========
+
+dm-queue-length increments/decrements 'in-flight' when an I/O is
+dispatched/completed respectively.
+dm-queue-length selects a path with the minimum 'in-flight'.
+
+
+Examples
+========
+In case that 2 paths (sda and sdb) are used with repeat_count == 128.
+
+# echo "0 10 multipath 0 0 1 1 queue-length 0 2 1 8:0 128 8:16 128" \
+  dmsetup create test
+#
+# dmsetup table
+test: 0 10 multipath 0 0 1 1 queue-length 0 2 1 8:0 128 8:16 128
+#
+# dmsetup status
+test: 0 10 multipath 2 0 0 0 1 1 E 0 2 1 8:0 A 0 0 8:16 A 0 0

doc/kernel/raid.txt

@@ -0,0 +1,224 @@
+dm-raid
+=======
+
+The device-mapper RAID (dm-raid) target provides a bridge from DM to MD.
+It allows the MD RAID drivers to be accessed using a device-mapper
+interface.
+
+
+Mapping Table Interface
+-----------------------
+The target is named "raid" and it accepts the following parameters:
+
+  <raid_type> <#raid_params> <raid_params> \
+    <#raid_devs> <metadata_dev0> <dev0> [.. <metadata_devN> <devN>]
+
+<raid_type>:
+  raid1		RAID1 mirroring
+  raid4		RAID4 dedicated parity disk
+  raid5_la	RAID5 left asymmetric
+		- rotating parity 0 with data continuation
+  raid5_ra	RAID5 right asymmetric
+		- rotating parity N with data continuation
+  raid5_ls	RAID5 left symmetric
+		- rotating parity 0 with data restart
+  raid5_rs 	RAID5 right symmetric
+		- rotating parity N with data restart
+  raid6_zr	RAID6 zero restart
+		- rotating parity zero (left-to-right) with data restart
+  raid6_nr	RAID6 N restart
+		- rotating parity N (right-to-left) with data restart
+  raid6_nc	RAID6 N continue
+		- rotating parity N (right-to-left) with data continuation
+  raid10        Various RAID10 inspired algorithms chosen by additional params
+		- RAID10: Striped Mirrors (aka 'Striping on top of mirrors')
+		- RAID1E: Integrated Adjacent Stripe Mirroring
+		- RAID1E: Integrated Offset Stripe Mirroring
+		-  and other similar RAID10 variants
+
+  Reference: Chapter 4 of
+  http://www.snia.org/sites/default/files/SNIA_DDF_Technical_Position_v2.0.pdf
+
+<#raid_params>: The number of parameters that follow.
+
+<raid_params> consists of
+    Mandatory parameters:
+        <chunk_size>: Chunk size in sectors.  This parameter is often known as
+		      "stripe size".  It is the only mandatory parameter and
+		      is placed first.
+
+    followed by optional parameters (in any order):
+	[sync|nosync]   Force or prevent RAID initialization.
+
+	[rebuild <idx>]	Rebuild drive number 'idx' (first drive is 0).
+
+	[daemon_sleep <ms>]
+		Interval between runs of the bitmap daemon that
+		clear bits.  A longer interval means less bitmap I/O but
+		resyncing after a failure is likely to take longer.
+
+	[min_recovery_rate <kB/sec/disk>]  Throttle RAID initialization
+	[max_recovery_rate <kB/sec/disk>]  Throttle RAID initialization
+	[write_mostly <idx>]		   Mark drive index 'idx' write-mostly.
+	[max_write_behind <sectors>]       See '--write-behind=' (man mdadm)
+	[stripe_cache <sectors>]           Stripe cache size (RAID 4/5/6 only)
+	[region_size <sectors>]
+		The region_size multiplied by the number of regions is the
+		logical size of the array.  The bitmap records the device
+		synchronisation state for each region.
+
+        [raid10_copies   <# copies>]
+        [raid10_format   <near|far|offset>]
+		These two options are used to alter the default layout of
+		a RAID10 configuration.  The number of copies is can be
+		specified, but the default is 2.  There are also three
+		variations to how the copies are laid down - the default
+		is "near".  Near copies are what most people think of with
+		respect to mirroring.  If these options are left unspecified,
+		or 'raid10_copies 2' and/or 'raid10_format near' are given,
+		then the layouts for 2, 3 and 4 devices	are:
+		2 drives         3 drives          4 drives
+		--------         ----------        --------------
+		A1  A1           A1  A1  A2        A1  A1  A2  A2
+		A2  A2           A2  A3  A3        A3  A3  A4  A4
+		A3  A3           A4  A4  A5        A5  A5  A6  A6
+		A4  A4           A5  A6  A6        A7  A7  A8  A8
+		..  ..           ..  ..  ..        ..  ..  ..  ..
+		The 2-device layout is equivalent 2-way RAID1.  The 4-device
+		layout is what a traditional RAID10 would look like.  The
+		3-device layout is what might be called a 'RAID1E - Integrated
+		Adjacent Stripe Mirroring'.
+
+		If 'raid10_copies 2' and 'raid10_format far', then the layouts
+		for 2, 3 and 4 devices are:
+		2 drives             3 drives             4 drives
+		--------             --------------       --------------------
+		A1  A2               A1   A2   A3         A1   A2   A3   A4
+		A3  A4               A4   A5   A6         A5   A6   A7   A8
+		A5  A6               A7   A8   A9         A9   A10  A11  A12
+		..  ..               ..   ..   ..         ..   ..   ..   ..
+		A2  A1               A3   A1   A2         A2   A1   A4   A3
+		A4  A3               A6   A4   A5         A6   A5   A8   A7
+		A6  A5               A9   A7   A8         A10  A9   A12  A11
+		..  ..               ..   ..   ..         ..   ..   ..   ..
+
+		If 'raid10_copies 2' and 'raid10_format offset', then the
+		layouts for 2, 3 and 4 devices are:
+		2 drives       3 drives           4 drives
+		--------       ------------       -----------------
+		A1  A2         A1  A2  A3         A1  A2  A3  A4
+		A2  A1         A3  A1  A2         A2  A1  A4  A3
+		A3  A4         A4  A5  A6         A5  A6  A7  A8
+		A4  A3         A6  A4  A5         A6  A5  A8  A7
+		A5  A6         A7  A8  A9         A9  A10 A11 A12
+		A6  A5         A9  A7  A8         A10 A9  A12 A11
+		..  ..         ..  ..  ..         ..  ..  ..  ..
+		Here we see layouts closely akin to 'RAID1E - Integrated
+		Offset Stripe Mirroring'.
+
+<#raid_devs>: The number of devices composing the array.
+	Each device consists of two entries.  The first is the device
+	containing the metadata (if any); the second is the one containing the
+	data.
+
+	If a drive has failed or is missing at creation time, a '-' can be
+	given for both the metadata and data drives for a given position.
+
+
+Example Tables
+--------------
+# RAID4 - 4 data drives, 1 parity (no metadata devices)
+# No metadata devices specified to hold superblock/bitmap info
+# Chunk size of 1MiB
+# (Lines separated for easy reading)
+
+0 1960893648 raid \
+        raid4 1 2048 \
+        5 - 8:17 - 8:33 - 8:49 - 8:65 - 8:81
+
+# RAID4 - 4 data drives, 1 parity (with metadata devices)
+# Chunk size of 1MiB, force RAID initialization,
+#       min recovery rate at 20 kiB/sec/disk
+
+0 1960893648 raid \
+        raid4 4 2048 sync min_recovery_rate 20 \
+        5 8:17 8:18 8:33 8:34 8:49 8:50 8:65 8:66 8:81 8:82
+
+
+Status Output
+-------------
+'dmsetup table' displays the table used to construct the mapping.
+The optional parameters are always printed in the order listed
+above with "sync" or "nosync" always output ahead of the other
+arguments, regardless of the order used when originally loading the table.
+Arguments that can be repeated are ordered by value.
+
+
+'dmsetup status' yields information on the state and health of the array.
+The output is as follows (normally a single line, but expanded here for
+clarity):
+1: <s> <l> raid \
+2:      <raid_type> <#devices> <health_chars> \
+3:      <sync_ratio> <sync_action> <mismatch_cnt>
+
+Line 1 is the standard output produced by device-mapper.
+Line 2 & 3 are produced by the raid target and are best explained by example:
+        0 1960893648 raid raid4 5 AAAAA 2/490221568 init 0
+Here we can see the RAID type is raid4, there are 5 devices - all of
+which are 'A'live, and the array is 2/490221568 complete with its initial
+recovery.  Here is a fuller description of the individual fields:
+	<raid_type>     Same as the <raid_type> used to create the array.
+	<health_chars>  One char for each device, indicating: 'A' = alive and
+			in-sync, 'a' = alive but not in-sync, 'D' = dead/failed.
+	<sync_ratio>    The ratio indicating how much of the array has undergone
+			the process described by 'sync_action'.  If the
+			'sync_action' is "check" or "repair", then the process
+			of "resync" or "recover" can be considered complete.
+	<sync_action>   One of the following possible states:
+			idle    - No synchronization action is being performed.
+			frozen  - The current action has been halted.
+			resync  - Array is undergoing its initial synchronization
+				  or is resynchronizing after an unclean shutdown
+				  (possibly aided by a bitmap).
+			recover - A device in the array is being rebuilt or
+				  replaced.
+			check   - A user-initiated full check of the array is
+				  being performed.  All blocks are read and
+				  checked for consistency.  The number of
+				  discrepancies found are recorded in
+				  <mismatch_cnt>.  No changes are made to the
+				  array by this action.
+			repair  - The same as "check", but discrepancies are
+				  corrected.
+			reshape - The array is undergoing a reshape.
+	<mismatch_cnt>  The number of discrepancies found between mirror copies
+			in RAID1/10 or wrong parity values found in RAID4/5/6.
+			This value is valid only after a "check" of the array
+			is performed.  A healthy array has a 'mismatch_cnt' of 0.
+
+Message Interface
+-----------------
+The dm-raid target will accept certain actions through the 'message' interface.
+('man dmsetup' for more information on the message interface.)  These actions
+include:
+	"idle"   - Halt the current sync action.
+	"frozen" - Freeze the current sync action.
+	"resync" - Initiate/continue a resync.
+	"recover"- Initiate/continue a recover process.
+	"check"  - Initiate a check (i.e. a "scrub") of the array.
+	"repair" - Initiate a repair of the array.
+	"reshape"- Currently unsupported (-EINVAL).
+
+Version History
+---------------
+1.0.0	Initial version.  Support for RAID 4/5/6
+1.1.0	Added support for RAID 1
+1.2.0	Handle creation of arrays that contain failed devices.
+1.3.0	Added support for RAID 10
+1.3.1	Allow device replacement/rebuild for RAID 10
+1.3.2   Fix/improve redundancy checking for RAID10
+1.4.0	Non-functional change.  Removes arg from mapping function.
+1.4.1   RAID10 fix redundancy validation checks (commit 55ebbb5).
+1.4.2   Add RAID10 "far" and "offset" algorithm support.
+1.5.0   Add message interface to allow manipulation of the sync_action.
+	New status (STATUSTYPE_INFO) fields: sync_action and mismatch_cnt.

doc/kernel/service-time.txt

@@ -0,0 +1,91 @@
+dm-service-time
+===============
+
+dm-service-time is a path selector module for device-mapper targets,
+which selects a path with the shortest estimated service time for
+the incoming I/O.
+
+The service time for each path is estimated by dividing the total size
+of in-flight I/Os on a path with the performance value of the path.
+The performance value is a relative throughput value among all paths
+in a path-group, and it can be specified as a table argument.
+
+The path selector name is 'service-time'.
+
+Table parameters for each path: [<repeat_count> [<relative_throughput>]]
+	<repeat_count>: The number of I/Os to dispatch using the selected
+			path before switching to the next path.
+			If not given, internal default is used.  To check
+			the default value, see the activated table.
+	<relative_throughput>: The relative throughput value of the path
+			among all paths in the path-group.
+			The valid range is 0-100.
+			If not given, minimum value '1' is used.
+			If '0' is given, the path isn't selected while
+			other paths having a positive value are available.
+
+Status for each path: <status> <fail-count> <in-flight-size> \
+		      <relative_throughput>
+	<status>: 'A' if the path is active, 'F' if the path is failed.
+	<fail-count>: The number of path failures.
+	<in-flight-size>: The size of in-flight I/Os on the path.
+	<relative_throughput>: The relative throughput value of the path
+			among all paths in the path-group.
+
+
+Algorithm
+=========
+
+dm-service-time adds the I/O size to 'in-flight-size' when the I/O is
+dispatched and subtracts when completed.
+Basically, dm-service-time selects a path having minimum service time
+which is calculated by:
+
+	('in-flight-size' + 'size-of-incoming-io') / 'relative_throughput'
+
+However, some optimizations below are used to reduce the calculation
+as much as possible.
+
+	1. If the paths have the same 'relative_throughput', skip
+	   the division and just compare the 'in-flight-size'.
+
+	2. If the paths have the same 'in-flight-size', skip the division
+	   and just compare the 'relative_throughput'.
+
+	3. If some paths have non-zero 'relative_throughput' and others
+	   have zero 'relative_throughput', ignore those paths with zero
+	   'relative_throughput'.
+
+If such optimizations can't be applied, calculate service time, and
+compare service time.
+If calculated service time is equal, the path having maximum
+'relative_throughput' may be better.  So compare 'relative_throughput'
+then.
+
+
+Examples
+========
+In case that 2 paths (sda and sdb) are used with repeat_count == 128
+and sda has an average throughput 1GB/s and sdb has 4GB/s,
+'relative_throughput' value may be '1' for sda and '4' for sdb.
+
+# echo "0 10 multipath 0 0 1 1 service-time 0 2 2 8:0 128 1 8:16 128 4" \
+  dmsetup create test
+#
+# dmsetup table
+test: 0 10 multipath 0 0 1 1 service-time 0 2 2 8:0 128 1 8:16 128 4
+#
+# dmsetup status
+test: 0 10 multipath 2 0 0 0 1 1 E 0 2 2 8:0 A 0 0 1 8:16 A 0 0 4
+
+
+Or '2' for sda and '8' for sdb would be also true.
+
+# echo "0 10 multipath 0 0 1 1 service-time 0 2 2 8:0 128 2 8:16 128 8" \
+  dmsetup create test
+#
+# dmsetup table
+test: 0 10 multipath 0 0 1 1 service-time 0 2 2 8:0 128 2 8:16 128 8
+#
+# dmsetup status
+test: 0 10 multipath 2 0 0 0 1 1 E 0 2 2 8:0 A 0 0 2 8:16 A 0 0 8

doc/kernel/snapshot.txt

@@ -0,0 +1,168 @@
+Device-mapper snapshot support
+==============================
+
+Device-mapper allows you, without massive data copying:
+
+*) To create snapshots of any block device i.e. mountable, saved states of
+the block device which are also writable without interfering with the
+original content;
+*) To create device "forks", i.e. multiple different versions of the
+same data stream.
+*) To merge a snapshot of a block device back into the snapshot's origin
+device.
+
+In the first two cases, dm copies only the chunks of data that get
+changed and uses a separate copy-on-write (COW) block device for
+storage.
+
+For snapshot merge the contents of the COW storage are merged back into
+the origin device.
+
+
+There are three dm targets available:
+snapshot, snapshot-origin, and snapshot-merge.
+
+*) snapshot-origin <origin>
+
+which will normally have one or more snapshots based on it.
+Reads will be mapped directly to the backing device. For each write, the
+original data will be saved in the <COW device> of each snapshot to keep
+its visible content unchanged, at least until the <COW device> fills up.
+
+
+*) snapshot <origin> <COW device> <persistent?> <chunksize>
+
+A snapshot of the <origin> block device is created. Changed chunks of
+<chunksize> sectors will be stored on the <COW device>.  Writes will
+only go to the <COW device>.  Reads will come from the <COW device> or
+from <origin> for unchanged data.  <COW device> will often be
+smaller than the origin and if it fills up the snapshot will become
+useless and be disabled, returning errors.  So it is important to monitor
+the amount of free space and expand the <COW device> before it fills up.
+
+<persistent?> is P (Persistent) or N (Not persistent - will not survive
+after reboot).
+The difference is that for transient snapshots less metadata must be
+saved on disk - they can be kept in memory by the kernel.
+
+
+* snapshot-merge <origin> <COW device> <persistent> <chunksize>
+
+takes the same table arguments as the snapshot target except it only
+works with persistent snapshots.  This target assumes the role of the
+"snapshot-origin" target and must not be loaded if the "snapshot-origin"
+is still present for <origin>.
+
+Creates a merging snapshot that takes control of the changed chunks
+stored in the <COW device> of an existing snapshot, through a handover
+procedure, and merges these chunks back into the <origin>.  Once merging
+has started (in the background) the <origin> may be opened and the merge
+will continue while I/O is flowing to it.  Changes to the <origin> are
+deferred until the merging snapshot's corresponding chunk(s) have been
+merged.  Once merging has started the snapshot device, associated with
+the "snapshot" target, will return -EIO when accessed.
+
+
+How snapshot is used by LVM2
+============================
+When you create the first LVM2 snapshot of a volume, four dm devices are used:
+
+1) a device containing the original mapping table of the source volume;
+2) a device used as the <COW device>;
+3) a "snapshot" device, combining #1 and #2, which is the visible snapshot
+   volume;
+4) the "original" volume (which uses the device number used by the original
+   source volume), whose table is replaced by a "snapshot-origin" mapping
+   from device #1.
+
+A fixed naming scheme is used, so with the following commands:
+
+lvcreate -L 1G -n base volumeGroup
+lvcreate -L 100M --snapshot -n snap volumeGroup/base
+
+we'll have this situation (with volumes in above order):
+
+# dmsetup table|grep volumeGroup
+
+volumeGroup-base-real: 0 2097152 linear 8:19 384
+volumeGroup-snap-cow: 0 204800 linear 8:19 2097536
+volumeGroup-snap: 0 2097152 snapshot 254:11 254:12 P 16
+volumeGroup-base: 0 2097152 snapshot-origin 254:11
+
+# ls -lL /dev/mapper/volumeGroup-*
+brw-------  1 root root 254, 11 29 ago 18:15 /dev/mapper/volumeGroup-base-real
+brw-------  1 root root 254, 12 29 ago 18:15 /dev/mapper/volumeGroup-snap-cow
+brw-------  1 root root 254, 13 29 ago 18:15 /dev/mapper/volumeGroup-snap
+brw-------  1 root root 254, 10 29 ago 18:14 /dev/mapper/volumeGroup-base
+
+
+How snapshot-merge is used by LVM2
+==================================
+A merging snapshot assumes the role of the "snapshot-origin" while
+merging.  As such the "snapshot-origin" is replaced with
+"snapshot-merge".  The "-real" device is not changed and the "-cow"
+device is renamed to <origin name>-cow to aid LVM2's cleanup of the
+merging snapshot after it completes.  The "snapshot" that hands over its
+COW device to the "snapshot-merge" is deactivated (unless using lvchange
+--refresh); but if it is left active it will simply return I/O errors.
+
+A snapshot will merge into its origin with the following command:
+
+lvconvert --merge volumeGroup/snap
+
+we'll now have this situation:
+
+# dmsetup table|grep volumeGroup
+
+volumeGroup-base-real: 0 2097152 linear 8:19 384
+volumeGroup-base-cow: 0 204800 linear 8:19 2097536
+volumeGroup-base: 0 2097152 snapshot-merge 254:11 254:12 P 16
+
+# ls -lL /dev/mapper/volumeGroup-*
+brw-------  1 root root 254, 11 29 ago 18:15 /dev/mapper/volumeGroup-base-real
+brw-------  1 root root 254, 12 29 ago 18:16 /dev/mapper/volumeGroup-base-cow
+brw-------  1 root root 254, 10 29 ago 18:16 /dev/mapper/volumeGroup-base
+
+
+How to determine when a merging is complete
+===========================================
+The snapshot-merge and snapshot status lines end with:
+  <sectors_allocated>/<total_sectors> <metadata_sectors>
+
+Both <sectors_allocated> and <total_sectors> include both data and metadata.
+During merging, the number of sectors allocated gets smaller and
+smaller.  Merging has finished when the number of sectors holding data
+is zero, in other words <sectors_allocated> == <metadata_sectors>.
+
+Here is a practical example (using a hybrid of lvm and dmsetup commands):
+
+# lvs
+  LV      VG          Attr   LSize Origin  Snap%  Move Log Copy%  Convert
+  base    volumeGroup owi-a- 4.00g
+  snap    volumeGroup swi-a- 1.00g base  18.97
+
+# dmsetup status volumeGroup-snap
+0 8388608 snapshot 397896/2097152 1560
+                                  ^^^^ metadata sectors
+
+# lvconvert --merge -b volumeGroup/snap
+  Merging of volume snap started.
+
+# lvs volumeGroup/snap
+  LV      VG          Attr   LSize Origin  Snap%  Move Log Copy%  Convert
+  base    volumeGroup Owi-a- 4.00g          17.23
+
+# dmsetup status volumeGroup-base
+0 8388608 snapshot-merge 281688/2097152 1104
+
+# dmsetup status volumeGroup-base
+0 8388608 snapshot-merge 180480/2097152 712
+
+# dmsetup status volumeGroup-base
+0 8388608 snapshot-merge 16/2097152 16
+
+Merging has finished.
+
+# lvs
+  LV      VG          Attr   LSize Origin  Snap%  Move Log Copy%  Convert
+  base    volumeGroup owi-a- 4.00g

doc/kernel/striped.txt

@@ -0,0 +1,57 @@
+dm-stripe
+=========
+
+Device-Mapper's "striped" target is used to create a striped (i.e. RAID-0)
+device across one or more underlying devices. Data is written in "chunks",
+with consecutive chunks rotating among the underlying devices. This can
+potentially provide improved I/O throughput by utilizing several physical
+devices in parallel.
+
+Parameters: <num devs> <chunk size> [<dev path> <offset>]+
+    <num devs>: Number of underlying devices.
+    <chunk size>: Size of each chunk of data. Must be at least as
+                  large as the system's PAGE_SIZE.
+    <dev path>: Full pathname to the underlying block-device, or a
+                "major:minor" device-number.
+    <offset>: Starting sector within the device.
+
+One or more underlying devices can be specified. The striped device size must
+be a multiple of the chunk size multiplied by the number of underlying devices.
+
+
+Example scripts
+===============
+
+[[
+#!/usr/bin/perl -w
+# Create a striped device across any number of underlying devices. The device
+# will be called "stripe_dev" and have a chunk-size of 128k.
+
+my $chunk_size = 128 * 2;
+my $dev_name = "stripe_dev";
+my $num_devs = @ARGV;
+my @devs = @ARGV;
+my ($min_dev_size, $stripe_dev_size, $i);
+
+if (!$num_devs) {
+        die("Specify at least one device\n");
+}
+
+$min_dev_size = `blockdev --getsize $devs[0]`;
+for ($i = 1; $i < $num_devs; $i++) {
+        my $this_size = `blockdev --getsize $devs[$i]`;
+        $min_dev_size = ($min_dev_size < $this_size) ?
+                        $min_dev_size : $this_size;
+}
+
+$stripe_dev_size = $min_dev_size * $num_devs;
+$stripe_dev_size -= $stripe_dev_size % ($chunk_size * $num_devs);
+
+$table = "0 $stripe_dev_size striped $num_devs $chunk_size";
+for ($i = 0; $i < $num_devs; $i++) {
+        $table .= " $devs[$i] 0";
+}
+
+`echo $table | dmsetup create $dev_name`;
+]]
+

doc/kernel/thin-provisioning.txt

@@ -0,0 +1,353 @@
+Introduction
+============
+
+This document describes a collection of device-mapper targets that
+between them implement thin-provisioning and snapshots.
+
+The main highlight of this implementation, compared to the previous
+implementation of snapshots, is that it allows many virtual devices to
+be stored on the same data volume.  This simplifies administration and
+allows the sharing of data between volumes, thus reducing disk usage.
+
+Another significant feature is support for an arbitrary depth of
+recursive snapshots (snapshots of snapshots of snapshots ...).  The
+previous implementation of snapshots did this by chaining together
+lookup tables, and so performance was O(depth).  This new
+implementation uses a single data structure to avoid this degradation
+with depth.  Fragmentation may still be an issue, however, in some
+scenarios.
+
+Metadata is stored on a separate device from data, giving the
+administrator some freedom, for example to:
+
+- Improve metadata resilience by storing metadata on a mirrored volume
+  but data on a non-mirrored one.
+
+- Improve performance by storing the metadata on SSD.
+
+Status
+======
+
+These targets are very much still in the EXPERIMENTAL state.  Please
+do not yet rely on them in production.  But do experiment and offer us
+feedback.  Different use cases will have different performance
+characteristics, for example due to fragmentation of the data volume.
+
+If you find this software is not performing as expected please mail
+dm-devel@redhat.com with details and we'll try our best to improve
+things for you.
+
+Userspace tools for checking and repairing the metadata are under
+development.
+
+Cookbook
+========
+
+This section describes some quick recipes for using thin provisioning.
+They use the dmsetup program to control the device-mapper driver
+directly.  End users will be advised to use a higher-level volume
+manager such as LVM2 once support has been added.
+
+Pool device
+-----------
+
+The pool device ties together the metadata volume and the data volume.
+It maps I/O linearly to the data volume and updates the metadata via
+two mechanisms:
+
+- Function calls from the thin targets
+
+- Device-mapper 'messages' from userspace which control the creation of new
+  virtual devices amongst other things.
+
+Setting up a fresh pool device
+------------------------------
+
+Setting up a pool device requires a valid metadata device, and a
+data device.  If you do not have an existing metadata device you can
+make one by zeroing the first 4k to indicate empty metadata.
+
+    dd if=/dev/zero of=$metadata_dev bs=4096 count=1
+
+The amount of metadata you need will vary according to how many blocks
+are shared between thin devices (i.e. through snapshots).  If you have
+less sharing than average you'll need a larger-than-average metadata device.
+
+As a guide, we suggest you calculate the number of bytes to use in the
+metadata device as 48 * $data_dev_size / $data_block_size but round it up
+to 2MB if the answer is smaller.  If you're creating large numbers of
+snapshots which are recording large amounts of change, you may find you
+need to increase this.
+
+The largest size supported is 16GB: If the device is larger,
+a warning will be issued and the excess space will not be used.
+
+Reloading a pool table
+----------------------
+
+You may reload a pool's table, indeed this is how the pool is resized
+if it runs out of space.  (N.B. While specifying a different metadata
+device when reloading is not forbidden at the moment, things will go
+wrong if it does not route I/O to exactly the same on-disk location as
+previously.)
+
+Using an existing pool device
+-----------------------------
+
+    dmsetup create pool \
+	--table "0 20971520 thin-pool $metadata_dev $data_dev \
+		 $data_block_size $low_water_mark"
+
+$data_block_size gives the smallest unit of disk space that can be
+allocated at a time expressed in units of 512-byte sectors.  People
+primarily interested in thin provisioning may want to use a value such
+as 1024 (512KB).  People doing lots of snapshotting may want a smaller value
+such as 128 (64KB).  If you are not zeroing newly-allocated data,
+a larger $data_block_size in the region of 256000 (128MB) is suggested.
+$data_block_size must be the same for the lifetime of the
+metadata device.
+
+$low_water_mark is expressed in blocks of size $data_block_size.  If
+free space on the data device drops below this level then a dm event
+will be triggered which a userspace daemon should catch allowing it to
+extend the pool device.  Only one such event will be sent.
+Resuming a device with a new table itself triggers an event so the
+userspace daemon can use this to detect a situation where a new table
+already exceeds the threshold.
+
+Thin provisioning
+-----------------
+
+i) Creating a new thinly-provisioned volume.
+
+  To create a new thinly- provisioned volume you must send a message to an
+  active pool device, /dev/mapper/pool in this example.
+
+    dmsetup message /dev/mapper/pool 0 "create_thin 0"
+
+  Here '0' is an identifier for the volume, a 24-bit number.  It's up
+  to the caller to allocate and manage these identifiers.  If the
+  identifier is already in use, the message will fail with -EEXIST.
+
+ii) Using a thinly-provisioned volume.
+
+  Thinly-provisioned volumes are activated using the 'thin' target:
+
+    dmsetup create thin --table "0 2097152 thin /dev/mapper/pool 0"
+
+  The last parameter is the identifier for the thinp device.
+
+Internal snapshots
+------------------
+
+i) Creating an internal snapshot.
+
+  Snapshots are created with another message to the pool.
+
+  N.B.  If the origin device that you wish to snapshot is active, you
+  must suspend it before creating the snapshot to avoid corruption.
+  This is NOT enforced at the moment, so please be careful!
+
+    dmsetup suspend /dev/mapper/thin
+    dmsetup message /dev/mapper/pool 0 "create_snap 1 0"
+    dmsetup resume /dev/mapper/thin
+
+  Here '1' is the identifier for the volume, a 24-bit number.  '0' is the
+  identifier for the origin device.
+
+ii) Using an internal snapshot.
+
+  Once created, the user doesn't have to worry about any connection
+  between the origin and the snapshot.  Indeed the snapshot is no
+  different from any other thinly-provisioned device and can be
+  snapshotted itself via the same method.  It's perfectly legal to
+  have only one of them active, and there's no ordering requirement on
+  activating or removing them both.  (This differs from conventional
+  device-mapper snapshots.)
+
+  Activate it exactly the same way as any other thinly-provisioned volume:
+
+    dmsetup create snap --table "0 2097152 thin /dev/mapper/pool 1"
+
+External snapshots
+------------------
+
+You can use an external _read only_ device as an origin for a
+thinly-provisioned volume.  Any read to an unprovisioned area of the
+thin device will be passed through to the origin.  Writes trigger
+the allocation of new blocks as usual.
+
+One use case for this is VM hosts that want to run guests on
+thinly-provisioned volumes but have the base image on another device
+(possibly shared between many VMs).
+
+You must not write to the origin device if you use this technique!
+Of course, you may write to the thin device and take internal snapshots
+of the thin volume.
+
+i) Creating a snapshot of an external device
+
+  This is the same as creating a thin device.
+  You don't mention the origin at this stage.
+
+    dmsetup message /dev/mapper/pool 0 "create_thin 0"
+
+ii) Using a snapshot of an external device.
+
+  Append an extra parameter to the thin target specifying the origin:
+
+    dmsetup create snap --table "0 2097152 thin /dev/mapper/pool 0 /dev/image"
+
+  N.B. All descendants (internal snapshots) of this snapshot require the
+  same extra origin parameter.
+
+Deactivation
+------------
+
+All devices using a pool must be deactivated before the pool itself
+can be.
+
+    dmsetup remove thin
+    dmsetup remove snap
+    dmsetup remove pool
+
+Reference
+=========
+
+'thin-pool' target
+------------------
+
+i) Constructor
+
+    thin-pool <metadata dev> <data dev> <data block size (sectors)> \
+	      <low water mark (blocks)> [<number of feature args> [<arg>]*]
+
+    Optional feature arguments:
+
+      skip_block_zeroing: Skip the zeroing of newly-provisioned blocks.
+
+      ignore_discard: Disable discard support.
+
+      no_discard_passdown: Don't pass discards down to the underlying
+			   data device, but just remove the mapping.
+
+      read_only: Don't allow any changes to be made to the pool
+		 metadata.
+
+    Data block size must be between 64KB (128 sectors) and 1GB
+    (2097152 sectors) inclusive.
+
+
+ii) Status
+
+    <transaction id> <used metadata blocks>/<total metadata blocks>
+    <used data blocks>/<total data blocks> <held metadata root>
+    [no_]discard_passdown ro|rw
+
+    transaction id:
+	A 64-bit number used by userspace to help synchronise with metadata
+	from volume managers.
+
+    used data blocks / total data blocks
+	If the number of free blocks drops below the pool's low water mark a
+	dm event will be sent to userspace.  This event is edge-triggered and
+	it will occur only once after each resume so volume manager writers
+	should register for the event and then check the target's status.
+
+    held metadata root:
+	The location, in sectors, of the metadata root that has been
+	'held' for userspace read access.  '-' indicates there is no
+	held root.  This feature is not yet implemented so '-' is
+	always returned.
+
+    discard_passdown|no_discard_passdown
+	Whether or not discards are actually being passed down to the
+	underlying device.  When this is enabled when loading the table,
+	it can get disabled if the underlying device doesn't support it.
+
+    ro|rw
+	If the pool encounters certain types of device failures it will
+	drop into a read-only metadata mode in which no changes to
+	the pool metadata (like allocating new blocks) are permitted.
+
+	In serious cases where even a read-only mode is deemed unsafe
+	no further I/O will be permitted and the status will just
+	contain the string 'Fail'.  The userspace recovery tools
+	should then be used.
+
+iii) Messages
+
+    create_thin <dev id>
+
+	Create a new thinly-provisioned device.
+	<dev id> is an arbitrary unique 24-bit identifier chosen by
+	the caller.
+
+    create_snap <dev id> <origin id>
+
+	Create a new snapshot of another thinly-provisioned device.
+	<dev id> is an arbitrary unique 24-bit identifier chosen by
+	the caller.
+	<origin id> is the identifier of the thinly-provisioned device
+	of which the new device will be a snapshot.
+
+    delete <dev id>
+
+	Deletes a thin device.  Irreversible.
+
+    set_transaction_id <current id> <new id>
+
+	Userland volume managers, such as LVM, need a way to
+	synchronise their external metadata with the internal metadata of the
+	pool target.  The thin-pool target offers to store an
+	arbitrary 64-bit transaction id and return it on the target's
+	status line.  To avoid races you must provide what you think
+	the current transaction id is when you change it with this
+	compare-and-swap message.
+
+    reserve_metadata_snap
+
+        Reserve a copy of the data mapping btree for use by userland.
+        This allows userland to inspect the mappings as they were when
+        this message was executed.  Use the pool's status command to
+        get the root block associated with the metadata snapshot.
+
+    release_metadata_snap
+
+        Release a previously reserved copy of the data mapping btree.
+
+'thin' target
+-------------
+
+i) Constructor
+
+    thin <pool dev> <dev id> [<external origin dev>]
+
+    pool dev:
+	the thin-pool device, e.g. /dev/mapper/my_pool or 253:0
+
+    dev id:
+	the internal device identifier of the device to be
+	activated.
+
+    external origin dev:
+	an optional block device outside the pool to be treated as a
+	read-only snapshot origin: reads to unprovisioned areas of the
+	thin target will be mapped to this device.
+
+The pool doesn't store any size against the thin devices.  If you
+load a thin target that is smaller than you've been using previously,
+then you'll have no access to blocks mapped beyond the end.  If you
+load a target that is bigger than before, then extra blocks will be
+provisioned as and when needed.
+
+If you wish to reduce the size of your thin device and potentially
+regain some space then send the 'trim' message to the pool.
+
+ii) Status
+
+     <nr mapped sectors> <highest mapped sector>
+
+	If the pool has encountered device errors and failed, the status
+	will just contain the string 'Fail'.  The userspace recovery
+	tools should then be used.

doc/kernel/uevent.txt

@@ -0,0 +1,97 @@
+The device-mapper uevent code adds the capability to device-mapper to create
+and send kobject uevents (uevents).  Previously device-mapper events were only
+available through the ioctl interface.  The advantage of the uevents interface
+is the event contains environment attributes providing increased context for
+the event avoiding the need to query the state of the device-mapper device after
+the event is received.
+
+There are two functions currently for device-mapper events.  The first function
+listed creates the event and the second function sends the event(s).
+
+void dm_path_uevent(enum dm_uevent_type event_type, struct dm_target *ti,
+                    const char *path, unsigned nr_valid_paths)
+
+void dm_send_uevents(struct list_head *events, struct kobject *kobj)
+
+
+The variables added to the uevent environment are:
+
+Variable Name: DM_TARGET
+Uevent Action(s): KOBJ_CHANGE
+Type: string
+Description:
+Value: Name of device-mapper target that generated the event.
+
+Variable Name: DM_ACTION
+Uevent Action(s): KOBJ_CHANGE
+Type: string
+Description:
+Value: Device-mapper specific action that caused the uevent action.
+	PATH_FAILED - A path has failed.
+	PATH_REINSTATED - A path has been reinstated.
+
+Variable Name: DM_SEQNUM
+Uevent Action(s): KOBJ_CHANGE
+Type: unsigned integer
+Description: A sequence number for this specific device-mapper device.
+Value: Valid unsigned integer range.
+
+Variable Name: DM_PATH
+Uevent Action(s): KOBJ_CHANGE
+Type: string
+Description: Major and minor number of the path device pertaining to this
+event.
+Value: Path name in the form of "Major:Minor"
+
+Variable Name: DM_NR_VALID_PATHS
+Uevent Action(s): KOBJ_CHANGE
+Type: unsigned integer
+Description:
+Value: Valid unsigned integer range.
+
+Variable Name: DM_NAME
+Uevent Action(s): KOBJ_CHANGE
+Type: string
+Description: Name of the device-mapper device.
+Value: Name
+
+Variable Name: DM_UUID
+Uevent Action(s): KOBJ_CHANGE
+Type: string
+Description: UUID of the device-mapper device.
+Value: UUID. (Empty string if there isn't one.)
+
+An example of the uevents generated as captured by udevmonitor is shown
+below.
+
+1.) Path failure.
+UEVENT[1192521009.711215] change@/block/dm-3
+ACTION=change
+DEVPATH=/block/dm-3
+SUBSYSTEM=block
+DM_TARGET=multipath
+DM_ACTION=PATH_FAILED
+DM_SEQNUM=1
+DM_PATH=8:32
+DM_NR_VALID_PATHS=0
+DM_NAME=mpath2
+DM_UUID=mpath-35333333000002328
+MINOR=3
+MAJOR=253
+SEQNUM=1130
+
+2.) Path reinstate.
+UEVENT[1192521132.989927] change@/block/dm-3
+ACTION=change
+DEVPATH=/block/dm-3
+SUBSYSTEM=block
+DM_TARGET=multipath
+DM_ACTION=PATH_REINSTATED
+DM_SEQNUM=2
+DM_PATH=8:32
+DM_NR_VALID_PATHS=1
+DM_NAME=mpath2
+DM_UUID=mpath-35333333000002328
+MINOR=3
+MAJOR=253
+SEQNUM=1131

doc/kernel/verity.txt

@@ -0,0 +1,155 @@
+dm-verity
+==========
+
+Device-Mapper's "verity" target provides transparent integrity checking of
+block devices using a cryptographic digest provided by the kernel crypto API.
+This target is read-only.
+
+Construction Parameters
+=======================
+    <version> <dev> <hash_dev>
+    <data_block_size> <hash_block_size>
+    <num_data_blocks> <hash_start_block>
+    <algorithm> <digest> <salt>
+
+<version>
+    This is the type of the on-disk hash format.
+
+    0 is the original format used in the Chromium OS.
+      The salt is appended when hashing, digests are stored continuously and
+      the rest of the block is padded with zeros.
+
+    1 is the current format that should be used for new devices.
+      The salt is prepended when hashing and each digest is
+      padded with zeros to the power of two.
+
+<dev>
+    This is the device containing data, the integrity of which needs to be
+    checked.  It may be specified as a path, like /dev/sdaX, or a device number,
+    <major>:<minor>.
+
+<hash_dev>
+    This is the device that supplies the hash tree data.  It may be
+    specified similarly to the device path and may be the same device.  If the
+    same device is used, the hash_start should be outside the configured
+    dm-verity device.
+
+<data_block_size>
+    The block size on a data device in bytes.
+    Each block corresponds to one digest on the hash device.
+
+<hash_block_size>
+    The size of a hash block in bytes.
+
+<num_data_blocks>
+    The number of data blocks on the data device.  Additional blocks are
+    inaccessible.  You can place hashes to the same partition as data, in this
+    case hashes are placed after <num_data_blocks>.
+
+<hash_start_block>
+    This is the offset, in <hash_block_size>-blocks, from the start of hash_dev
+    to the root block of the hash tree.
+
+<algorithm>
+    The cryptographic hash algorithm used for this device.  This should
+    be the name of the algorithm, like "sha1".
+
+<digest>
+    The hexadecimal encoding of the cryptographic hash of the root hash block
+    and the salt.  This hash should be trusted as there is no other authenticity
+    beyond this point.
+
+<salt>
+    The hexadecimal encoding of the salt value.
+
+Theory of operation
+===================
+
+dm-verity is meant to be set up as part of a verified boot path.  This
+may be anything ranging from a boot using tboot or trustedgrub to just
+booting from a known-good device (like a USB drive or CD).
+
+When a dm-verity device is configured, it is expected that the caller
+has been authenticated in some way (cryptographic signatures, etc).
+After instantiation, all hashes will be verified on-demand during
+disk access.  If they cannot be verified up to the root node of the
+tree, the root hash, then the I/O will fail.  This should detect
+tampering with any data on the device and the hash data.
+
+Cryptographic hashes are used to assert the integrity of the device on a
+per-block basis. This allows for a lightweight hash computation on first read
+into the page cache. Block hashes are stored linearly, aligned to the nearest
+block size.
+
+Hash Tree
+---------
+
+Each node in the tree is a cryptographic hash.  If it is a leaf node, the hash
+of some data block on disk is calculated. If it is an intermediary node,
+the hash of a number of child nodes is calculated.
+
+Each entry in the tree is a collection of neighboring nodes that fit in one
+block.  The number is determined based on block_size and the size of the
+selected cryptographic digest algorithm.  The hashes are linearly-ordered in
+this entry and any unaligned trailing space is ignored but included when
+calculating the parent node.
+
+The tree looks something like:
+
+alg = sha256, num_blocks = 32768, block_size = 4096
+
+                                 [   root    ]
+                                /    . . .    \
+                     [entry_0]                 [entry_1]
+                    /  . . .  \                 . . .   \
+         [entry_0_0]   . . .  [entry_0_127]    . . . .  [entry_1_127]
+           / ... \             /   . . .  \             /           \
+     blk_0 ... blk_127  blk_16256   blk_16383      blk_32640 . . . blk_32767
+
+
+On-disk format
+==============
+
+The verity kernel code does not read the verity metadata on-disk header.
+It only reads the hash blocks which directly follow the header.
+It is expected that a user-space tool will verify the integrity of the
+verity header.
+
+Alternatively, the header can be omitted and the dmsetup parameters can
+be passed via the kernel command-line in a rooted chain of trust where
+the command-line is verified.
+
+Directly following the header (and with sector number padded to the next hash
+block boundary) are the hash blocks which are stored a depth at a time
+(starting from the root), sorted in order of increasing index.
+
+The full specification of kernel parameters and on-disk metadata format
+is available at the cryptsetup project's wiki page
+  http://code.google.com/p/cryptsetup/wiki/DMVerity
+
+Status
+======
+V (for Valid) is returned if every check performed so far was valid.
+If any check failed, C (for Corruption) is returned.
+
+Example
+=======
+Set up a device:
+  # dmsetup create vroot --readonly --table \
+    "0 2097152 verity 1 /dev/sda1 /dev/sda2 4096 4096 262144 1 sha256 "\
+    "4392712ba01368efdf14b05c76f9e4df0d53664630b5d48632ed17a137f39076 "\
+    "1234000000000000000000000000000000000000000000000000000000000000"
+
+A command line tool veritysetup is available to compute or verify
+the hash tree or activate the kernel device. This is available from
+the cryptsetup upstream repository http://code.google.com/p/cryptsetup/
+(as a libcryptsetup extension).
+
+Create hash on the device:
+  # veritysetup format /dev/sda1 /dev/sda2
+  ...
+  Root hash: 4392712ba01368efdf14b05c76f9e4df0d53664630b5d48632ed17a137f39076
+
+Activate the device:
+  # veritysetup create vroot /dev/sda1 /dev/sda2 \
+    4392712ba01368efdf14b05c76f9e4df0d53664630b5d48632ed17a137f39076

doc/kernel/zero.txt

@@ -0,0 +1,37 @@
+dm-zero
+=======
+
+Device-Mapper's "zero" target provides a block-device that always returns
+zero'd data on reads and silently drops writes. This is similar behavior to
+/dev/zero, but as a block-device instead of a character-device.
+
+Dm-zero has no target-specific parameters.
+
+One very interesting use of dm-zero is for creating "sparse" devices in
+conjunction with dm-snapshot. A sparse device reports a device-size larger
+than the amount of actual storage space available for that device. A user can
+write data anywhere within the sparse device and read it back like a normal
+device. Reads to previously unwritten areas will return a zero'd buffer. When
+enough data has been written to fill up the actual storage space, the sparse
+device is deactivated. This can be very useful for testing device and
+filesystem limitations.
+
+To create a sparse device, start by creating a dm-zero device that's the
+desired size of the sparse device. For this example, we'll assume a 10TB
+sparse device.
+
+TEN_TERABYTES=`expr 10 \* 1024 \* 1024 \* 1024 \* 2`   # 10 TB in sectors
+echo "0 $TEN_TERABYTES zero" | dmsetup create zero1
+
+Then create a snapshot of the zero device, using any available block-device as
+the COW device. The size of the COW device will determine the amount of real
+space available to the sparse device. For this example, we'll assume /dev/sdb1
+is an available 10GB partition.
+
+echo "0 $TEN_TERABYTES snapshot /dev/mapper/zero1 /dev/sdb1 p 128" | \
+   dmsetup create sparse1
+
+This will create a 10TB sparse device called /dev/mapper/sparse1 that has
+10GB of actual storage space available. If more than 10GB of data is written
+to this device, it will start returning I/O errors.
+