Preparing for release of 3.2.2pre2

The release script & the patch management script now require the use of
an auto-build-save dir that makes it much easier to keep the generated
files from melding together, and remembers the configure setup for each
patch branch.

.cvsignore

@@ -1,25 +0,0 @@
-ID
-Makefile
-autom4te*.cache
-confdefs.h
-config.cache
-config.h
-config.log
-config.status
-conftest.c
-conftest.log
-dox
-getgroups
-gmon.out
-rsync
-shconfig
-testdir
-tests-dont-exist
-testtmp
-testtmp.*
-tls
-trimslash
-t_unsafe
-wildtest
-getfsdev
-.rsync-filter

.github/workflows/ccpp.yml

@@ -0,0 +1,31 @@
+name: build
+
+on:
+  push:
+    branches: [ master ]
+  pull_request:
+    branches: [ master ]
+
+jobs:
+  build:
+
+    runs-on: ubuntu-20.04
+
+    steps:
+    - uses: actions/checkout@v2
+    - name: prepare-packages
+      run: sudo apt-get install fakeroot acl libacl1-dev attr libattr1-dev liblz4-dev libzstd-dev libxxhash-dev python3-cmarkgfm
+    - name: prepare-source
+      run: ./prepare-source
+    - name: configure
+      run: ./configure --with-included-popt --with-included-zlib
+    - name: make
+      run: make
+    - name: version-summary
+      run: ./rsync --version
+    - name: make check
+      run: make check
+    - name: make check30
+      run: make check30
+    - name: make check29
+      run: make check29

.gitignore

@@ -0,0 +1,54 @@
+*.[oa]
+*~
+dummy
+ID
+Makefile
+Makefile.old
+configure.sh
+configure.sh.old
+config.cache
+config.h
+config.h.in
+config.h.in.old
+config.log
+config.status
+aclocal.m4
+/proto.h
+/proto.h-tstamp
+/rsync*.1
+/rsync*.5
+/rsync*.html
+/help-rsync*.h
+/default-cvsignore.h
+/default-dont-compress.h
+/.md2man-works
+/autom4te*.cache
+/confdefs.h
+/conftest*
+/dox
+/getgroups
+/gmon.out
+/rsync
+/stunnel-rsyncd.conf
+/shconfig
+/testdir
+/tests-dont-exist
+/testtmp
+/tls
+/testrun
+/trimslash
+/t_unsafe
+/wildtest
+/getfsdev
+/rounding.h
+/doc/rsync.pdf
+/doc/rsync.ps
+/support/savetransfer
+/testsuite/chown-fake.test
+/testsuite/devices-fake.test
+/testsuite/xattrs-hlink.test
+/patches
+/patches.gen
+/build
+/auto-build-save
+.deps

.ignore

@@ -1,19 +0,0 @@
-.#*
-*.log
-Makefile
-config.h
-*.o
-CVS
-.ignore
-.cvsignore
-*~
-rsync
-config.status
-config.cache
-TAGS
-config.log
-test
-*.gz
-rsync-*
-*.dvi
-*.aux

COPYING

@@ -1,285 +1,626 @@
-		    GNU GENERAL PUBLIC LICENSE
-		       Version 2, June 1991
+                    GNU GENERAL PUBLIC LICENSE
+                       Version 3, 29 June 2007
 
- Copyright (C) 1989, 1991 Free Software Foundation, Inc.
-                          675 Mass Ave, Cambridge, MA 02139, USA
+ Copyright (C) 2007 Free Software Foundation, Inc. <http://fsf.org/>
  Everyone is permitted to copy and distribute verbatim copies
  of this license document, but changing it is not allowed.
 
-			    Preamble
+                            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
+  The GNU General Public License is a free, copyleft license for
+software and other kinds of works.
+
+  The licenses for most software and other practical works are designed
+to take away your freedom to share and change the works.  By contrast,
+the GNU General Public License is intended to guarantee your freedom to
+share and change all versions of a program--to make sure it remains free
+software for all its users.  We, the Free Software Foundation, use the
+GNU General Public License for most of our software; it applies also to
+any other work released this way by its authors.  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.
+them 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.
+  To protect your rights, we need to prevent others from denying you
+these rights or asking you to surrender the rights.  Therefore, you have
+certain responsibilities if you distribute copies of the software, or if
+you modify it: responsibilities to respect the freedom of others.
 
   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.
+gratis or for a fee, you must pass on to the recipients the same
+freedoms that you received.  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.
+  Developers that use the GNU GPL protect your rights with two steps:
+(1) assert copyright on the software, and (2) offer you this License
+giving you legal permission to copy, distribute and/or modify it.
 
-  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.
+  For the developers' and authors' protection, the GPL clearly explains
+that there is no warranty for this free software.  For both users' and
+authors' sake, the GPL requires that modified versions be marked as
+changed, so that their problems will not be attributed erroneously to
+authors of previous versions.
 
-  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.
+  Some devices are designed to deny users access to install or run
+modified versions of the software inside them, although the manufacturer
+can do so.  This is fundamentally incompatible with the aim of
+protecting users' freedom to change the software.  The systematic
+pattern of such abuse occurs in the area of products for individuals to
+use, which is precisely where it is most unacceptable.  Therefore, we
+have designed this version of the GPL to prohibit the practice for those
+products.  If such problems arise substantially in other domains, we
+stand ready to extend this provision to those domains in future versions
+of the GPL, as needed to protect the freedom of users.
+
+  Finally, every program is threatened constantly by software patents.
+States should not allow patents to restrict development and use of
+software on general-purpose computers, but in those that do, we wish to
+avoid the special danger that patents applied to a free program could
+make it effectively proprietary.  To prevent this, the GPL assures that
+patents cannot be used to render the program non-free.
 
   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".
+                       TERMS AND CONDITIONS
 
-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.
+  0. Definitions.
 
-  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.
+  "This License" refers to version 3 of the GNU General Public License.
 
-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.
+  "Copyright" also means copyright-like laws that apply to other kinds of
+works, such as semiconductor masks.
 
-  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:
+  "The Program" refers to any copyrightable work licensed under this
+License.  Each licensee is addressed as "you".  "Licensees" and
+"recipients" may be individuals or organizations.
 
-    a) You must cause the modified files to carry prominent notices
-    stating that you changed the files and the date of any change.
+  To "modify" a work means to copy from or adapt all or part of the work
+in a fashion requiring copyright permission, other than the making of an
+exact copy.  The resulting work is called a "modified version" of the
+earlier work or a work "based on" the earlier work.
 
-    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.
+  A "covered work" means either the unmodified Program or a work based
+on the Program.
 
-    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.
+  To "propagate" a work means to do anything with it that, without
+permission, would make you directly or secondarily liable for
+infringement under applicable copyright law, except executing it on a
+computer or modifying a private copy.  Propagation includes copying,
+distribution (with or without modification), making available to the
+public, and in some countries other activities as well.
 
-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.
+  To "convey" a work means any kind of propagation that enables other
+parties to make or receive copies.  Mere interaction with a user through
+a computer network, with no transfer of a copy, is not conveying.
 
-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.
+  An interactive user interface displays "Appropriate Legal Notices"
+to the extent that it includes a convenient and prominently visible
+feature that (1) displays an appropriate copyright notice, and (2)
+tells the user that there is no warranty for the work (except to the
+extent that warranties are provided), that licensees may convey the
+work under this License, and how to view a copy of this License.  If
+the interface presents a list of user commands or options, such as a
+menu, a prominent item in the list meets this criterion.
 
-  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:
+  1. Source Code.
 
-    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,
+  The "source code" for a work means the preferred form of the work
+for making modifications to it.  "Object code" means any non-source
+form of a work.
 
-    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,
+  A "Standard Interface" means an interface that either is an official
+standard defined by a recognized standards body, or, in the case of
+interfaces specified for a particular programming language, one that
+is widely used among developers working in that language.
 
-    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 "System Libraries" of an executable work include anything, other
+than the work as a whole, that (a) is included in the normal form of
+packaging a Major Component, but which is not part of that Major
+Component, and (b) serves only to enable use of the work with that
+Major Component, or to implement a Standard Interface for which an
+implementation is available to the public in source code form.  A
+"Major Component", in this context, means a major essential component
+(kernel, window system, and so on) of the specific operating system
+(if any) on which the executable work runs, or a compiler used to
+produce the work, or an object code interpreter used to run it.
 
-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.
+  The "Corresponding Source" for a work in object code form means all
+the source code needed to generate, install, and (for an executable
+work) run the object code and to modify the work, including scripts to
+control those activities.  However, it does not include the work's
+System Libraries, or general-purpose tools or generally available free
+programs which are used unmodified in performing those activities but
+which are not part of the work.  For example, Corresponding Source
+includes interface definition files associated with source files for
+the work, and the source code for shared libraries and dynamically
+linked subprograms that the work is specifically designed to require,
+such as by intimate data communication or control flow between those
+subprograms and other parts of the work.
 
-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.
+  The Corresponding Source need not include anything that users
+can regenerate automatically from other parts of the Corresponding
+Source.
 
-  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.
+  The Corresponding Source for a work in source code form is that
+same work.
 
-  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
+  2. Basic Permissions.
+
+  All rights granted under this License are granted for the term of
+copyright on the Program, and are irrevocable provided the stated
+conditions are met.  This License explicitly affirms your unlimited
+permission to run the unmodified Program.  The output from running a
+covered work is covered by this License only if the output, given its
+content, constitutes a covered work.  This License acknowledges your
+rights of fair use or other equivalent, as provided by copyright law.
+
+  You may make, run and propagate covered works that you do not
+convey, without conditions so long as your license otherwise remains
+in force.  You may convey covered works to others for the sole purpose
+of having them make modifications exclusively for you, or provide you
+with facilities for running those works, provided that you comply with
+the terms of this License in conveying all material for which you do
+not control copyright.  Those thus making or running the covered works
+for you must do so exclusively on your behalf, under your direction
+and control, on terms that prohibit them from making any copies of
+your copyrighted material outside their relationship with you.
+
+  Conveying under any other circumstances is permitted solely under
+the conditions stated below.  Sublicensing is not allowed; section 10
+makes it unnecessary.
+
+  3. Protecting Users' Legal Rights From Anti-Circumvention Law.
+
+  No covered work shall be deemed part of an effective technological
+measure under any applicable law fulfilling obligations under article
+11 of the WIPO copyright treaty adopted on 20 December 1996, or
+similar laws prohibiting or restricting circumvention of such
+measures.
+
+  When you convey a covered work, you waive any legal power to forbid
+circumvention of technological measures to the extent such circumvention
+is effected by exercising rights under this License with respect to
+the covered work, and you disclaim any intention to limit operation or
+modification of the work as a means of enforcing, against the work's
+users, your or third parties' legal rights to forbid circumvention of
+technological measures.
+
+  4. Conveying Verbatim Copies.
+
+  You may convey 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;
+keep intact all notices stating that this License and any
+non-permissive terms added in accord with section 7 apply to the code;
+keep intact all notices of the absence of any warranty; and give all
+recipients a copy of this License along with the Program.
+
+  You may charge any price or no price for each copy that you convey,
+and you may offer support or warranty protection for a fee.
+
+  5. Conveying Modified Source Versions.
+
+  You may convey a work based on the Program, or the modifications to
+produce it from the Program, in the form of source code under the
+terms of section 4, provided that you also meet all of these conditions:
+
+    a) The work must carry prominent notices stating that you modified
+    it, and giving a relevant date.
+
+    b) The work must carry prominent notices stating that it is
+    released under this License and any conditions added under section
+    7.  This requirement modifies the requirement in section 4 to
+    "keep intact all notices".
+
+    c) You must license the entire work, as a whole, under this
+    License to anyone who comes into possession of a copy.  This
+    License will therefore apply, along with any applicable section 7
+    additional terms, to the whole of the work, and all its parts,
+    regardless of how they are packaged.  This License gives no
+    permission to license the work in any other way, but it does not
+    invalidate such permission if you have separately received it.
+
+    d) If the work has interactive user interfaces, each must display
+    Appropriate Legal Notices; however, if the Program has interactive
+    interfaces that do not display Appropriate Legal Notices, your
+    work need not make them do so.
+
+  A compilation of a covered work with other separate and independent
+works, which are not by their nature extensions of the covered work,
+and which are not combined with it such as to form a larger program,
+in or on a volume of a storage or distribution medium, is called an
+"aggregate" if the compilation and its resulting copyright are not
+used to limit the access or legal rights of the compilation's users
+beyond what the individual works permit.  Inclusion of a covered work
+in an aggregate does not cause this License to apply to the other
+parts of the aggregate.
+
+  6. Conveying Non-Source Forms.
+
+  You may convey a covered work in object code form under the terms
+of sections 4 and 5, provided that you also convey the
+machine-readable Corresponding Source under the terms of this License,
+in one of these ways:
+
+    a) Convey the object code in, or embodied in, a physical product
+    (including a physical distribution medium), accompanied by the
+    Corresponding Source fixed on a durable physical medium
+    customarily used for software interchange.
+
+    b) Convey the object code in, or embodied in, a physical product
+    (including a physical distribution medium), accompanied by a
+    written offer, valid for at least three years and valid for as
+    long as you offer spare parts or customer support for that product
+    model, to give anyone who possesses the object code either (1) a
+    copy of the Corresponding Source for all the software in the
+    product that is covered by this License, on a durable physical
+    medium customarily used for software interchange, for a price no
+    more than your reasonable cost of physically performing this
+    conveying of source, or (2) access to copy the
+    Corresponding Source from a network server at no charge.
+
+    c) Convey individual copies of the object code with a copy of the
+    written offer to provide the Corresponding Source.  This
+    alternative is allowed only occasionally and noncommercially, and
+    only if you received the object code with such an offer, in accord
+    with subsection 6b.
+
+    d) Convey the object code by offering access from a designated
+    place (gratis or for a charge), and offer equivalent access to the
+    Corresponding Source in the same way through the same place at no
+    further charge.  You need not require recipients to copy the
+    Corresponding Source along with the object code.  If the place to
+    copy the object code is a network server, the Corresponding Source
+    may be on a different server (operated by you or a third party)
+    that supports equivalent copying facilities, provided you maintain
+    clear directions next to the object code saying where to find the
+    Corresponding Source.  Regardless of what server hosts the
+    Corresponding Source, you remain obligated to ensure that it is
+    available for as long as needed to satisfy these requirements.
+
+    e) Convey the object code using peer-to-peer transmission, provided
+    you inform other peers where the object code and Corresponding
+    Source of the work are being offered to the general public at no
+    charge under subsection 6d.
+
+  A separable portion of the object code, whose source code is excluded
+from the Corresponding Source as a System Library, need not be
+included in conveying the object code work.
+
+  A "User Product" is either (1) a "consumer product", which means any
+tangible personal property which is normally used for personal, family,
+or household purposes, or (2) anything designed or sold for incorporation
+into a dwelling.  In determining whether a product is a consumer product,
+doubtful cases shall be resolved in favor of coverage.  For a particular
+product received by a particular user, "normally used" refers to a
+typical or common use of that class of product, regardless of the status
+of the particular user or of the way in which the particular user
+actually uses, or expects or is expected to use, the product.  A product
+is a consumer product regardless of whether the product has substantial
+commercial, industrial or non-consumer uses, unless such uses represent
+the only significant mode of use of the product.
+
+  "Installation Information" for a User Product means any methods,
+procedures, authorization keys, or other information required to install
+and execute modified versions of a covered work in that User Product from
+a modified version of its Corresponding Source.  The information must
+suffice to ensure that the continued functioning of the modified object
+code is in no case prevented or interfered with solely because
+modification has been made.
+
+  If you convey an object code work under this section in, or with, or
+specifically for use in, a User Product, and the conveying occurs as
+part of a transaction in which the right of possession and use of the
+User Product is transferred to the recipient in perpetuity or for a
+fixed term (regardless of how the transaction is characterized), the
+Corresponding Source conveyed under this section must be accompanied
+by the Installation Information.  But this requirement does not apply
+if neither you nor any third party retains the ability to install
+modified object code on the User Product (for example, the work has
+been installed in ROM).
+
+  The requirement to provide Installation Information does not include a
+requirement to continue to provide support service, warranty, or updates
+for a work that has been modified or installed by the recipient, or for
+the User Product in which it has been modified or installed.  Access to a
+network may be denied when the modification itself materially and
+adversely affects the operation of the network or violates the rules and
+protocols for communication across the network.
+
+  Corresponding Source conveyed, and Installation Information provided,
+in accord with this section must be in a format that is publicly
+documented (and with an implementation available to the public in
+source code form), and must require no special password or key for
+unpacking, reading or copying.
+
+  7. Additional Terms.
+
+  "Additional permissions" are terms that supplement the terms of this
+License by making exceptions from one or more of its conditions.
+Additional permissions that are applicable to the entire Program shall
+be treated as though they were included in this License, to the extent
+that they are valid under applicable law.  If additional permissions
+apply only to part of the Program, that part may be used separately
+under those permissions, but the entire Program remains governed by
+this License without regard to the additional permissions.
+
+  When you convey a copy of a covered work, you may at your option
+remove any additional permissions from that copy, or from any part of
+it.  (Additional permissions may be written to require their own
+removal in certain cases when you modify the work.)  You may place
+additional permissions on material, added by you to a covered work,
+for which you have or can give appropriate copyright permission.
+
+  Notwithstanding any other provision of this License, for material you
+add to a covered work, you may (if authorized by the copyright holders of
+that material) supplement the terms of this License with terms:
+
+    a) Disclaiming warranty or limiting liability differently from the
+    terms of sections 15 and 16 of this License; or
+
+    b) Requiring preservation of specified reasonable legal notices or
+    author attributions in that material or in the Appropriate Legal
+    Notices displayed by works containing it; or
+
+    c) Prohibiting misrepresentation of the origin of that material, or
+    requiring that modified versions of such material be marked in
+    reasonable ways as different from the original version; or
+
+    d) Limiting the use for publicity purposes of names of licensors or
+    authors of the material; or
+
+    e) Declining to grant rights under trademark law for use of some
+    trade names, trademarks, or service marks; or
+
+    f) Requiring indemnification of licensors and authors of that
+    material by anyone who conveys the material (or modified versions of
+    it) with contractual assumptions of liability to the recipient, for
+    any liability that these contractual assumptions directly impose on
+    those licensors and authors.
+
+  All other non-permissive additional terms are considered "further
+restrictions" within the meaning of section 10.  If the Program as you
+received it, or any part of it, contains a notice stating that it is
+governed by this License along with a term that is a further
+restriction, you may remove that term.  If a license document contains
+a further restriction but permits relicensing or conveying under this
+License, you may add to a covered work material governed by the terms
+of that license document, provided that the further restriction does
+not survive such relicensing or conveying.
+
+  If you add terms to a covered work in accord with this section, you
+must place, in the relevant source files, a statement of the
+additional terms that apply to those files, or a notice indicating
+where to find the applicable terms.
+
+  Additional terms, permissive or non-permissive, may be stated in the
+form of a separately written license, or stated as exceptions;
+the above requirements apply either way.
+
+  8. Termination.
+
+  You may not propagate or modify a covered work except as expressly
+provided under this License.  Any attempt otherwise to propagate or
+modify it is void, and will automatically terminate your rights under
+this License (including any patent licenses granted under the third
+paragraph of section 11).
+
+  However, if you cease all violation of this License, then your
+license from a particular copyright holder is reinstated (a)
+provisionally, unless and until the copyright holder explicitly and
+finally terminates your license, and (b) permanently, if the copyright
+holder fails to notify you of the violation by some reasonable means
+prior to 60 days after the cessation.
+
+  Moreover, your license from a particular copyright holder is
+reinstated permanently if the copyright holder notifies you of the
+violation by some reasonable means, this is the first time you have
+received notice of violation of this License (for any work) from that
+copyright holder, and you cure the violation prior to 30 days after
+your receipt of the notice.
+
+  Termination of your rights under this section does not terminate the
+licenses of parties who have received copies or rights from you under
+this License.  If your rights have been terminated and not permanently
+reinstated, you do not qualify to receive new licenses for the same
+material under section 10.
+
+  9. Acceptance Not Required for Having Copies.
+
+  You are not required to accept this License in order to receive or
+run a copy of the Program.  Ancillary propagation of a covered work
+occurring solely as a consequence of using peer-to-peer transmission
+to receive a copy likewise does not require acceptance.  However,
+nothing other than this License grants you permission to propagate or
+modify any covered work.  These actions infringe copyright if you do
+not accept this License.  Therefore, by modifying or propagating a
+covered work, you indicate your acceptance of this License to do so.
+
+  10. Automatic Licensing of Downstream Recipients.
+
+  Each time you convey a covered work, the recipient automatically
+receives a license from the original licensors, to run, modify and
+propagate that work, subject to this License.  You are not responsible
+for enforcing compliance by third parties with this License.
+
+  An "entity transaction" is a transaction transferring control of an
+organization, or substantially all assets of one, or subdividing an
+organization, or merging organizations.  If propagation of a covered
+work results from an entity transaction, each party to that
+transaction who receives a copy of the work also receives whatever
+licenses to the work the party's predecessor in interest had or could
+give under the previous paragraph, plus a right to possession of the
+Corresponding Source of the work from the predecessor in interest, if
+the predecessor has it or can get it with reasonable efforts.
+
+  You may not impose any further restrictions on the exercise of the
+rights granted or affirmed under this License.  For example, you may
+not impose a license fee, royalty, or other charge for exercise of
+rights granted under this License, and you may not initiate litigation
+(including a cross-claim or counterclaim in a lawsuit) alleging that
+any patent claim is infringed by making, using, selling, offering for
+sale, or importing the Program or any portion of it.
+
+  11. Patents.
+
+  A "contributor" is a copyright holder who authorizes use under this
+License of the Program or a work on which the Program is based.  The
+work thus licensed is called the contributor's "contributor version".
+
+  A contributor's "essential patent claims" are all patent claims
+owned or controlled by the contributor, whether already acquired or
+hereafter acquired, that would be infringed by some manner, permitted
+by this License, of making, using, or selling its contributor version,
+but do not include claims that would be infringed only as a
+consequence of further modification of the contributor version.  For
+purposes of this definition, "control" includes the right to grant
+patent sublicenses in a manner consistent with the requirements of
 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
+  Each contributor grants you a non-exclusive, worldwide, royalty-free
+patent license under the contributor's essential patent claims, to
+make, use, sell, offer for sale, import and otherwise run, modify and
+propagate the contents of its contributor version.
+
+  In the following three paragraphs, a "patent license" is any express
+agreement or commitment, however denominated, not to enforce a patent
+(such as an express permission to practice a patent or covenant not to
+sue for patent infringement).  To "grant" such a patent license to a
+party means to make such an agreement or commitment not to enforce a
+patent against the party.
+
+  If you convey a covered work, knowingly relying on a patent license,
+and the Corresponding Source of the work is not available for anyone
+to copy, free of charge and under the terms of this License, through a
+publicly available network server or other readily accessible means,
+then you must either (1) cause the Corresponding Source to be so
+available, or (2) arrange to deprive yourself of the benefit of the
+patent license for this particular work, or (3) arrange, in a manner
+consistent with the requirements of this License, to extend the patent
+license to downstream recipients.  "Knowingly relying" means you have
+actual knowledge that, but for the patent license, your conveying the
+covered work in a country, or your recipient's use of the covered work
+in a country, would infringe one or more identifiable patents in that
+country that you have reason to believe are valid.
+
+  If, pursuant to or in connection with a single transaction or
+arrangement, you convey, or propagate by procuring conveyance of, a
+covered work, and grant a patent license to some of the parties
+receiving the covered work authorizing them to use, propagate, modify
+or convey a specific copy of the covered work, then the patent license
+you grant is automatically extended to all recipients of the covered
+work and works based on it.
+
+  A patent license is "discriminatory" if it does not include within
+the scope of its coverage, prohibits the exercise of, or is
+conditioned on the non-exercise of one or more of the rights that are
+specifically granted under this License.  You may not convey a covered
+work if you are a party to an arrangement with a third party that is
+in the business of distributing software, under which you make payment
+to the third party based on the extent of your activity of conveying
+the work, and under which the third party grants, to any of the
+parties who would receive the covered work from you, a discriminatory
+patent license (a) in connection with copies of the covered work
+conveyed by you (or copies made from those copies), or (b) primarily
+for and in connection with specific products or compilations that
+contain the covered work, unless you entered into that arrangement,
+or that patent license was granted, prior to 28 March 2007.
+
+  Nothing in this License shall be construed as excluding or limiting
+any implied license or other defenses to infringement that may
+otherwise be available to you under applicable patent law.
+
+  12. No Surrender of Others' Freedom.
+
+  If 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.
+excuse you from the conditions of this License.  If you cannot convey a
+covered work so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you may
+not convey it at all.  For example, if you agree to terms that obligate you
+to collect a royalty for further conveying from those to whom you convey
+the Program, the only way you could satisfy both those terms and this
+License would be to refrain entirely from conveying 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.
+  13. Use with the GNU Affero General Public License.
 
-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.
+  Notwithstanding any other provision of this License, you have
+permission to link or combine any covered work with a work licensed
+under version 3 of the GNU Affero General Public License into a single
+combined work, and to convey the resulting work.  The terms of this
+License will continue to apply to the part which is the covered work,
+but the special requirements of the GNU Affero General Public License,
+section 13, concerning interaction through a network will apply to the
+combination as such.
 
-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.
+  14. Revised Versions 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
+  The Free Software Foundation may publish revised and/or new versions of
+the GNU 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.
+  Each version is given a distinguishing version number.  If the
+Program specifies that a certain numbered version of the GNU General
+Public License "or any later version" applies to it, you have the
+option of following the terms and conditions either of that numbered
+version or of any later version published by the Free Software
+Foundation.  If the Program does not specify a version number of the
+GNU General Public 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.
+  If the Program specifies that a proxy can decide which future
+versions of the GNU General Public License can be used, that proxy's
+public statement of acceptance of a version permanently authorizes you
+to choose that version for the Program.
 
-			    NO WARRANTY
+  Later license versions may give you additional or different
+permissions.  However, no additional obligations are imposed on any
+author or copyright holder as a result of your choosing to follow a
+later version.
 
-  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.
+  15. Disclaimer of Warranty.
 
-  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.
+  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.
 
-		     END OF TERMS AND CONDITIONS
-
-	Appendix: How to Apply These Terms to Your New Programs
+  16. Limitation of Liability.
+
+  IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
+WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS
+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.
+
+  17. Interpretation of Sections 15 and 16.
+
+  If the disclaimer of warranty and limitation of liability provided
+above cannot be given local legal effect according to their terms,
+reviewing courts shall apply local law that most closely approximates
+an absolute waiver of all civil liability in connection with the
+Program, unless a warranty or assumption of liability accompanies a
+copy of the Program in return for a fee.
+
+                     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
@@ -287,15 +628,15 @@ 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
+state 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) 19yy  <name of author>
+    Copyright (C) <year>  <name of author>
 
-    This program is free software; you can redistribute it and/or modify
+    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
+    the Free Software Foundation, either version 3 of the License, or
     (at your option) any later version.
 
     This program is distributed in the hope that it will be useful,
@@ -304,36 +645,39 @@ the "copyright" line and a pointer to where the full notice is found.
     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., 675 Mass Ave, Cambridge, MA 02139, USA.
+    along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
 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:
+  If the program does terminal interaction, make it output a short
+notice like this when it starts in an interactive mode:
 
-    Gnomovision version 69, Copyright (C) 19yy name of author
-    Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
+    <program>  Copyright (C) <year>  <name of author>
+    This program 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.
+parts of the General Public License.  Of course, your program's commands
+might be different; for a GUI interface, you would use an "about box".
 
-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:
+  You should also get your employer (if you work as a programmer) or school,
+if any, to sign a "copyright disclaimer" for the program, if necessary.
+For more information on this, and how to apply and follow the GNU GPL, see
+<http://www.gnu.org/licenses/>.
 
-  Yoyodyne, Inc., hereby disclaims all copyright interest in the program
-  `Gnomovision' (which makes passes at compilers) written by James Hacker.
+  The GNU 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 Lesser General
+Public License instead of this License.  But first, please read
+<http://www.gnu.org/philosophy/why-not-lgpl.html>.
 
-  <signature of Ty Coon>, 1 April 1989
-  Ty Coon, President of Vice
+REGARDING OPENSSL AND XXHASH
 
-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.
+In addition, as a special exception, the copyright holders give
+permission to dynamically link rsync with the OpenSSL and xxhash
+libraries when those libraries are being distributed in compliance
+with their license terms, and to distribute a dynamically linked
+combination of rsync and these libraries.  This is also considered
+to be covered under the GPL's System Libraries exception.

INSTALL.md

@@ -1,16 +1,23 @@
-To build and install rsync
+To build and install rsync:
 
-  $ ./configure
-  $ make
-  # make install
+    $ ./configure
+    $ make
+    # make install
 
 You may set the installation directory and other parameters by options
 to ./configure.  To see them, use:
 
-  $ ./configure --help
+    $ ./configure --help
+
+Configure tries to figure out if the local system uses group "nobody" or
+"nogroup" by looking in the /etc/group file.  (This is only used for the
+default group of an rsync daemon, which attempts to run with "nobody"
+user and group permissions.)  You can change the default user and group
+for the daemon by editing the NOBODY_USER and NOBODY_GROUP defines in
+config.h, or just override them in your /etc/rsyncd.conf file.
 
 As of 2.4.7, rsync uses Eric Troan's popt option-parsing library.  A
-cut-down copy of release 1.5 is included in the rsync distribution,
+cut-down copy of a recent release is included in the rsync distribution,
 and will be used if there is no popt library on your build host, or if
 the --with-included-popt option is passed to ./configure.
 
@@ -18,6 +25,16 @@ If you configure using --enable-maintainer-mode, then rsync will try
 to pop up an xterm on DISPLAY=:0 if it crashes.  You might find this
 useful, but it should be turned off for production builds.
 
+MAKE COMPATIBILITY
+------------------
+
+Note that Makefile.in has a rule that uses a wildcard in a prerequisite.  If
+your make has a problem with this rule, you will see an error like this:
+
+    Don't know how to make ./*.c
+
+You can change the "proto.h-tstamp" target in Makefile.in to list all the *.c
+filenames explicitly in order to avoid this issue.
 
 RPM NOTES
 ---------
@@ -37,16 +54,15 @@ fails:
 
 Install gcc or HP's "ANSI/C Compiler".
 
-
-
 MAC OSX NOTES
 -------------
 
-Mac OS X (Darwin) seems to have an IPv6 stack, but it does not
-completely implement the "New Sockets" API.  
+Some versions of Mac OS X (Darwin) seem to have an IPv6 stack, but do
+not completely implement the "New Sockets" API.  
 
-<http://www.ipv6.org/impl/mac.html> says that Apple do not support
-IPv6 yet.  If your build fails, try again with --disable-ipv6.
+<http://www.ipv6.org/impl/mac.html> says that Apple started to support
+IPv6 in 10.2 (Jaguar).  If your build fails, try again after running
+configure with --disable-ipv6.
 
 IBM AIX NOTES
 -------------

Makefile.in

@@ -1,54 +1,65 @@
-# Makefile for rsync. This is processed by configure to produce the final
-# Makefile
+# The input file that configure uses to create the Makefile for rsync.
 
 prefix=@prefix@
+datarootdir=@datarootdir@
 exec_prefix=@exec_prefix@
 bindir=@bindir@
+libdir=@libdir@/rsync
 mandir=@mandir@
 
 LIBS=@LIBS@
 CC=@CC@
+AWK=@AWK@
 CFLAGS=@CFLAGS@
 CPPFLAGS=@CPPFLAGS@
+CXX=@CXX@
+CXXFLAGS=@CXXFLAGS@
 EXEEXT=@EXEEXT@
 LDFLAGS=@LDFLAGS@
+LIBOBJDIR=lib/
 
 INSTALLCMD=@INSTALL@
 INSTALLMAN=@INSTALL@
 
 srcdir=@srcdir@
+MKDIR_P=@MKDIR_P@
 VPATH=$(srcdir)
 SHELL=/bin/sh
 
-VERSION=@VERSION@
-
 .SUFFIXES:
 .SUFFIXES: .c .o
 
-HEADERS=byteorder.h config.h errcode.h proto.h rsync.h lib/pool_alloc.h
-LIBOBJ=lib/wildmatch.o lib/compat.o lib/snprintf.o lib/mdfour.o \
-	lib/permstring.o lib/pool_alloc.o @LIBOBJS@
-ZLIBOBJ=zlib/deflate.o zlib/infblock.o zlib/infcodes.o zlib/inffast.o \
-	zlib/inflate.o zlib/inftrees.o zlib/infutil.o zlib/trees.o \
-	zlib/zutil.o zlib/adler32.o
-OBJS1=rsync.o generator.o receiver.o cleanup.o sender.o exclude.o util.o \
-	main.o checksum.o match.o syscall.o log.o backup.o
-OBJS2=options.o flist.o io.o compat.o hlink.o token.o uidlist.o socket.o \
-	fileio.o batch.o clientname.o
-OBJS3=progress.o pipe.o
+SIMD_x86_64=simd-checksum-x86_64.o
+ASM_x86_64=lib/md5-asm-x86_64.o
+
+GENFILES=configure.sh aclocal.m4 config.h.in proto.h proto.h-tstamp rsync.1 rsync.1.html \
+	 rsync-ssl.1 rsync-ssl.1.html rsyncd.conf.5 rsyncd.conf.5.html
+HEADERS=byteorder.h config.h errcode.h proto.h rsync.h ifuncs.h itypes.h inums.h \
+	lib/pool_alloc.h lib/mdigest.h lib/md-defines.h version.h
+LIBOBJ=lib/wildmatch.o lib/compat.o lib/snprintf.o lib/mdfour.o lib/md5.o \
+	lib/permstring.o lib/pool_alloc.o lib/sysacls.o lib/sysxattrs.o @LIBOBJS@
+zlib_OBJS=zlib/deflate.o zlib/inffast.o zlib/inflate.o zlib/inftrees.o \
+	zlib/trees.o zlib/zutil.o zlib/adler32.o zlib/compress.o zlib/crc32.o
+OBJS1=flist.o rsync.o generator.o receiver.o cleanup.o sender.o exclude.o \
+	util.o util2.o main.o checksum.o match.o syscall.o log.o backup.o delete.o
+OBJS2=options.o io.o compat.o hlink.o token.o uidlist.o socket.o hashtable.o \
+	fileio.o batch.o clientname.o chmod.o acls.o xattrs.o
+OBJS3=progress.o pipe.o @ASM@
 DAEMON_OBJ = params.o loadparm.o clientserver.o access.o connection.o authenticate.o
 popt_OBJS=popt/findme.o  popt/popt.o  popt/poptconfig.o \
 	popt/popthelp.o popt/poptparse.o
-OBJS=$(OBJS1) $(OBJS2) $(OBJS3) $(DAEMON_OBJ) $(LIBOBJ) $(ZLIBOBJ) @BUILD_POPT@
+OBJS=$(OBJS1) $(OBJS2) $(OBJS3) @SIMD@ $(DAEMON_OBJ) $(LIBOBJ) @BUILD_ZLIB@ @BUILD_POPT@
 
-TLS_OBJ = tls.o syscall.o lib/compat.o lib/snprintf.o lib/permstring.o
+TLS_OBJ = tls.o syscall.o t_stub.o lib/compat.o lib/snprintf.o lib/permstring.o lib/sysxattrs.o @BUILD_POPT@
 
 # Programs we must have to run the test cases
 CHECK_PROGS = rsync$(EXEEXT) tls$(EXEEXT) getgroups$(EXEEXT) getfsdev$(EXEEXT) \
-	trimslash$(EXEEXT) t_unsafe$(EXEEXT) wildtest$(EXEEXT)
+	testrun$(EXEEXT) trimslash$(EXEEXT) t_unsafe$(EXEEXT) wildtest$(EXEEXT)
+
+CHECK_SYMLINKS = testsuite/chown-fake.test testsuite/devices-fake.test testsuite/xattrs-hlink.test
 
 # Objects for CHECK_PROGS to clean
-CHECK_OBJS=getgroups.o getfsdev.o t_stub.o t_unsafe.o trimslash.o wildtest.o
+CHECK_OBJS=tls.o testrun.o getgroups.o getfsdev.o t_stub.o t_unsafe.o trimslash.o wildtest.o
 
 # note that the -I. is needed to handle config.h when using VPATH
 .c.o:
@@ -56,15 +67,28 @@ CHECK_OBJS=getgroups.o getfsdev.o t_stub.o t_unsafe.o trimslash.o wildtest.o
 	$(CC) -I. -I$(srcdir) $(CFLAGS) $(CPPFLAGS) -c $< @CC_SHOBJ_FLAG@
 @OBJ_RESTORE@
 
-all: rsync$(EXEEXT)
+.PHONY: all
+all: Makefile rsync$(EXEEXT) stunnel-rsyncd.conf @MAKE_MAN@
 
+.PHONY: install
 install: all
-	-mkdir -p ${DESTDIR}${bindir}
+	-${MKDIR_P} ${DESTDIR}${bindir}
 	${INSTALLCMD} ${INSTALL_STRIP} -m 755 rsync$(EXEEXT) ${DESTDIR}${bindir}
-	-mkdir -p ${DESTDIR}${mandir}/man1
-	-mkdir -p ${DESTDIR}${mandir}/man5
-	${INSTALLMAN} -m 644 $(srcdir)/rsync.1 ${DESTDIR}${mandir}/man1
-	${INSTALLMAN} -m 644 $(srcdir)/rsyncd.conf.5 ${DESTDIR}${mandir}/man5
+	${INSTALLCMD} -m 755 $(srcdir)/rsync-ssl ${DESTDIR}${bindir}
+	-${MKDIR_P} ${DESTDIR}${mandir}/man1
+	-${MKDIR_P} ${DESTDIR}${mandir}/man5
+	if test -f rsync.1; then ${INSTALLMAN} -m 644 rsync.1 ${DESTDIR}${mandir}/man1; fi
+	if test -f rsync-ssl.1; then ${INSTALLMAN} -m 644 rsync-ssl.1 ${DESTDIR}${mandir}/man1; fi
+	if test -f rsyncd.conf.5; then ${INSTALLMAN} -m 644 rsyncd.conf.5 ${DESTDIR}${mandir}/man5; fi
+
+install-ssl-daemon: stunnel-rsyncd.conf
+	-${MKDIR_P} ${DESTDIR}/etc/stunnel
+	${INSTALLCMD} -m 644 stunnel-rsyncd.conf ${DESTDIR}/etc/stunnel/rsyncd.conf
+	@if ! ls /etc/rsync-ssl/certs/server.* >/dev/null 2>/dev/null; then \
+	    echo "Note that you'll need to install the certificate used by /etc/stunnel/rsyncd.conf"; \
+	fi
+
+install-all: install install-ssl-client install-ssl-daemon
 
 install-strip:
 	$(MAKE) INSTALL_STRIP='-s' install
@@ -73,80 +97,190 @@ rsync$(EXEEXT): $(OBJS)
 	$(CC) $(CFLAGS) $(LDFLAGS) -o $@ $(OBJS) $(LIBS)
 
 $(OBJS): $(HEADERS)
+$(CHECK_OBJS): $(HEADERS)
+tls.o xattrs.o: lib/sysxattrs.h
+options.o: latest-year.h help-rsync.h help-rsyncd.h
+exclude.o: default-cvsignore.h
+loadparm.o: default-dont-compress.h
+
+flist.o: rounding.h
+
+default-cvsignore.h default-dont-compress.h: rsync.1.md define-from-md.awk
+	$(AWK) -f $(srcdir)/define-from-md.awk -v hfile=$@ $(srcdir)/rsync.1.md
+
+help-rsync.h help-rsyncd.h: rsync.1.md help-from-md.awk
+	$(AWK) -f $(srcdir)/help-from-md.awk -v hfile=$@ $(srcdir)/rsync.1.md
+
+rounding.h: rounding.c rsync.h proto.h
+	@for r in 0 1 3; do \
+	    if $(CC) $(CFLAGS) $(CPPFLAGS) $(LDFLAGS) -o rounding -DEXTRA_ROUNDING=$$r -I. $(srcdir)/rounding.c >rounding.out 2>&1; then \
+		echo "#define EXTRA_ROUNDING $$r" >rounding.h; \
+		if test -f "$$HOME/build_farm/build_test.fns"; then \
+		    echo "EXTRA_ROUNDING is $$r" >&2; \
+		fi; \
+		break; \
+	    fi; \
+	done
+	@rm -f rounding
+	@if test -f rounding.h; then : ; else \
+	    cat rounding.out 1>&2; \
+	    echo "Failed to create rounding.h!" 1>&2; \
+	    exit 1; \
+	fi
+	@rm -f rounding.out
+
+simd-checksum-x86_64.o: simd-checksum-x86_64.cpp
+	@$(srcdir)/cmdormsg disable-simd $(CXX) -I. $(CXXFLAGS) $(CPPFLAGS) -c -o $@ $(srcdir)/simd-checksum-x86_64.cpp
+
+lib/md5-asm-x86_64.o: lib/md5-asm-x86_64.S config.h lib/md-defines.h
+	@$(srcdir)/cmdormsg disable-asm $(CC) -I. @NOEXECSTACK@ -c -o $@ $(srcdir)/lib/md5-asm-x86_64.S
 
 tls$(EXEEXT): $(TLS_OBJ)
 	$(CC) $(CFLAGS) $(LDFLAGS) -o $@ $(TLS_OBJ) $(LIBS)
 
+testrun$(EXEEXT): testrun.o
+	$(CC) $(CFLAGS) $(LDFLAGS) -o $@ testrun.o
+
 getgroups$(EXEEXT): getgroups.o
 	$(CC) $(CFLAGS) $(LDFLAGS) -o $@ getgroups.o $(LIBS)
 
 getfsdev$(EXEEXT): getfsdev.o
 	$(CC) $(CFLAGS) $(LDFLAGS) -o $@ getfsdev.o $(LIBS)
 
-TRIMSLASH_OBJ = trimslash.o syscall.o lib/compat.o lib/snprintf.o
+TRIMSLASH_OBJ = trimslash.o syscall.o t_stub.o lib/compat.o lib/snprintf.o
 trimslash$(EXEEXT): $(TRIMSLASH_OBJ)
 	$(CC) $(CFLAGS) $(LDFLAGS) -o $@ $(TRIMSLASH_OBJ) $(LIBS)
 
-T_UNSAFE_OBJ = t_unsafe.o syscall.o util.o t_stub.o lib/compat.o lib/snprintf.o
+T_UNSAFE_OBJ = t_unsafe.o syscall.o util.o util2.o t_stub.o lib/compat.o lib/snprintf.o lib/wildmatch.o
 t_unsafe$(EXEEXT): $(T_UNSAFE_OBJ)
 	$(CC) $(CFLAGS) $(LDFLAGS) -o $@ $(T_UNSAFE_OBJ) $(LIBS)
 
-gen: $(srcdir)/configure $(srcdir)/config.h.in proto man
+.PHONY: conf
+conf: configure.sh config.h.in
 
-man: $(srcdir)/rsync.1 $(srcdir)/rsyncd.conf.5
+.PHONY: gen
+gen: conf proto.h man
 
-$(srcdir)/configure: $(srcdir)/configure.in $(srcdir)/aclocal.m4
-	cd $(srcdir); autoconf
+.PHONY: gensend
+gensend: gen
+	rsync -aic $(GENFILES) $${SAMBA_HOST-samba.org}:/home/ftp/pub/rsync/generated-files/
 
-$(srcdir)/config.h.in: $(srcdir)/configure.in $(srcdir)/aclocal.m4
-	cd $(srcdir); autoheader
+aclocal.m4: $(srcdir)/m4/*.m4
+	aclocal -I $(srcdir)/m4
 
-$(srcdir)/rsync.1: $(srcdir)/rsync.yo
-	yodl2man -o $(srcdir)/rsync.1 $(srcdir)/rsync.yo
-
-$(srcdir)/rsyncd.conf.5: $(srcdir)/rsyncd.conf.yo
-	yodl2man -o $(srcdir)/rsyncd.conf.5 $(srcdir)/rsyncd.conf.yo
-
-proto:
-	cat $(srcdir)/*.c $(srcdir)/lib/compat.c | awk -f $(srcdir)/mkproto.awk >$(srcdir)/proto.h.new
-	if diff $(srcdir)/proto.h $(srcdir)/proto.h.new >/dev/null; then \
-	  rm $(srcdir)/proto.h.new; \
+configure.sh config.h.in: configure.ac aclocal.m4
+	@if test -f configure.sh; then cp -p configure.sh configure.sh.old; else touch configure.sh.old; fi
+	@if test -f config.h.in; then cp -p config.h.in config.h.in.old; else touch config.h.in.old; fi
+	autoconf -o configure.sh
+	autoheader && touch config.h.in
+	@if diff configure.sh configure.sh.old >/dev/null 2>&1; then \
+	    echo "configure.sh is unchanged."; \
+	    rm configure.sh.old; \
 	else \
-	  mv $(srcdir)/proto.h.new $(srcdir)/proto.h; \
+	    echo "configure.sh has CHANGED."; \
+	fi
+	@if diff config.h.in config.h.in.old >/dev/null 2>&1; then \
+	    echo "config.h.in is unchanged."; \
+	    rm config.h.in.old; \
+	else \
+	    echo "config.h.in has CHANGED."; \
+	fi
+	@if test -f configure.sh.old -o -f config.h.in.old; then \
+	    if test "$(MAKECMDGOALS)" = reconfigure; then \
+		echo 'Continuing with "make reconfigure".'; \
+	    else \
+		echo 'You may need to run:'; \
+		echo '  make reconfigure'; \
+		exit 1; \
+	    fi \
 	fi
 
-clean: cleantests
-	rm -f *~ $(OBJS) $(TLS_OBJ) $(CHECK_PROGS) $(CHECK_OBJS)
+.PHONY: reconfigure
+reconfigure: configure.sh
+	./config.status --recheck
+	./config.status
 
+.PHONY: restatus
+restatus:
+	./config.status
+
+Makefile: Makefile.in config.status configure.sh config.h.in
+	@if test -f Makefile; then cp -p Makefile Makefile.old; else touch Makefile.old; fi
+	@./config.status
+	@if diff Makefile Makefile.old >/dev/null 2>&1; then \
+	    echo "Makefile is unchanged."; \
+	    rm Makefile.old; \
+	else \
+	    if test "$(MAKECMDGOALS)" = reconfigure; then \
+		echo 'Continuing with "make reconfigure".'; \
+	    else \
+		echo "Makefile updated -- rerun your make command."; \
+		exit 1; \
+	    fi \
+	fi
+
+stunnel-rsyncd.conf: $(srcdir)/stunnel-rsyncd.conf.in Makefile
+	sed 's;\@bindir\@;$(bindir);g' <$(srcdir)/stunnel-rsyncd.conf.in >stunnel-rsyncd.conf
+
+.PHONY: proto
+proto: proto.h-tstamp
+
+proto.h: proto.h-tstamp
+	@if test -f proto.h; then :; else cp -p $(srcdir)/proto.h .; fi
+
+proto.h-tstamp: $(srcdir)/*.c $(srcdir)/lib/compat.c config.h
+	$(AWK) -f $(srcdir)/mkproto.awk $(srcdir)/*.c $(srcdir)/lib/compat.c
+
+.PHONY: man
+man: rsync.1 rsync-ssl.1 rsyncd.conf.5
+
+rsync.1: rsync.1.md md2man version.h Makefile
+	@$(srcdir)/maybe-make-man $(srcdir) rsync.1.md
+
+rsync-ssl.1: rsync-ssl.1.md md2man version.h Makefile
+	@$(srcdir)/maybe-make-man $(srcdir) rsync-ssl.1.md
+
+rsyncd.conf.5: rsyncd.conf.5.md md2man version.h Makefile
+	@$(srcdir)/maybe-make-man $(srcdir) rsyncd.conf.5.md
+
+.PHONY: clean
+clean: cleantests
+	rm -f *~ $(OBJS) $(CHECK_PROGS) $(CHECK_OBJS) $(CHECK_SYMLINKS) \
+		rounding rounding.h *.old rsync*.1 rsync*.5 rsync*.html
+
+.PHONY: cleantests
 cleantests:
 	rm -rf ./testtmp*
 
 # We try to delete built files from both the source and build
 # directories, just in case somebody previously configured things in
 # the source directory.
+.PHONY: distclean
 distclean: clean
 	rm -f Makefile config.h config.status
+	rm -f stunnel-rsyncd.conf
+	rm -f lib/dummy popt/dummy zlib/dummy
 	rm -f $(srcdir)/Makefile $(srcdir)/config.h $(srcdir)/config.status
-
+	rm -f $(srcdir)/lib/dummy $(srcdir)/popt/dummy $(srcdir)/zlib/dummy
 	rm -f config.cache config.log
 	rm -f $(srcdir)/config.cache $(srcdir)/config.log
-
 	rm -f shconfig $(srcdir)/shconfig
+	rm -f $(GENFILES)
+	rm -rf autom4te.cache
 
 # this target is really just for my use. It only works on a limited
 # range of machines and is used to produce a list of potentially
 # dead (ie. unused) functions in the code. (tridge)
+.PHONY: finddead
 finddead:
 	nm *.o */*.o |grep 'U ' | awk '{print $$2}' | sort -u > nmused.txt
 	nm *.o */*.o |grep 'T ' | awk '{print $$3}' | sort -u > nmfns.txt
 	comm -13 nmused.txt nmfns.txt
 
 # 'check' is the GNU name, 'test' is the name for everybody else :-)
-.PHONY: check test
-
+.PHONY: test
 test: check
 
-
 # There seems to be no standard way to specify some variables as
 # exported from a Makefile apart from listing them like this.
 
@@ -157,18 +291,37 @@ test: check
 # catch Bash-isms earlier even if we're running on GNU.  Of course, we
 # might lose in the future where POSIX diverges from old sh.
 
-check: all $(CHECK_PROGS)
-	POSIXLY_CORRECT=1 TOOLDIR=`pwd` rsync_bin=`pwd`/rsync$(EXEEXT) srcdir="$(srcdir)" $(srcdir)/runtests.sh
+.PHONY: check
+check: all $(CHECK_PROGS) $(CHECK_SYMLINKS)
+	rsync_bin=`pwd`/rsync$(EXEEXT) $(srcdir)/runtests.sh
 
-wildtest.o: wildtest.c lib/wildmatch.c rsync.h
-wildtest$(EXEEXT): wildtest.o
-	$(CC) $(CFLAGS) $(LDFLAGS) -o $@ wildtest.o @BUILD_POPT@ $(LIBS)
+.PHONY: check29
+check29: all $(CHECK_PROGS) $(CHECK_SYMLINKS)
+	rsync_bin=`pwd`/rsync$(EXEEXT) $(srcdir)/runtests.sh --protocol=29
+
+.PHONY: check30
+check30: all $(CHECK_PROGS) $(CHECK_SYMLINKS)
+	rsync_bin=`pwd`/rsync$(EXEEXT) $(srcdir)/runtests.sh --protocol=30
+
+wildtest.o: wildtest.c t_stub.o lib/wildmatch.c rsync.h config.h
+wildtest$(EXEEXT): wildtest.o lib/compat.o lib/snprintf.o @BUILD_POPT@
+	$(CC) $(CFLAGS) $(LDFLAGS) -o $@ wildtest.o lib/compat.o lib/snprintf.o @BUILD_POPT@ $(LIBS)
+
+testsuite/chown-fake.test:
+	ln -s chown.test $(srcdir)/testsuite/chown-fake.test
+
+testsuite/devices-fake.test:
+	ln -s devices.test $(srcdir)/testsuite/devices-fake.test
+
+testsuite/xattrs-hlink.test:
+	ln -s xattrs.test $(srcdir)/testsuite/xattrs-hlink.test
 
 # This does *not* depend on building or installing: you can use it to
 # check a version installed from a binary or some other source tree,
 # if you want.
 
-installcheck: $(CHECK_PROGS)
+.PHONY: installcheck
+installcheck: $(CHECK_PROGS) $(CHECK_SYMLINKS)
 	POSIXLY_CORRECT=1 TOOLDIR=`pwd` rsync_bin="$(bindir)/rsync$(EXEEXT)" srcdir="$(srcdir)" $(srcdir)/runtests.sh
 
 # TODO: Add 'dist' target; need to know which files will be included
@@ -178,21 +331,12 @@ installcheck: $(CHECK_PROGS)
 splint:
 	splint +unixlib +gnuextensions -weak rsync.c
 
-
-rsync.dvi: doc/rsync.texinfo
-	texi2dvi -o $@ $<
-
-rsync.ps: rsync.dvi
-	dvips -ta4 -o $@ $<
-
-rsync.pdf: doc/rsync.texinfo
-	texi2dvi -o $@ --pdf $<
-
-
+.PHONY: doxygen
 doxygen:
 	cd $(srcdir) && rm dox/html/* && doxygen
 
 # for maintainers only
+.PHONY: doxygen-upload
 doxygen-upload:
 	rsync -avzv $(srcdir)/dox/html/ --delete \
-	samba.org:/home/httpd/html/rsync/doxygen/head/
+	$${SAMBA_HOST-samba.org}:/home/httpd/html/rsync/doxygen/head/

NEWS

@@ -1,384 +0,0 @@
-NEWS for rsync 2.6.4 (30 March 2005)
-Protocol: 29 (changed)
-Changes since 2.6.3:
-
-  OUTPUT CHANGES:
-
-    - When rsync deletes a directory and outputs a verbose message about
-      it, it now appends a trailing slash to the name instead of (only
-      sometimes) outputting a preceding "directory " string.
-
-    - The --stats output will contain file-list time-statistics if both
-      sides are 2.6.4, or if the local side is 2.6.4 and the files are
-      being pushed (since the stats come from the sending side).
-      (Requires protocol 29 for a pull.)
-
-    - The "%o" (operation) log-format escape now has a third value (besides
-      "send" and "recv"):  "del." (with trailing dot to make it 4 chars).
-      This changes the way deletions are logged in the daemon's log file.
-
-    - When the --log-format option is combined with --verbose, rsync now
-      avoids outputting the name of the file twice in most circumstances.
-      As long as the --log-format item does not refer to any post-transfer
-      items (such as %b or %c), the --log-format message is output prior to
-      the transfer, so --verbose is now the equivalent of a --log-format of
-      '%n%L' (which outputs the name and any link info).  If the log output
-      must occur after the transfer to be complete, the only time the name
-      is also output prior to the transfer is when --progress was specified
-      (so that the name will precede the progress stats, and the full
-      --log-format output will come after).
-  
-  BUG FIXES:
-
-    - Restore the list-clearing behavior of "!" in a .cvsignore file (2.6.3
-      was only treating it as a special token in an rsync include/exclude
-      file).
-
-    - The combination of --verbose and --dry-run now mentions the full list
-      of changes that would be output without --dry-run.
-
-    - Avoid a mkdir warning when removing a directory in the destination
-      that already exists in the --backup-dir.
-
-    - An OS that has a binary mode for its files (such as cygwin) needed
-      setmode(fd, O_BINARY) called on the temp-file we opened with
-      mkstemp().  (Fix derived from the cygwin's 2.6.3 rsync package.)
-
-    - Fixed a potential hang when verbosity is high, the client side is
-      the sender, and the file-list is large.
-
-    - Fixed a potential protocol-corrupting bug where the generator could
-      merge a message from the receiver into the middle of a multiplexed
-      packet of data if only part of that data had been written out to the
-      socket when the message from the generator arrived.
-
-    - We now check if the OS doesn't support using mknod() for creating
-      FIFOs and sockets, and compile-in some compatibility code using
-      mkfifo() and socket() when necessary.
-
-    - Fixed an off-by-one error in the handling of --max-delete=N.  Also,
-      if the --max-delete limit is exceeded during a run, we now output a
-      warning about this at the end of the run and exit with a new error
-      code (25).
-
-    - One place in the code wasn't checking if fork() failed.
-
-    - The "ignore nonreadable" daemon parameter used to erroneously affect
-      readable symlinks that pointed to a non-existent file.
-
-    - If the OS does not have lchown() and a chown() of a symlink will
-      affect the referent of a symlink (as it should), we no longer try
-      to set the user and group of a symlink.
-
-    - The generator now properly runs the hard-link loop and the dir-time
-      rewriting loop after we're sure that the redo phase is complete.
-
-    - When --backup was specified with --partial-dir=DIR, where DIR is a
-      relative path, the backup code was erroneously trying to backup a
-      file that was put into the partial-dir.
-
-    - If a file gets resent in a single transfer and the --backup option is
-      enabled along with --inplace, rsync no longer performs a duplicate
-      backup (it used to overwrite the first backup with the failed file).
-
-    - One call to flush_write_file() was not being checked for an error.
-
-    - The --no-relative option was not being sent from the client to a
-      server sender.
-
-    - If an rsync daemon specified "dont compress = ..." for a file and the
-      client tried to specify --compress, the libz code was not handling a
-      compression level of 0 properly.  This could cause a transfer failure
-      if the block-size for a file was large enough (e.g. rsync might have
-      exited with an error for large files).
-
-    - Fixed a bug that would sometimes surface when using --compress and
-      sending a file with a block-size larger than 64K (either manually
-      specified, or computed due to the file being really large).  Prior
-      versions of rsync would sometimes fail to decompress the data
-      properly, and thus the transferred file would fail its verification.
-
-    - If a daemon can't open the specified log file (i.e. syslog is not
-      being used), die without crashing.  We also output an error about
-      the failure on stderr (which will only be seen if --no-detach was
-      specified) and exit with a new error code (6).
-
-    - A local transfer no longer duplicates all its include/exclude options
-      (since the forked process already has a copy of the exclude list,
-      there's no need to send them a set of duplicates).
-
-    - When --progress is specified, the output of items that the generator
-      is creating (e.g. dirs, symlinks) is now integrated into the progress
-      output without overlapping it.  (Requires protocol 29.)
-
-    - When --timeout is specified, lulls that occur in the transfer while
-      the generator is doing work that does not generate socket traffic
-      (looking for changed files, deleting files, doing directory-time
-      touch-ups, etc.) will cause a new keep-alive packet to be sent that
-      should keep the transfer going as long as the generator continues to
-      make progress.  (Requires protocol 29.)
-
-    - The stat size of a device is not added to the total file size of the
-      items in the transfer (the size might be undefined on some OSes).
-
-    - Fixed a problem with refused-option messages sometimes not making it
-      back to the client side when a remote --files-from was in effect and
-      the daemon was the receiver.
-
-    - The --compare-dest option was not updating a file that differred in
-      (the preserved) attributes from the version in the compare-dest DIR.
-
-    - When rsync is copying files into a write-protected directory, fixed
-      the change-report output for the directory so that we don't report
-      an identical directory as changed.
-
-  ENHANCEMENTS:
-
-    - Rsync now supports popt's option aliases, which means that you can
-      use /etc/popt and/or ~/.popt to create your own option aliases.
-
-    - Added the --delete-during (--del) option which will delete files
-      from the receiving side incrementally as each directory in the
-      transfer is being processed.  This makes it more efficient than the
-      default, before-the-transfer behavior, which is now also available as
-      --delete-before (and is still the default --delete-WHEN option that
-      will be chosen if --delete or --delete-excluded is specified without
-      a --delete-WHEN choice).  All the --del* options infer --delete, so
-      an rsync daemon that refuses "delete" will still refuse to allow any
-      file-deleting options (including the new --remove-sent-files option).
-
-    - All the --delete-WHEN options are now more memory efficient:
-      Previously an duplicate set of file-list objects was created on the
-      receiving side for the entire destination hierarchy.  The new
-      algorithm only creates one directory of objects at a time (for files
-      inside the transfer).
-
-    - Added the --copy-dest option, which works like --link-dest except
-      that it locally copies identical files instead of hard-linking them.
-
-    - Added support for specifying multiple --compare-dest, --copy-dest, or
-      --link-dest options, but only of a single type. (Promoted from the
-      patches dir and enhanced.) (Requires protocol 29.)
-
-    - Added the --max-size option. (Promoted from the patches dir.)
-
-    - The daemon-mode options are now separated from the normal rsync
-      options so that they can't be mixed together.  This makes it
-      impossible to start a daemon that has improper default option values
-      (which could cause problems when a client connects, such as hanging
-      or crashing).
-
-    - The --bwlimit option may now be used in combination with --daemon
-      to specify both a default value for the daemon side and a value
-      that cannot be exceeded by a user-specified --bwlimit option.
-
-    - Added the "port" parameter to the rsyncd.conf file. (Promoted from
-      the patches dir.)  Also added "address".  The command-line options
-      take precedence over a config-file option, as expected.
-
-    - In _exit_cleanup(): when we are exiting with a partially-received
-      file, we now flush any data in the write-cache before closing the
-      partial file.
-
-    - The --inplace support was enhanced to work with --compare-dest,
-      --link-dest, and (the new) --copy-dest options. (Requires protocol
-      29.)
-
-    - Added the --dirs (-d) option for an easier way to copy directories
-      without recursion.
-
-    - Added the --list-only option, which is mainly a way for the client to
-      put the server into listing mode without needing to resort to any
-      internal option kluges (e.g. the age-old use of "-r --exclude="/*/*"
-      for a non-recursive listing).  This option is used automatically
-      (behind the scenes) when a modern rsync speaks to a modern daemon,
-      but may also be specified manually if you want to force the use of
-      the --list-only option over a remote-shell connection.
-
-    - Added the --omit-dir-times (-O) option, which will avoid updating
-      the modified time for directories when --times was specified.  This
-      option will avoid an extra pass through the file-list at the end of
-      the transfer (to tweak all the directory times), which may provide
-      an appreciable speedup for a really large transfer. (Promoted from
-      the patches dir.)
-
-    - Added the --filter (-f) option and its helper option, -F.  Filter
-      rules are an extension to the existing include/exclude handling
-      that also supports nested filter files as well as per-directory
-      filter files (like .cvsignore, but with full filter-rule parsing).
-      This new option was chosen in order to ensure that all existing
-      include/exclude processing remained 100% compatible with older
-      versions.  Protocol 29 is needed for full filter-rule support, but
-      backward-compatible rules work with earlier protocol versions.
-      (Promoted from the patches dir and enhanced.)
-
-    - Added the --delay-updates option that puts all updated files into
-      a temporary directory (by default ".~tmp~", but settable via the
-      --partial-dir=DIR option) until the end of the transfer.  This
-      makes the updates a little more atomic for a large transfer.
-
-    - If rsync is put into the background, any output from --progress is
-      reduced.
-
-    - Documented the "max verbosity" setting for rsyncd.conf.  (This
-      setting was added a couple releases ago, but left undocumented.)
-
-    - The sender and the generator now double-check the file-list index
-      they are given, and refuse to try to do a file transfer on a
-      non-file index (since that would indicate that something had gone
-      very wrong).
-
-    - Added the --itemize-changes (-i) option, which is a way to output a
-      more detailed list of what files changed and in what way.  The effect
-      is the same as specifying a --log-format of "%i %n%L" (see both the
-      rsync and rsyncd.conf manpages).  Works with --dry-run too.
-
-    - Added the --fuzzy (-y) option, which attempts to find a basis file
-      for a file that is being created from scratch.  The current algorithm
-      only looks in the destination directory for the created file, but it
-      does attempt to find a match based on size/mod-time (in case the file
-      was renamed with no other changes) as well as based on a fuzzy
-      name-matching algorithm.  This option requires protocol 29 because it
-      needs the new file-sorting order.  (Promoted from patches dir and
-      enhanced.) (Requires protocol 29.)
-
-    - Added the --remove-sent-files option, which lets you move files
-      between systems.
-
-    - The hostname in HOST:PATH or HOST::PATH may now be an IPv6 literal
-      enclosed in '[' and ']' (e.g. "[::1]").  (We already allowed IPv6
-      literals in the rsync://HOST:PORT/PATH format.)
-
-    - When rsync recurses to build the file list, it no longer keeps open
-      one or more directory handles from the dir's parent dirs.
-
-    - When building under windows, the default for --daemon is now to
-      avoid detaching, requiring the new --detach option to force rsync
-      to detach.
-
-    - The --dry-run option can now be combined with either --write-batch or
-      --read-batch, allowing you to run a do-nothing test command to see
-      what would happen without --dry-run.
-
-    - The daemon's "read only" config item now sets an internal read_only
-      variable that makes extra sure that no write/delete calls on the
-      read-only side can succeed.
-
-    - The log-format % escapes can now have a numeric field width in
-      between the % and the escape letter (e.g. "%-40n %08p").
-
-    - Improved the option descriptions in the --help text.
-
-  SUPPORT FILES:
-
-    - Added atomic-rsync to the support dir: a perl script that will
-      transfer some files using rsync, and then move the updated files into
-      place all at once at the end of the transfer.  Only works when
-      pulling, and uses --link-dest and a parallel hierarchy of files to
-      effect its update.
-
-    - Added mnt-excl to the support dir: a perl script that takes the
-      /proc/mounts file and translates it into a set of excludes that will
-      exclude all mount points (even mapped mounts to the same disk).  The
-      excludes are made relative to the specified source dir and properly
-      anchored.
-
-    - Added savetransfer.c to the support dir: a C program that can make
-      a copy of all the data that flows over the wire.  This lets you test
-      for data corruption (by saving the data on both the sending side and
-      the receiving side) and provides one way to debug a protocol error.
-
-    - Added rrsync to the support dir: this is an updated version of Joe
-      Smith's restricted rsync perl script.  This helps to ensure that only
-      certain rsync commands can be run by an ssh invocation.
-
-  INTERNAL:
-
-    - Added better checking of the checksum-header values that come over
-      the socket.
-
-    - Merged a variety of file-deleting functions into a single function so
-      that it is easier to maintain.
-
-    - Improved the type of some variables (particularly blocksize vars) for
-      consistency and proper size.
-
-    - Got rid of the uint64 type (which we didn't need).
-
-    - Use a slightly more compatible set of core #include directives.
-
-    - Defined int32 in a way that ensures that the build dies if we can't
-      find a variable with at least 32 bits.
-
-  PROTOCOL DIFFERENCES FOR VERSION 29:
-
-    - A 16-bit flag-word is transmitted after every file-list index.  This
-      indicates what is changing between the sender and the receiver.  The
-      generator now transmits an index and a flag-word to indicate when
-      dirs and symlinks have changed (instead of producing a message),
-      which makes the outputting of the information more consistent and
-      less prone to screen corruption (because the local receiver/sender is
-      now outputting all the file-change info messages).
-
-    - If a file is being hard-linked, the ITEM_XNAME_FOLLOWS bit is enabled
-      in the flag-word and the name of the file that was linked immediately
-      follows in vstring format (see below).
-
-    - If a file is being transferred with an alternate-basis file, the
-      ITEM_BASIS_TYPE_FOLLOWS bit is enabled in the flag-word and a single
-      byte follows, indicating what type of basis file was chosen.  If that
-      indicates that a fuzzy-match was selected, the ITEM_XNAME_FOLLOWS bit
-      is set in the flag-word and the name of the match in vstring format
-      follows the basis byte.  A vstring is a variable length string that
-      has its size written prior to the string, and no terminating null.
-      If the string is from 1-127 bytes, the length is a single byte.  If
-      it is from 128-32767 bytes, the length is written as ((len >> 8) |
-      0x80) followed by (len % 0x100).
-
-    - The sending of exclude names is done using filter-rule syntax.  This
-      means that all names have a prefixed rule indicator, even excludes
-      (which used to be sent as a bare pattern, when possible).  The -C
-      option will include the per-dir .cvsignore merge file in the list of
-      filter rules so it is positioned correctly (unlike in some older
-      transfer scenarios).
-
-    - Rsync sorts the filename list in a different way: it sorts the subdir
-      names after the non-subdir names for each dir's contents, and it
-      always puts a dir's contents immediately after the dir's name in the
-      list.  (Previously an item named "foo.txt" would sort in between
-      directory "foo/" and "foo/bar".)
-
-    - When talking to a protocol 29 rsync daemon, a list-only request
-      is able to note this before the options are sent over the wire and
-      the new --list-only option is included in the options.
-
-    - When the --stats bytes are sent over the wire (or stored in a batch),
-      they now include two elapsed-time values: one for how long it took to
-      build the file-list, and one for how long it took to send it over the
-      wire (each expressed in thousandths of a second).
-
-    - When --delete-excluded is specified with some filter rules (AKA
-      excludes), a client sender will now initiate a send of the rules to
-      the receiver (older protocols used to omit the sending of excludes in
-      this situation since there were no receiver-specific rules that
-      survived --delete-excluded back then).  Note that, as with all the
-      filter-list sending, only items that are significant to the other
-      side will actually be sent over the wire, so the filter-rule list
-      that is sent in this scenario is often empty.
-
-    - An index equal to the file-list count is sent as a keep-alive packet
-      from the generator to the sender, which then forwards it on to the
-      receiver.  This normally invalid index is only a valid keep-alive
-      packet if the 16-bit flag-word that follows it contains a single bit
-      (ITEM_IS_NEW, which is normally an illegal flag to appear alone).
-
-    - A protocol-29 batch file includes a bit for the setting of the --dirs
-      option and for the setting of the --compress option.  Also, the shell
-      script created by --write-batch will use the --filter option instead
-      of --exclude-from to capture any filter rules.
-
-  BUILD CHANGES:
-
-    - Handle an operating system that use mkdev() in place of makedev().
-
-    - Improved configure to better handle cross-compiling.

README

@@ -1,124 +0,0 @@
-WHAT IS RSYNC?
---------------
-
-rsync is a replacement for scp/rcp that has many more features.
-
-rsync uses the "rsync algorithm" which provides a very fast method for
-bringing remote files into sync. It does this by sending just the
-differences in the files across the link, without requiring that both
-sets of files are present at one of the ends of the link beforehand.
-At first glance this may seem impossible because the calculation of
-diffs between two files normally requires local access to both
-files. 
-
-A technical report describing the rsync algorithm is included with
-this package. 
-
-
-USAGE
------
-
-Basically you use rsync just like rcp, but rsync has many additional
-options.  To get a complete list of supported options type
-
-  rsync --help
-
-and see the manual for more information.
-
-
-SETUP
------
-
-Rsync normally uses ssh or rsh for communication.  It does not need to
-be setuid and requires no special privileges for installation.  You
-must, however, have a working ssh or rsh system.  Using ssh is
-recommended for its security features.
-
-Alternatively, rsync can run in `daemon' mode, listening on a socket.
-This is generally used for public file distribution, although
-authentication and access control are available.
-
-To install rsync, first run the "configure" script. This will create a
-Makefile and config.h appropriate for your system. Then type
-"make". 
-
-Note that on some systems you will have to force configure not to use
-gcc because gcc may not support some features (such as 64 bit file
-offsets) that your system may support. Set the environment variable CC
-to the name of your native compiler before running configure in this
-case. 
-
-Once built put a copy of rsync in your search path on the local and
-remote systems (or use "make install"). That's it!
-
-RSYNC SERVERS
--------------
-
-rsync can also talk to "rsync servers" which can provide anonymous or
-authenticated rsync. See the rsyncd.conf(5) man page for details on how
-to setup a rsync server. See the rsync(1) man page for info on how to
-connect to a rsync server.
-
-
-MAILING LIST
-------------
-
-There is a mailing list for the discussion of rsync and its
-applications. It is open to anyone to join. I will announce new
-versions on this list. 
-
-To join the mailing list see the web page at http://lists.samba.org/
-
-To send mail to everyone on the list send it to rsync@samba.org
-
-
-BUG REPORTS
------------
-
-If you have web access then please look at
-
-  http://rsync.samba.org/
-
-That page contains links to the current bug list, and information on
-how to report a bug well.  You might also like to try searching the
-internet for the error message you've received, or looking in the
-mailing list archives at
-
-  http://mail-archive.com/rsync@lists.samba.org/
-
-To send a bug report, follow the instructions on the bug-tracking
-page of the web site.
-
-
-CVS TREE
---------
-
-If you want to get the very latest version of rsync direct from the
-source code repository then you can use anonymous cvs. You will need a
-recent version of cvs then use the following commands:
-
-	cvs -d :pserver:cvs@pserver.samba.org:/cvsroot login
-	Password: cvs
-
-	cvs -d :pserver:cvs@pserver.samba.org:/cvsroot co rsync
-
-Look at the cvs documentation for more details.
-
-
-COPYRIGHT
----------
-
-rsync was originally written by Andrew Tridgell and has been improved
-by many developers around the world.  rsync may be used, modified and
-redistributed only under the terms of the GNU General Public License,
-found in the file COPYING in this distribution, or at 
-
-  http://www.fsf.org/licenses/gpl.html
-
-
-AVAILABILITY
-------------
-
-The main web site for rsync is http://rsync.samba.org/
-The main ftp site is ftp://rsync.samba.org/pub/rsync/
-This is also available as rsync://rsync.samba.org/rsyncftp/

README.md

@@ -0,0 +1,142 @@
+WHAT IS RSYNC?
+--------------
+
+Rsync is a fast and extraordinarily versatile file copying tool for
+both remote and local files.
+
+Rsync uses a delta-transfer algorithm which provides a very fast method
+for bringing remote files into sync.  It does this by sending just the
+differences in the files across the link, without requiring that both
+sets of files are present at one of the ends of the link beforehand.  At
+first glance this may seem impossible because the calculation of diffs
+between two files normally requires local access to both files.
+
+A technical report describing the rsync algorithm is included with this
+package.
+
+
+USAGE
+-----
+
+Basically you use rsync just like scp, but rsync has many additional
+options.  To get a complete list of supported options type:
+
+    rsync --help
+
+See the manpage for more detailed information.
+
+
+SETUP
+-----
+
+Rsync normally uses ssh or rsh for communication with remote systems.
+It does not need to be setuid and requires no special privileges for
+installation.  You must, however, have a working ssh or rsh system.
+Using ssh is recommended for its security features.
+
+Alternatively, rsync can run in `daemon' mode, listening on a socket.
+This is generally used for public file distribution, although
+authentication and access control are available.
+
+To install rsync, first run the "configure" script.  This will create a
+Makefile and config.h appropriate for your system.  Then type "make".
+
+Note that on some systems you will have to force configure not to use
+gcc because gcc may not support some features (such as 64 bit file
+offsets) that your system may support.  Set the environment variable CC
+to the name of your native compiler before running configure in this
+case.
+
+Once built put a copy of rsync in your search path on the local and
+remote systems (or use "make install").  That's it!
+
+
+RSYNC DAEMONS
+-------------
+
+Rsync can also talk to "rsync daemons" which can provide anonymous or
+authenticated rsync.  See the rsyncd.conf(5) man page for details on how
+to setup an rsync daemon.  See the rsync(1) man page for info on how to
+connect to an rsync daemon.
+
+
+WEB SITE
+--------
+
+The main rsync web site is here:
+
+>  https://rsync.samba.org/
+
+You'll find a FAQ list, downloads, resources, HTML versions of the
+manpages, etc.
+
+
+MAILING LISTS
+-------------
+
+There is a mailing list for the discussion of rsync and its applications
+that is open to anyone to join.  New releases are announced on this
+list, and there is also an announcement-only mailing list for those that
+want official announcements.  See the mailing-list page for full
+details:
+
+>  https://rsync.samba.org/lists.html
+
+
+BUG REPORTS
+-----------
+
+To visit this web page for full the details on bug reporting:
+
+>  https://rsync.samba.org/bugtracking.html
+
+That page contains links to the current bug list, and information on how
+to report a bug well.  You might also like to try searching the Internet
+for the error message you've received, or looking in the mailing list
+archives at:
+
+>  https://mail-archive.com/rsync@lists.samba.org/
+
+To send a bug report, follow the instructions on the bug-tracking
+page of the web site.
+
+Alternately, email your bug report to <rsync@lists.samba.org>.
+
+
+GIT REPOSITORY
+--------------
+
+If you want to get the very latest version of rsync direct from the
+source code repository, then you will need to use git.  The git repo
+is hosted on github and on samba's site.  Feel free to access it here:
+
+>  https://github.com/WayneD/rsync
+
+A backup git repo is available on the samba site:
+
+>     git clone git://git.samba.org/rsync.git
+
+See the download page for full details on all the ways to grab the
+source:
+
+>  https://rsync.samba.org/download.html
+
+
+COPYRIGHT
+---------
+
+Rsync was originally written by Andrew Tridgell and is currently
+maintained by Wayne Davison.  It has been improved by many developers
+from around the world.
+
+Rsync may be used, modified and redistributed only under the terms of
+the GNU General Public License, found in the file COPYING in this
+distribution, or at:
+
+>  https://www.fsf.org/licenses/gpl.html
+
+
+AVAILABILITY
+------------
+
+The main web site for rsync is https://rsync.samba.org/

TODO

@@ -1,22 +1,17 @@
 -*- indented-text -*-
 
-BUGS ---------------------------------------------------------------
-Do not rely on having a group called "nobody"
-
 FEATURES ------------------------------------------------------------
 Use chroot only if supported
 Allow supplementary groups in rsyncd.conf			2002/04/09
 Handling IPv6 on old machines
-Other IPv6 stuff:
+Other IPv6 stuff
 Add ACL support							2001/12/02
-Lazy directory creation
 proxy authentication						2002/01/23
 SOCKS								2002/01/23
 FAT support
-Allow forcing arbitrary permissions				2002/03/12
 --diff						david.e.sewell	2002/03/15
 Add daemon --no-fork option
-Create more granular verbosity				jw	2003/05/15
+Create more granular verbosity					2003/05/15
 
 DOCUMENTATION --------------------------------------------------------
 Keep list of open issues and todos on the web site
@@ -25,22 +20,19 @@ Perhaps redo manual as SGML
 LOGGING --------------------------------------------------------------
 Memory accounting
 Improve error messages
-Better statistics:					Rasmus	2002/03/08
+Better statistics					Rasmus	2002/03/08
 Perhaps flush stdout like syslog
-Log deamon sessions that just list modules
 Log child death on signal
-Log errors with function that reports process of origin
 verbose output					David Stein	2001/12/20
 internationalization
 
 DEVELOPMENT --------------------------------------------------------
 Handling duplicate names
 Use generic zlib						2002/02/25
-TDB:								2002/03/12
+TDB								2002/03/12
 Splint								2002/03/12
 
 PERFORMANCE ----------------------------------------------------------
-File list structure in memory
 Traverse just one directory at a time
 Allow skipping MD4 file_sum					2002/04/08
 Accelerate MD4
@@ -52,31 +44,18 @@ Test on kernel source
 Test large files
 Create mutator program for testing
 Create configure option to enable dangerous tests
-If tests are skipped, say why.
-Test daemon feature to disallow particular options.
 Create pipe program for testing
 Create test makefile target for some tests
 
 RELATED PROJECTS -----------------------------------------------------
 rsyncsh
-http://rsync.samba.org/rsync-and-debian/
+https://rsync.samba.org/rsync-and-debian/
 rsyncable gzip patch
 rsyncsplit as alternative to real integration with gzip?
 reverse rsync over HTTP Range
 
 
 
-BUGS ---------------------------------------------------------------
-
-
-Do not rely on having a group called "nobody"
-
-  http://www.linuxbase.org/spec/refspecs/LSB_1.1.0/gLSB/usernames.html
-
-  On Debian it's "nogroup"
-
-                      --          --
-
 FEATURES ------------------------------------------------------------
 
 
@@ -87,8 +66,8 @@ Use chroot only if supported
   If running as non-root, then don't fail, just give a warning.
   (There was a thread about this a while ago?)
 
-    http://lists.samba.org/pipermail/rsync/2001-August/thread.html
-    http://lists.samba.org/pipermail/rsync/2001-September/thread.html
+    https://lists.samba.org/pipermail/rsync/2001-August/thread.html
+    https://lists.samba.org/pipermail/rsync/2001-September/thread.html
 
                       --          --
 
@@ -115,7 +94,7 @@ Handling IPv6 on old machines
   platforms that have a half-working implementation, so redefining
   these functions clashes with system headers, and leaving them out
   breaks.  This affects at least OSF/1, RedHat 5, and Cobalt, which
-  are moderately improtant.
+  are moderately important.
 
   Perhaps the simplest solution would be to have two different files
   implementing the same interface, and choose either the new or the
@@ -133,7 +112,7 @@ Handling IPv6 on old machines
                       --          --
 
 
-Other IPv6 stuff:
+Other IPv6 stuff
   
   Implement suggestions from http://www.kame.net/newsletter/19980604/
   and ftp://ftp.iij.ad.jp/pub/RFC/rfc2553.txt
@@ -159,15 +138,6 @@ Add ACL support							2001/12/02
                       --          --
 
 
-Lazy directory creation
-
-  With the current common --include '*/' --exclude '*' pattern, people
-  can end up with many empty directories.  We might avoid this by
-  lazily creating such directories.
-
-                      --          --
-
-
 proxy authentication						2002/01/23
 
   Allow RSYNC_PROXY to be http://user:pass@proxy.foo:3128/, and do
@@ -199,35 +169,6 @@ FAT support
                       --          --
 
 
-Allow forcing arbitrary permissions				2002/03/12
-
-  On 12 Mar 2002, Dave Dykstra <dwd@bell-labs.com> wrote:
-  > If we would add an option to do that functionality, I
-  > would vote for one that was more general which could mask
-  > off any set of permission bits and possibly add any set of
-  > bits.  Perhaps a chmod-like syntax if it could be
-  > implemented simply.
-
-  I think that would be good too.  For example, people uploading files
-  to a web server might like to say
-
-  rsync -avzP --chmod a+rX ./ sourcefrog.net:/home/www/sourcefrog/
-
-  Ideally the patch would implement as many of the gnu chmod semantics
-  as possible.  I think the mode parser should be a separate function
-  that passes back something like (mask,set) description to the rest
-  of the program.  For bonus points there would be a test case for the
-  parser.
-
-  Possibly also --chown
-
-  (Debian #23628)
-
-  NOTE: there is a patch that implements this in the "patches" subdir.
-
-                      --          --
-
-
 --diff						david.e.sewell	2002/03/15
 
   Allow people to specify the diff command.  (Might want to use wdiff,
@@ -252,7 +193,7 @@ Add daemon --no-fork option
                       --          --
 
 
-Create more granular verbosity				jw	2003/05/15
+Create more granular verbosity					2003/05/15
 
   Control output with the --report option.
 
@@ -263,7 +204,7 @@ Create more granular verbosity				jw	2003/05/15
   fine grained selection of statistical reporting and what
   actions are logged.
 
-  http://lists.samba.org/archive/rsync/2003-May/006059.html
+  https://lists.samba.org/archive/rsync/2003-May/006059.html
 
                       --          --
 
@@ -295,7 +236,7 @@ Memory accounting
 
   At exit, show how much memory was used for the file list, etc.
 
-  Also we do a wierd exponential-growth allocation in flist.c.  I'm
+  We also do a weird exponential-growth allocation in flist.c.  I'm
   not sure this makes sense with modern mallocs.  At any rate it will
   make us allocate a huge amount of memory for large file lists.
 
@@ -322,14 +263,10 @@ Improve error messages
   our load?  (Debian #28416) Probably fixed now, but a test case would
   be good.
 
-  When running as a daemon, some errors should both be returned to the
-  user and logged.  This will make interacting with a daemon less
-  cryptic.
-
                       --          --
 
 
-Better statistics:					Rasmus	2002/03/08
+Better statistics					Rasmus	2002/03/08
 
   <Rasmus>
       hey, how about an rsync option that just gives you the
@@ -350,15 +287,7 @@ Perhaps flush stdout like syslog
 
   Perhaps flush stdout after each filename, so that people trying to
   monitor progress in a log file can do so more easily.  See
-  http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=48108
-
-                      --          --
-
-
-Log deamon sessions that just list modules
-
-  At the connections that just get a list of modules are not logged,
-  but they should be.
+  https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=48108
 
                       --          --
 
@@ -371,15 +300,6 @@ Log child death on signal
                       --          --
 
 
-Log errors with function that reports process of origin
-
-  Use a separate function for reporting errors; prefix it with
-  "rsync:" or "rsync(remote)", or perhaps even "rsync(local
-  generator): ".
-
-                      --          --
-
-
 verbose output					David Stein	2001/12/20
   
   At end of transfer, show how many files were or were not transferred
@@ -445,21 +365,6 @@ Use generic zlib						2002/02/25
                       --          --
 
 
-TDB:								2002/03/12
-
-  Rather than storing the file list in memory, store it in a TDB.
-
-  This *might* make memory usage lower while building the file list.
-
-  Hashtable lookup will mean files are not transmitted in order,
-  though... hm.
-
-  This would neatly eliminate one of the major post-fork shared data
-  structures.
-
-                      --          --
-
-
 Splint								2002/03/12
 
   Build rsync with SPLINT to try to find security holes.  Add
@@ -472,31 +377,6 @@ Splint								2002/03/12
 
 PERFORMANCE ----------------------------------------------------------
 
-File list structure in memory
-
-  Rather than one big array, perhaps have a tree in memory mirroring
-  the directory tree.
-
-  This might make sorting much faster!  (I'm not sure it's a big CPU
-  problem, mind you.)
-
-  It might also reduce memory use in storing repeated directory names
-  -- again I'm not sure this is a problem.
-
-                      --          --
-
-
-Traverse just one directory at a time
-
-  Traverse just one directory at a time.  Tridge says it's possible.
-
-  At the moment rsync reads the whole file list into memory at the
-  start, which makes us use a lot of memory and also not pipeline
-  network access as much as we could.
-
-                      --          --
-
-
 Allow skipping MD4 file_sum					2002/04/08
 
   If we're doing a local transfer, or using -W, then perhaps don't
@@ -586,16 +466,6 @@ Create configure option to enable dangerous tests
                       --          --
 
 
-If tests are skipped, say why.
-
-                      --          --
-
-
-Test daemon feature to disallow particular options.
-
-                      --          --
-
-
 Create pipe program for testing
 
   Create pipe program that makes slow/jerky connections for
@@ -625,7 +495,7 @@ rsyncsh
                       --          --
 
 
-http://rsync.samba.org/rsync-and-debian/
+https://rsync.samba.org/rsync-and-debian/
 
 
                       --          --

access.c

@@ -1,37 +1,68 @@
-/* 
-   Copyright (C) Andrew Tridgell 1998
-   
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
-
 /*
-  hosts allow/deny code for rsync
-
-  */
+ * Routines to authenticate access to a daemon (hosts allow/deny).
+ *
+ * Copyright (C) 1998 Andrew Tridgell
+ * Copyright (C) 2004-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
 #include "rsync.h"
+#include "ifuncs.h"
 
+static int allow_forward_dns;
 
-static int match_hostname(char *host, char *tok)
+extern const char undetermined_hostname[];
+
+static int match_hostname(const char **host_ptr, const char *addr, const char *tok)
 {
+	struct hostent *hp;
+	unsigned int i;
+	const char *host = *host_ptr;
+
 	if (!host || !*host)
 		return 0;
-	return wildmatch(tok, host);
+
+	/* First check if the reverse-DNS-determined hostname matches. */
+	if (iwildmatch(tok, host))
+		return 1;
+
+	if (!allow_forward_dns)
+		return 0;
+
+	/* Fail quietly if tok is an address or wildcarded entry, not a simple hostname. */
+	if (!tok[strspn(tok, ".0123456789")] || tok[strcspn(tok, ":/*?[")])
+		return 0;
+
+	/* Now try forward-DNS on the token (config-specified hostname) and see if the IP matches. */
+	if (!(hp = gethostbyname(tok)))
+		return 0;
+
+	for (i = 0; hp->h_addr_list[i] != NULL; i++) {
+		if (strcmp(addr, inet_ntoa(*(struct in_addr*)(hp->h_addr_list[i]))) == 0) {
+			/* If reverse lookups are off, we'll use the conf-specified
+			 * hostname in preference to UNDETERMINED. */
+			if (host == undetermined_hostname)
+				*host_ptr = strdup(tok);
+			return 1;
+		}
+	}
+
+	return 0;
 }
 
-static int match_binary(char *b1, char *b2, char *mask, int addrlen)
+static int match_binary(const char *b1, const char *b2, const char *mask, int addrlen)
 {
 	int i;
 
@@ -60,7 +91,7 @@ static void make_mask(char *mask, int plen, int addrlen)
 	return;
 }
 
-static int match_address(char *addr, char *tok)
+static int match_address(const char *addr, const char *tok)
 {
 	char *p;
 	struct addrinfo hints, *resa, *rest;
@@ -74,24 +105,16 @@ static int match_address(char *addr, char *tok)
 #endif
 	char mask[16];
 	char *a = NULL, *t = NULL;
-	unsigned int len;
 
 	if (!addr || !*addr)
 		return 0;
 
 	p = strchr(tok,'/');
-	if (p) {
+	if (p)
 		*p = '\0';
-		len = p - tok;
-	} else
-		len = strlen(tok);
 
-	/* Fail quietly if tok is a hostname (not an address) */
-	if (strspn(tok, ".0123456789") != len
-#ifdef INET6
-	    && strchr(tok, ':') == NULL
-#endif
-	) {
+	/* Fail quietly if tok is a hostname, not an address. */
+	if (tok[strspn(tok, ".0123456789")] && strchr(tok, ':') == NULL) {
 		if (p)
 			*p = '/';
 		return 0;
@@ -134,8 +157,7 @@ static int match_address(char *addr, char *tok)
 		break;
 
 #ifdef INET6
-	case PF_INET6:
-	    {
+	case PF_INET6: {
 		struct sockaddr_in6 *sin6a, *sin6t;
 
 		sin6a = (struct sockaddr_in6 *)resa->ai_addr;
@@ -147,20 +169,19 @@ static int match_address(char *addr, char *tok)
 		addrlen = 16;
 
 #ifdef HAVE_SOCKADDR_IN6_SCOPE_ID
-		if (sin6t->sin6_scope_id &&
-		    sin6a->sin6_scope_id != sin6t->sin6_scope_id) {
+		if (sin6t->sin6_scope_id && sin6a->sin6_scope_id != sin6t->sin6_scope_id) {
 			ret = 0;
 			goto out;
 		}
 #endif
 
 		break;
-	    }
+	}
 #endif
 	default:
-	    rprintf(FLOG, "unknown family %u\n", rest->ai_family);
-	    ret = 0;
-	    goto out;
+		rprintf(FLOG, "unknown family %u\n", rest->ai_family);
+		ret = 0;
+		goto out;
 	}
 
 	bits = -1;
@@ -208,26 +229,21 @@ static int match_address(char *addr, char *tok)
 
 	ret = match_binary(a, t, mask, addrlen);
 
-out:
+  out:
 	freeaddrinfo(resa);
 	freeaddrinfo(rest);
 	return ret;
 }
 
-static int access_match(char *list, char *addr, char *host)
+static int access_match(const char *list, const char *addr, const char **host_ptr)
 {
 	char *tok;
 	char *list2 = strdup(list);
 
-	if (!list2)
-		out_of_memory("access_match");
-
 	strlower(list2);
-	if (host)
-		strlower(host);
 
 	for (tok = strtok(list2, " ,\t"); tok; tok = strtok(NULL, " ,\t")) {
-		if (match_hostname(host, tok) || match_address(addr, tok)) {
+		if (match_hostname(host_ptr, addr, tok) || match_address(addr, tok)) {
 			free(list2);
 			return 1;
 		}
@@ -237,16 +253,21 @@ static int access_match(char *list, char *addr, char *host)
 	return 0;
 }
 
-int allow_access(char *addr, char *host, char *allow_list, char *deny_list)
+int allow_access(const char *addr, const char **host_ptr, int i)
 {
+	const char *allow_list = lp_hosts_allow(i);
+	const char *deny_list = lp_hosts_deny(i);
+
 	if (allow_list && !*allow_list)
 		allow_list = NULL;
 	if (deny_list && !*deny_list)
 		deny_list = NULL;
 
+	allow_forward_dns = lp_forward_lookup(i);
+
 	/* If we match an allow-list item, we always allow access. */
 	if (allow_list) {
-		if (access_match(allow_list, addr, host))
+		if (access_match(allow_list, addr, host_ptr))
 			return 1;
 		/* For an allow-list w/o a deny-list, disallow non-matches. */
 		if (!deny_list)
@@ -255,7 +276,7 @@ int allow_access(char *addr, char *host, char *allow_list, char *deny_list)
 
 	/* If we match a deny-list item (and got past any allow-list
 	 * items), we always disallow access. */
-	if (deny_list && access_match(deny_list, addr, host))
+	if (deny_list && access_match(deny_list, addr, host_ptr))
 		return 0;
 
 	/* Allow all other access. */

authenticate.c

@@ -1,41 +1,41 @@
-/* -*- c-file-style: "linux"; -*-
+/*
+ * Support rsync daemon authentication.
+ *
+ * Copyright (C) 1998-2000 Andrew Tridgell
+ * Copyright (C) 2002-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
-   Copyright (C) 1998-2000 by Andrew Tridgell
-
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
-
-/* support rsync authentication */
 #include "rsync.h"
+#include "itypes.h"
+#include "ifuncs.h"
 
+extern int read_only;
 extern char *password_file;
-extern int am_root;
 
 /***************************************************************************
 encode a buffer using base64 - simple and slow algorithm. null terminates
 the result.
   ***************************************************************************/
-void base64_encode(char *buf, int len, char *out)
+void base64_encode(const char *buf, int len, char *out, int pad)
 {
 	char *b64 = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
 	int bit_offset, byte_offset, idx, i;
-	unsigned char *d = (unsigned char *)buf;
+	const uchar *d = (const uchar *)buf;
 	int bytes = (len*8 + 5)/6;
 
-	memset(out, 0, bytes+1);
-
 	for (i = 0; i < bytes; i++) {
 		byte_offset = (i*6)/8;
 		bit_offset = (i*6)%8;
@@ -49,163 +49,170 @@ void base64_encode(char *buf, int len, char *out)
 		}
 		out[i] = b64[idx];
 	}
+
+	while (pad && (i % 4))
+		out[i++] = '=';
+
+	out[i] = '\0';
 }
 
-/* create a 16 byte challenge buffer */
-static void gen_challenge(char *addr, char *challenge)
+/* Generate a challenge buffer and return it base64-encoded. */
+static void gen_challenge(const char *addr, char *challenge)
 {
 	char input[32];
+	char digest[MAX_DIGEST_LEN];
 	struct timeval tv;
+	int len;
 
 	memset(input, 0, sizeof input);
 
-	strlcpy((char *)input, addr, 17);
+	strlcpy(input, addr, 17);
 	sys_gettimeofday(&tv);
 	SIVAL(input, 16, tv.tv_sec);
 	SIVAL(input, 20, tv.tv_usec);
 	SIVAL(input, 24, getpid());
 
-	sum_init(0);
+	sum_init(-1, 0);
 	sum_update(input, sizeof input);
-	sum_end(challenge);
+	len = sum_end(digest);
+
+	base64_encode(digest, len, challenge, 0);
 }
 
+/* Generate an MD4 hash created from the combination of the password
+ * and the challenge string and return it base64-encoded. */
+static void generate_hash(const char *in, const char *challenge, char *out)
+{
+	char buf[MAX_DIGEST_LEN];
+	int len;
+
+	sum_init(-1, 0);
+	sum_update(in, strlen(in));
+	sum_update(challenge, strlen(challenge));
+	len = sum_end(buf);
+
+	base64_encode(buf, len, out, 0);
+}
 
 /* Return the secret for a user from the secret file, null terminated.
  * Maximum length is len (not counting the null). */
-static int get_secret(int module, char *user, char *secret, int len)
+static const char *check_secret(int module, const char *user, const char *group,
+				const char *challenge, const char *pass)
 {
-	char *fname = lp_secrets_file(module);
+	char line[1024];
+	char pass2[MAX_DIGEST_LEN*2];
+	const char *fname = lp_secrets_file(module);
 	STRUCT_STAT st;
-	int fd, ok = 1;
-	char ch, *p;
+	int ok = 1;
+	int user_len = strlen(user);
+	int group_len = group ? strlen(group) : 0;
+	char *err;
+	FILE *fh;
 
-	if (!fname || !*fname)
-		return 0;
+	if (!fname || !*fname || (fh = fopen(fname, "r")) == NULL)
+		return "no secrets file";
 
-	if ((fd = open(fname, O_RDONLY)) < 0)
-		return 0;
-
-	if (do_stat(fname, &st) == -1) {
-		rsyserr(FLOG, errno, "stat(%s)", safe_fname(fname));
+	if (do_fstat(fileno(fh), &st) == -1) {
+		rsyserr(FLOG, errno, "fstat(%s)", fname);
 		ok = 0;
 	} else if (lp_strict_modes(module)) {
 		if ((st.st_mode & 06) != 0) {
 			rprintf(FLOG, "secrets file must not be other-accessible (see strict modes option)\n");
 			ok = 0;
-		} else if (am_root && (st.st_uid != 0)) {
+		} else if (MY_UID() == 0 && st.st_uid != 0) {
 			rprintf(FLOG, "secrets file must be owned by root when running as root (see strict modes)\n");
 			ok = 0;
 		}
 	}
 	if (!ok) {
-		rprintf(FLOG, "continuing without secrets file\n");
-		close(fd);
-		return 0;
+		fclose(fh);
+		return "ignoring secrets file";
 	}
 
 	if (*user == '#') {
 		/* Reject attempt to match a comment. */
-		close(fd);
-		return 0;
+		fclose(fh);
+		return "invalid username";
 	}
 
-	/* Try to find a line that starts with the user name and a ':'. */
-	p = user;
-	while (1) {
-		if (read(fd, &ch, 1) != 1) {
-			close(fd);
-			return 0;
-		}
-		if (ch == '\n')
-			p = user;
-		else if (p) {
-			if (*p == ch)
-				p++;
-			else if (!*p && ch == ':')
-				break;
-			else
-				p = NULL;
-		}
-	}
-
-	/* Slurp the secret into the "secret" buffer. */
-	p = secret;
-	while (len > 0) {
-		if (read(fd, p, 1) != 1 || *p == '\n')
-			break;
-		if (*p == '\r')
+	/* Try to find a line that starts with the user (or @group) name and a ':'. */
+	err = "secret not found";
+	while ((user || group) && fgets(line, sizeof line, fh) != NULL) {
+		const char **ptr, *s = strtok(line, "\n\r");
+		int len;
+		if (!s)
 			continue;
-		p++;
-		len--;
+		if (*s == '@') {
+			ptr = &group;
+			len = group_len;
+			s++;
+		} else {
+			ptr = &user;
+			len = user_len;
+		}
+		if (!*ptr || strncmp(s, *ptr, len) != 0 || s[len] != ':')
+			continue;
+		generate_hash(s+len+1, challenge, pass2);
+		if (strcmp(pass, pass2) == 0) {
+			err = NULL;
+			break;
+		}
+		err = "password mismatch";
+		*ptr = NULL; /* Don't look for name again. */
 	}
-	*p = '\0';
-	close(fd);
 
-	return 1;
+	fclose(fh);
+
+	force_memzero(line, sizeof line);
+	force_memzero(pass2, sizeof pass2);
+
+	return err;
 }
 
-static char *getpassf(char *filename)
+static const char *getpassf(const char *filename)
 {
 	STRUCT_STAT st;
 	char buffer[512], *p;
-	int fd, n, ok = 1;
-	char *envpw = getenv("RSYNC_PASSWORD");
+	int n;
 
 	if (!filename)
 		return NULL;
 
-	if ((fd = open(filename,O_RDONLY)) < 0) {
-		rsyserr(FERROR, errno, "could not open password file \"%s\"",
-			safe_fname(filename));
-		if (envpw)
-			rprintf(FERROR, "falling back to RSYNC_PASSWORD environment variable.\n");
-		return NULL;
-	}
+	if (strcmp(filename, "-") == 0) {
+		n = fgets(buffer, sizeof buffer, stdin) == NULL ? -1 : (int)strlen(buffer);
+	} else {
+		int fd;
 
-	if (do_stat(filename, &st) == -1) {
-		rsyserr(FERROR, errno, "stat(%s)", safe_fname(filename));
-		ok = 0;
-	} else if ((st.st_mode & 06) != 0) {
-		rprintf(FERROR,"password file must not be other-accessible\n");
-		ok = 0;
-	} else if (am_root && st.st_uid != 0) {
-		rprintf(FERROR,"password file must be owned by root when running as root\n");
-		ok = 0;
-	}
-	if (!ok) {
-		rprintf(FERROR,"continuing without password file\n");
-		if (envpw)
-			rprintf(FERROR, "using RSYNC_PASSWORD environment variable.\n");
+		if ((fd = open(filename,O_RDONLY)) < 0) {
+			rsyserr(FERROR, errno, "could not open password file %s", filename);
+			exit_cleanup(RERR_SYNTAX);
+		}
+
+		if (do_stat(filename, &st) == -1) {
+			rsyserr(FERROR, errno, "stat(%s)", filename);
+			exit_cleanup(RERR_SYNTAX);
+		}
+		if ((st.st_mode & 06) != 0) {
+			rprintf(FERROR, "ERROR: password file must not be other-accessible\n");
+			exit_cleanup(RERR_SYNTAX);
+		}
+		if (MY_UID() == 0 && st.st_uid != 0) {
+			rprintf(FERROR, "ERROR: password file must be owned by root when running as root\n");
+			exit_cleanup(RERR_SYNTAX);
+		}
+
+		n = read(fd, buffer, sizeof buffer - 1);
 		close(fd);
-		return NULL;
 	}
 
-	if (envpw)
-		rprintf(FERROR, "RSYNC_PASSWORD environment variable ignored\n");
-
-	n = read(fd, buffer, sizeof buffer - 1);
-	close(fd);
 	if (n > 0) {
 		buffer[n] = '\0';
 		if ((p = strtok(buffer, "\n\r")) != NULL)
 			return strdup(p);
 	}
 
-	return NULL;
-}
-
-/* generate a 16 byte hash from a password and challenge */
-static void generate_hash(char *in, char *challenge, char *out)
-{
-	char buf[16];
-
-	sum_init(0);
-	sum_update(in, strlen(in));
-	sum_update(challenge, strlen(challenge));
-	sum_end(buf);
-
-	base64_encode(buf, 16, out);
+	rprintf(FERROR, "ERROR: failed to read a password from %s\n", filename);
+	exit_cleanup(RERR_SYNTAX);
 }
 
 /* Possibly negotiate authentication with the client.  Use "leader" to
@@ -214,17 +221,18 @@ static void generate_hash(char *in, char *challenge, char *out)
  * Return NULL if authentication failed.  Return "" if anonymous access.
  * Otherwise return username.
  */
-char *auth_server(int f_in, int f_out, int module, char *addr, char *leader)
+char *auth_server(int f_in, int f_out, int module, const char *host,
+		  const char *addr, const char *leader)
 {
 	char *users = lp_auth_users(module);
-	char challenge[16];
-	char b64_challenge[30];
-	char line[MAXPATHLEN];
-	static char user[100];
-	char secret[100];
-	char pass[30];
-	char pass2[30];
-	char *tok;
+	char challenge[MAX_DIGEST_LEN*2];
+	char line[BIGPATHBUFLEN];
+	char **auth_uid_groups = NULL;
+	int auth_uid_groups_cnt = -1;
+	const char *err = NULL;
+	int group_match = -1;
+	char *tok, *pass;
+	char opt_ch = '\0';
 
 	/* if no auth list then allow anyone in! */
 	if (!users || !*users)
@@ -232,52 +240,113 @@ char *auth_server(int f_in, int f_out, int module, char *addr, char *leader)
 
 	gen_challenge(addr, challenge);
 
-	base64_encode(challenge, 16, b64_challenge);
+	io_printf(f_out, "%s%s\n", leader, challenge);
 
-	io_printf(f_out, "%s%s\n", leader, b64_challenge);
-
-	if (!read_line(f_in, line, sizeof line - 1))
-		return NULL;
-
-	memset(user, 0, sizeof user);
-	memset(pass, 0, sizeof pass);
-
-	if (sscanf(line,"%99s %29s", user, pass) != 2)
+	if (!read_line_old(f_in, line, sizeof line, 0)
+	 || (pass = strchr(line, ' ')) == NULL) {
+		rprintf(FLOG, "auth failed on module %s from %s (%s): "
+			"invalid challenge response\n",
+			lp_name(module), host, addr);
 		return NULL;
+	}
+	*pass++ = '\0';
 
 	users = strdup(users);
-	if (!users)
-		return NULL;
 
-	for (tok=strtok(users," ,\t"); tok; tok = strtok(NULL," ,\t")) {
-		if (wildmatch(tok, user))
-			break;
+	for (tok = strtok(users, " ,\t"); tok; tok = strtok(NULL, " ,\t")) {
+		char *opts;
+		/* See if the user appended :deny, :ro, or :rw. */
+		if ((opts = strchr(tok, ':')) != NULL) {
+			*opts++ = '\0';
+			opt_ch = isUpper(opts) ? toLower(opts) : *opts;
+			if (opt_ch == 'r') { /* handle ro and rw */
+				opt_ch = isUpper(opts+1) ? toLower(opts+1) : opts[1];
+				if (opt_ch == 'o')
+					opt_ch = 'r';
+				else if (opt_ch != 'w')
+					opt_ch = '\0';
+			} else if (opt_ch != 'd') /* if it's not deny, ignore it */
+				opt_ch = '\0';
+		} else
+			opt_ch = '\0';
+		if (*tok != '@') {
+			/* Match the username */
+			if (wildmatch(tok, line))
+				break;
+		} else {
+#ifdef HAVE_GETGROUPLIST
+			int j;
+			/* See if authorizing user is a real user, and if so, see
+			 * if it is in a group that matches tok+1 wildmat. */
+			if (auth_uid_groups_cnt < 0) {
+				item_list gid_list = EMPTY_ITEM_LIST;
+				uid_t auth_uid;
+				if (!user_to_uid(line, &auth_uid, False)
+				 || getallgroups(auth_uid, &gid_list) != NULL)
+					auth_uid_groups_cnt = 0;
+				else {
+					gid_t *gid_array = gid_list.items;
+					auth_uid_groups_cnt = gid_list.count;
+					auth_uid_groups = new_array(char *, auth_uid_groups_cnt);
+					for (j = 0; j < auth_uid_groups_cnt; j++)
+						auth_uid_groups[j] = gid_to_group(gid_array[j]);
+				}
+			}
+			for (j = 0; j < auth_uid_groups_cnt; j++) {
+				if (auth_uid_groups[j] && wildmatch(tok+1, auth_uid_groups[j])) {
+					group_match = j;
+					break;
+				}
+			}
+			if (group_match >= 0)
+				break;
+#else
+			rprintf(FLOG, "your computer doesn't support getgrouplist(), so no @group authorization is possible.\n");
+#endif
+		}
 	}
+
 	free(users);
 
 	if (!tok)
-		return NULL;
+		err = "no matching rule";
+	else if (opt_ch == 'd')
+		err = "denied by rule";
+	else {
+		char *group = group_match >= 0 ? auth_uid_groups[group_match] : NULL;
+		err = check_secret(module, line, group, challenge, pass);
+	}
 
-	memset(secret, 0, sizeof secret);
-	if (!get_secret(module, user, secret, sizeof secret - 1)) {
-		memset(secret, 0, sizeof secret);
+	force_memzero(challenge, sizeof challenge);
+	force_memzero(pass, strlen(pass));
+
+	if (auth_uid_groups) {
+		int j;
+		for (j = 0; j < auth_uid_groups_cnt; j++) {
+			if (auth_uid_groups[j])
+				free(auth_uid_groups[j]);
+		}
+		free(auth_uid_groups);
+	}
+
+	if (err) {
+		rprintf(FLOG, "auth failed on module %s from %s (%s) for %s: %s\n",
+			lp_name(module), host, addr, line, err);
 		return NULL;
 	}
 
-	generate_hash(secret, b64_challenge, pass2);
-	memset(secret, 0, sizeof secret);
+	if (opt_ch == 'r')
+		read_only = 1;
+	else if (opt_ch == 'w')
+		read_only = 0;
 
-	if (strcmp(pass, pass2) == 0)
-		return user;
-
-	return NULL;
+	return strdup(line);
 }
 
-
-void auth_client(int fd, char *user, char *challenge)
+void auth_client(int fd, const char *user, const char *challenge)
 {
-	char *pass;
-	char pass2[30];
+	const char *pass;
+	char pass2[MAX_DIGEST_LEN*2];
 
 	if (!user || !*user)
 		user = "nobody";
@@ -287,12 +356,12 @@ void auth_client(int fd, char *user, char *challenge)
 		/* XXX: cyeoh says that getpass is deprecated, because
 		 * it may return a truncated password on some systems,
 		 * and it is not in the LSB.
-                 *
-                 * Andrew Klein says that getpassphrase() is present
-                 * on Solaris and reads up to 256 characters.
-                 *
-                 * OpenBSD has a readpassphrase() that might be more suitable.
-                 */
+		 *
+		 * Andrew Klein says that getpassphrase() is present
+		 * on Solaris and reads up to 256 characters.
+		 *
+		 * OpenBSD has a readpassphrase() that might be more suitable.
+		 */
 		pass = getpass("Password: ");
 	}
 
@@ -302,5 +371,3 @@ void auth_client(int fd, char *user, char *challenge)
 	generate_hash(pass, challenge, pass2);
 	io_printf(fd, "%s %s\n", user, pass2);
 }
-
-

backup.c

@@ -1,279 +1,354 @@
 /*
-   Copyright (C) Andrew Tridgell 1999
-
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
-
-/* backup handling code */
+ * Backup handling code.
+ *
+ * Copyright (C) 1999 Andrew Tridgell
+ * Copyright (C) 2003-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
 #include "rsync.h"
+#include "ifuncs.h"
 
-extern int verbose;
-extern int backup_suffix_len;
+extern int am_root;
+extern int preserve_acls;
+extern int preserve_xattrs;
+extern int preserve_devices;
+extern int preserve_specials;
+extern int preserve_links;
+extern int safe_symlinks;
 extern int backup_dir_len;
 extern unsigned int backup_dir_remainder;
 extern char backup_dir_buf[MAXPATHLEN];
 extern char *backup_suffix;
 extern char *backup_dir;
 
-extern int am_root;
-extern int preserve_devices;
-extern int preserve_links;
-extern int preserve_hard_links;
-extern int orig_umask;
-extern int safe_symlinks;
+/* Returns -1 on error, 0 on missing dir, and 1 on present dir. */
+static int validate_backup_dir(void)
+{
+	STRUCT_STAT st;
 
-/* make a complete pathname for backup file */
-char *get_backup_name(char *fname)
+	if (do_lstat(backup_dir_buf, &st) < 0) {
+		if (errno == ENOENT)
+			return 0;
+		rsyserr(FERROR, errno, "backup lstat %s failed", backup_dir_buf);
+		return -1;
+	}
+	if (!S_ISDIR(st.st_mode)) {
+		int flags = get_del_for_flag(st.st_mode) | DEL_FOR_BACKUP | DEL_RECURSE;
+		if (delete_item(backup_dir_buf, st.st_mode, flags) == 0)
+			return 0;
+		return -1;
+	}
+	return 1;
+}
+
+/* Create a backup path from the given fname, putting the result into
+ * backup_dir_buf.  Any new directories (compared to the prior backup
+ * path) are ensured to exist as directories, replacing anything else
+ * that may be in the way (e.g. a symlink). */
+static BOOL copy_valid_path(const char *fname)
+{
+	const char *f;
+	int val;
+	BOOL ret = True;
+	stat_x sx;
+	char *b, *rel = backup_dir_buf + backup_dir_len, *name = rel;
+
+	for (f = fname, b = rel; *f && *f == *b; f++, b++) {
+		if (*b == '/')
+			name = b + 1;
+	}
+
+	if (stringjoin(rel, backup_dir_remainder, fname, backup_suffix, NULL) >= backup_dir_remainder) {
+		rprintf(FERROR, "backup filename too long\n");
+		*name = '\0';
+		return False;
+	}
+
+	for ( ; ; name = b + 1) {
+		if ((b = strchr(name, '/')) == NULL)
+			return True;
+		*b = '\0';
+
+		val = validate_backup_dir();
+		if (val == 0)
+			break;
+		if (val < 0) {
+			*name = '\0';
+			return False;
+		}
+
+		*b = '/';
+	}
+
+	init_stat_x(&sx);
+
+	for ( ; b; name = b + 1, b = strchr(name, '/')) {
+		*b = '\0';
+
+		while (do_mkdir(backup_dir_buf, ACCESSPERMS) < 0) {
+			if (errno == EEXIST) {
+				val = validate_backup_dir();
+				if (val > 0)
+					break;
+				if (val == 0)
+					continue;
+			} else
+				rsyserr(FERROR, errno, "backup mkdir %s failed", backup_dir_buf);
+			*name = '\0';
+			ret = False;
+			goto cleanup;
+		}
+
+		/* Try to transfer the directory settings of the actual dir
+		 * that the files are coming from. */
+		if (x_stat(rel, &sx.st, NULL) < 0)
+			rsyserr(FERROR, errno, "backup stat %s failed", full_fname(rel));
+		else {
+			struct file_struct *file;
+			if (!(file = make_file(rel, NULL, NULL, 0, NO_FILTERS)))
+				continue;
+#ifdef SUPPORT_ACLS
+			if (preserve_acls && !S_ISLNK(file->mode)) {
+				get_acl(rel, &sx);
+				cache_tmp_acl(file, &sx);
+				free_acl(&sx);
+			}
+#endif
+#ifdef SUPPORT_XATTRS
+			if (preserve_xattrs) {
+				get_xattr(rel, &sx);
+				cache_tmp_xattr(file, &sx);
+				free_xattr(&sx);
+			}
+#endif
+			set_file_attrs(backup_dir_buf, file, NULL, NULL, 0);
+			unmake_file(file);
+		}
+
+		*b = '/';
+	}
+
+  cleanup:
+
+#ifdef SUPPORT_ACLS
+	uncache_tmp_acls();
+#endif
+#ifdef SUPPORT_XATTRS
+	uncache_tmp_xattrs();
+#endif
+
+	return ret;
+}
+
+/* Make a complete pathname for backup file and verify any new path elements. */
+char *get_backup_name(const char *fname)
 {
 	if (backup_dir) {
-		if (stringjoin(backup_dir_buf + backup_dir_len, backup_dir_remainder,
-			       fname, backup_suffix, NULL) < backup_dir_remainder)
-			return backup_dir_buf;
-	} else {
-		if (stringjoin(backup_dir_buf, MAXPATHLEN,
-			       fname, backup_suffix, NULL) < MAXPATHLEN)
+		static int initialized = 0;
+		if (!initialized) {
+			int ret;
+			if (backup_dir_len > 1)
+				backup_dir_buf[backup_dir_len-1] = '\0';
+			ret = make_path(backup_dir_buf, 0);
+			if (backup_dir_len > 1)
+				backup_dir_buf[backup_dir_len-1] = '/';
+			if (ret < 0)
+				return NULL;
+			initialized = 1;
+		}
+		/* copy fname into backup_dir_buf while validating the dirs. */
+		if (copy_valid_path(fname))
 			return backup_dir_buf;
+		/* copy_valid_path() has printed an error message. */
+		return NULL;
 	}
 
+	if (stringjoin(backup_dir_buf, MAXPATHLEN, fname, backup_suffix, NULL) < MAXPATHLEN)
+		return backup_dir_buf;
+
 	rprintf(FERROR, "backup filename too long\n");
 	return NULL;
 }
 
-/* simple backup creates a backup with a suffix in the same directory */
-static int make_simple_backup(char *fname)
+/* Has same return codes as make_backup(). */
+static inline int link_or_rename(const char *from, const char *to,
+				 BOOL prefer_rename, STRUCT_STAT *stp)
 {
-	int rename_errno;
-	char *fnamebak = get_backup_name(fname);
-
-	if (!fnamebak)
-		return 0;
-
-	while (1) {
-		if (do_rename(fname, fnamebak) == 0) {
-			if (verbose > 1) {
-				rprintf(FINFO, "backed up %s to %s\n",
-					safe_fname(fname),
-					safe_fname(fnamebak));
-			}
-			break;
+#ifdef SUPPORT_HARD_LINKS
+	if (!prefer_rename) {
+#ifndef CAN_HARDLINK_SYMLINK
+		if (S_ISLNK(stp->st_mode))
+			return 0; /* Use copy code. */
+#endif
+#ifndef CAN_HARDLINK_SPECIAL
+		if (IS_SPECIAL(stp->st_mode) || IS_DEVICE(stp->st_mode))
+			return 0; /* Use copy code. */
+#endif
+		if (do_link(from, to) == 0) {
+			if (DEBUG_GTE(BACKUP, 1))
+				rprintf(FINFO, "make_backup: HLINK %s successful.\n", from);
+			return 2;
 		}
-		/* cygwin (at least version b19) reports EINVAL */
-		if (errno == ENOENT || errno == EINVAL)
-			break;
-
-		rename_errno = errno;
-		if (errno == EISDIR && do_rmdir(fnamebak) == 0)
-			continue;
-		if (errno == ENOTDIR && do_unlink(fnamebak) == 0)
-			continue;
-
-		rsyserr(FERROR, rename_errno, "rename %s to backup %s",
-			safe_fname(fname), safe_fname(fnamebak));
-		errno = rename_errno;
-		return 0;
+		/* We prefer to rename a regular file rather than copy it. */
+		if (!S_ISREG(stp->st_mode) || errno == EEXIST || errno == EISDIR)
+			return 0;
 	}
-
-	return 1;
-}
-
-
-/****************************************************************************
-Create a directory given an absolute path, perms based upon another directory
-path
-****************************************************************************/
-static int make_bak_dir(char *fullpath)
-{
-	STRUCT_STAT st;
-	char *rel = fullpath + backup_dir_len;
-	char *end = rel + strlen(rel);
-	char *p = end;
-
-	while (strncmp(fullpath, "./", 2) == 0)
-		fullpath += 2;
-
-	/* Try to find an existing dir, starting from the deepest dir. */
-	while (1) {
-		if (--p == fullpath) {
-			p += strlen(p);
-			goto failure;
+#endif
+	if (do_rename(from, to) == 0) {
+		if (stp->st_nlink > 1 && !S_ISDIR(stp->st_mode)) {
+			/* If someone has hard-linked the file into the backup
+			 * dir, rename() might return success but do nothing! */
+			robust_unlink(from); /* Just in case... */
 		}
-		if (*p == '/') {
-			*p = '\0';
-			if (do_mkdir(fullpath, 0777 & ~orig_umask) == 0)
-				break;
-			if (errno != ENOENT) {
-				rsyserr(FERROR, errno,
-					"make_bak_dir mkdir %s failed",
-					full_fname(fullpath));
-				goto failure;
-			}
-		}
-	}
-
-	/* Make all the dirs that we didn't find on the way here. */
-	while (1) {
-		if (p >= rel) {
-			/* Try to transfer the directory settings of the
-			 * actual dir that the files are coming from. */
-			if (do_stat(rel, &st) < 0) {
-				rsyserr(FERROR, errno,
-					"make_bak_dir stat %s failed",
-					full_fname(rel));
-			} else {
-				do_lchown(fullpath, st.st_uid, st.st_gid);
-				do_chmod(fullpath, st.st_mode);
-			}
-		}
-		*p = '/';
-		p += strlen(p);
-		if (p == end)
-			break;
-		if (do_mkdir(fullpath, 0777 & ~orig_umask) < 0) {
-			rsyserr(FERROR, errno, "make_bak_dir mkdir %s failed",
-				full_fname(fullpath));
-			goto failure;
-		}
-	}
-	return 0;
-
-failure:
-	while (p != end) {
-		*p = '/';
-		p += strlen(p);
-	}
-	return -1;
-}
-
-/* robustly move a file, creating new directory structures if necessary */
-static int robust_move(char *src, char *dst)
-{
-	if (robust_rename(src, dst, 0755) < 0 && (errno != ENOENT
-	    || make_bak_dir(dst) < 0 || robust_rename(src, dst, 0755) < 0))
-		return -1;
-	return 0;
-}
-
-
-/* If we have a --backup-dir, then we get here from make_backup().
- * We will move the file to be deleted into a parallel directory tree. */
-static int keep_backup(char *fname)
-{
-	STRUCT_STAT st;
-	struct file_struct *file;
-	char *buf;
-	int kept = 0;
-	int ret_code;
-
-	/* return if no file to keep */
-	if (do_lstat(fname, &st) < 0)
+		if (DEBUG_GTE(BACKUP, 1))
+			rprintf(FINFO, "make_backup: RENAME %s successful.\n", from);
 		return 1;
+	}
+	return 0;
+}
 
-	if (!(file = make_file(fname, NULL, NO_FILTERS)))
-		return 1; /* the file could have disappeared */
+/* Hard-link, rename, or copy an item to the backup name.  Returns 0 for
+ * failure, 1 if item was moved, 2 if item was duplicated or hard linked
+ * into backup area, or 3 if item doesn't exist or isn't a regular file. */
+int make_backup(const char *fname, BOOL prefer_rename)
+{
+	stat_x sx;
+	struct file_struct *file;
+	int save_preserve_xattrs;
+	char *buf;
+	int ret = 0;
+
+	init_stat_x(&sx);
+	/* Return success if no file to keep. */
+	if (x_lstat(fname, &sx.st, NULL) < 0)
+		return 3;
 
 	if (!(buf = get_backup_name(fname)))
 		return 0;
 
-	/* Check to see if this is a device file, or link */
-	if (IS_DEVICE(file->mode)) {
-		if (am_root && preserve_devices) {
-			if (do_mknod(buf, file->mode, file->u.rdev) < 0
-			    && (errno != ENOENT || make_bak_dir(buf) < 0
-			     || do_mknod(buf, file->mode, file->u.rdev) < 0)) {
-				rsyserr(FERROR, errno, "mknod %s failed",
-					full_fname(buf));
-			} else if (verbose > 2) {
-				rprintf(FINFO,
-					"make_backup: DEVICE %s successful.\n",
-					safe_fname(fname));
-			}
+	/* Try a hard-link or a rename first.  Using rename is not atomic, but
+	 * is more efficient than forcing a copy for larger files when no hard-
+	 * linking is possible. */
+	if ((ret = link_or_rename(fname, buf, prefer_rename, &sx.st)) != 0)
+		goto success;
+	if (errno == EEXIST || errno == EISDIR) {
+		STRUCT_STAT bakst;
+		if (do_lstat(buf, &bakst) == 0) {
+			int flags = get_del_for_flag(bakst.st_mode) | DEL_FOR_BACKUP | DEL_RECURSE;
+			if (delete_item(buf, bakst.st_mode, flags) != 0)
+				return 0;
 		}
-		kept = 1;
-		do_unlink(fname);
+		if ((ret = link_or_rename(fname, buf, prefer_rename, &sx.st)) != 0)
+			goto success;
 	}
 
-	if (!kept && S_ISDIR(file->mode)) {
-		/* make an empty directory */
-		if (do_mkdir(buf, file->mode) < 0
-		    && (errno != ENOENT || make_bak_dir(buf) < 0
-		     || do_mkdir(buf, file->mode) < 0)) {
-			rsyserr(FINFO, errno, "mkdir %s failed",
-				full_fname(buf));
-		}
+	/* Fall back to making a copy. */
+	if (!(file = make_file(fname, NULL, &sx.st, 0, NO_FILTERS)))
+		return 3; /* the file could have disappeared */
 
-		ret_code = do_rmdir(fname);
-		if (verbose > 2) {
-			rprintf(FINFO, "make_backup: RMDIR %s returns %i\n",
-				full_fname(fname), ret_code);
-		}
-		kept = 1;
+#ifdef SUPPORT_ACLS
+	if (preserve_acls && !S_ISLNK(file->mode)) {
+		get_acl(fname, &sx);
+		cache_tmp_acl(file, &sx);
+		free_acl(&sx);
 	}
-
-#ifdef SUPPORT_LINKS
-	if (!kept && preserve_links && S_ISLNK(file->mode)) {
-		if (safe_symlinks && unsafe_symlink(file->u.link, buf)) {
-			if (verbose) {
-				rprintf(FINFO, "ignoring unsafe symlink %s -> %s\n",
-					full_fname(buf), file->u.link);
-			}
-			kept = 1;
-		}
-		if (do_symlink(file->u.link, buf) < 0
-		    && (errno != ENOENT || make_bak_dir(buf) < 0
-		     || do_symlink(file->u.link, buf) < 0)) {
-			rsyserr(FERROR, errno, "link %s -> \"%s\"",
-				full_fname(buf), safe_fname(file->u.link));
-		}
-		do_unlink(fname);
-		kept = 1;
+#endif
+#ifdef SUPPORT_XATTRS
+	if (preserve_xattrs) {
+		get_xattr(fname, &sx);
+		cache_tmp_xattr(file, &sx);
+		free_xattr(&sx);
 	}
 #endif
 
-	if (!kept && !S_ISREG(file->mode)) {
-		rprintf(FINFO, "make_bak: skipping non-regular file %s\n",
-			safe_fname(fname));
-		return 1;
+	/* Check to see if this is a device file, or link */
+	if ((am_root && preserve_devices && IS_DEVICE(file->mode))
+	 || (preserve_specials && IS_SPECIAL(file->mode))) {
+		if (do_mknod(buf, file->mode, sx.st.st_rdev) < 0)
+			rsyserr(FERROR, errno, "mknod %s failed", full_fname(buf));
+		else if (DEBUG_GTE(BACKUP, 1))
+			rprintf(FINFO, "make_backup: DEVICE %s successful.\n", fname);
+		ret = 2;
 	}
 
-	/* move to keep tree if a file */
-	if (!kept) {
-		if (robust_move(fname, buf) != 0) {
-			rsyserr(FERROR, errno, "keep_backup failed: %s -> \"%s\"",
-				full_fname(fname), safe_fname(buf));
-		} else if (st.st_nlink > 1) {
-			/* If someone has hard-linked the file into the backup
-			 * dir, rename() might return success but do nothing! */
-			robust_unlink(fname); /* Just in case... */
+#ifdef SUPPORT_LINKS
+	if (!ret && preserve_links && S_ISLNK(file->mode)) {
+		const char *sl = F_SYMLINK(file);
+		if (safe_symlinks && unsafe_symlink(sl, fname)) {
+			if (INFO_GTE(SYMSAFE, 1)) {
+				rprintf(FINFO, "not backing up unsafe symlink \"%s\" -> \"%s\"\n",
+					fname, sl);
+			}
+			ret = 2;
+		} else {
+			if (do_symlink(sl, buf) < 0)
+				rsyserr(FERROR, errno, "link %s -> \"%s\"", full_fname(buf), sl);
+			else if (DEBUG_GTE(BACKUP, 1))
+				rprintf(FINFO, "make_backup: SYMLINK %s successful.\n", fname);
+			ret = 2;
 		}
 	}
-	set_perms(buf, file, NULL, 0);
-	free(file);
+#endif
 
-	if (verbose > 1) {
-		rprintf(FINFO, "backed up %s to %s\n",
-			safe_fname(fname), safe_fname(buf));
+	if (!ret && !S_ISREG(file->mode)) {
+		rprintf(FINFO, "make_bak: skipping non-regular file %s\n", fname);
+		unmake_file(file);
+#ifdef SUPPORT_ACLS
+		uncache_tmp_acls();
+#endif
+#ifdef SUPPORT_XATTRS
+		uncache_tmp_xattrs();
+#endif
+		return 3;
 	}
-	return 1;
-}
 
+	/* Copy to backup tree if a file. */
+	if (!ret) {
+		if (copy_file(fname, buf, -1, file->mode) < 0) {
+			rsyserr(FERROR, errno, "keep_backup failed: %s -> \"%s\"",
+				full_fname(fname), buf);
+			unmake_file(file);
+#ifdef SUPPORT_ACLS
+			uncache_tmp_acls();
+#endif
+#ifdef SUPPORT_XATTRS
+			uncache_tmp_xattrs();
+#endif
+			return 0;
+		}
+		if (DEBUG_GTE(BACKUP, 1))
+			rprintf(FINFO, "make_backup: COPY %s successful.\n", fname);
+		ret = 2;
+	}
 
-/* main backup switch routine */
-int make_backup(char *fname)
-{
-	if (backup_dir)
-		return keep_backup(fname);
-	return make_simple_backup(fname);
+	save_preserve_xattrs = preserve_xattrs;
+	preserve_xattrs = 0;
+	set_file_attrs(buf, file, NULL, fname, ATTRS_ACCURATE_TIME);
+	preserve_xattrs = save_preserve_xattrs;
+
+	unmake_file(file);
+#ifdef SUPPORT_ACLS
+	uncache_tmp_acls();
+#endif
+#ifdef SUPPORT_XATTRS
+	uncache_tmp_xattrs();
+#endif
+
+  success:
+	if (INFO_GTE(BACKUP, 1))
+		rprintf(FINFO, "backed up %s to %s\n", fname, buf);
+	return ret;
 }

batch.c

@@ -1,14 +1,28 @@
-/* -*- c-file-style: "linux" -*-
-
-   Weiss 1/1999
-   Batch utilities for rsync.
-
-*/
+/*
+ * Support for the batch-file options.
+ *
+ * Copyright (C) 1999 Weiss
+ * Copyright (C) 2004 Chris Shoemaker
+ * Copyright (C) 2004-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
 #include "rsync.h"
+#include <zlib.h>
 #include <time.h>
 
-extern int am_sender;
 extern int eol_nulls;
 extern int recurse;
 extern int xfer_dirs;
@@ -17,12 +31,30 @@ extern int preserve_hard_links;
 extern int preserve_devices;
 extern int preserve_uid;
 extern int preserve_gid;
+extern int preserve_acls;
+extern int preserve_xattrs;
 extern int always_checksum;
 extern int do_compression;
+extern int inplace;
+extern int append_mode;
+extern int write_batch;
 extern int protocol_version;
+extern int raw_argc, cooked_argc;
+extern char **raw_argv, **cooked_argv;
 extern char *batch_name;
+#ifdef ICONV_OPTION
+extern char *iconv_opt;
+#endif
 
-extern struct filter_list_struct filter_list;
+extern filter_rule_list filter_list;
+
+int batch_fd = -1;
+int batch_sh_fd = -1;
+int batch_stream_flags;
+
+static int tweaked_append;
+static int tweaked_append_verify;
+static int tweaked_iconv;
 
 static int *flag_ptr[] = {
 	&recurse,		/* 0 */
@@ -34,6 +66,12 @@ static int *flag_ptr[] = {
 	&always_checksum,	/* 6 */
 	&xfer_dirs,		/* 7 (protocol 29) */
 	&do_compression,	/* 8 (protocol 29) */
+	&tweaked_iconv,		/* 9  (protocol 30) */
+	&preserve_acls,		/* 10 (protocol 30) */
+	&preserve_xattrs,	/* 11 (protocol 30) */
+	&inplace,		/* 12 (protocol 30) */
+	&tweaked_append,	/* 13 (protocol 30) */
+	&tweaked_append_verify,	/* 14 (protocol 30) */
 	NULL
 };
 
@@ -47,6 +85,12 @@ static char *flag_name[] = {
 	"--checksum (-c)",
 	"--dirs (-d)",
 	"--compress (-z)",
+	"--iconv",
+	"--acls (-A)",
+	"--xattrs (-X)",
+	"--inplace",
+	"--append",
+	"--append-verify",
 	NULL
 };
 
@@ -54,10 +98,14 @@ void write_stream_flags(int fd)
 {
 	int i, flags;
 
+	tweaked_append = append_mode == 1;
+	tweaked_append_verify = append_mode == 2;
+#ifdef ICONV_OPTION
+	tweaked_iconv = iconv_opt != NULL;
+#endif
+
 	/* Start the batch file with a bitmap of data-stream-affecting
 	 * flags. */
-	if (protocol_version < 29)
-		flag_ptr[7] = NULL;
 	for (i = 0, flags = 0; flag_ptr[i]; i++) {
 		if (*flag_ptr[i])
 			flags |= 1 << i;
@@ -67,14 +115,32 @@ void write_stream_flags(int fd)
 
 void read_stream_flags(int fd)
 {
-	int i, flags;
+	batch_stream_flags = read_int(fd);
+}
+
+void check_batch_flags(void)
+{
+	int i;
 
 	if (protocol_version < 29)
 		flag_ptr[7] = NULL;
-	for (i = 0, flags = read_int(fd); flag_ptr[i]; i++) {
-		int set = flags & (1 << i) ? 1 : 0;
+	else if (protocol_version < 30)
+		flag_ptr[9] = NULL;
+	tweaked_append = append_mode == 1;
+	tweaked_append_verify = append_mode == 2;
+#ifdef ICONV_OPTION
+	tweaked_iconv = iconv_opt != NULL;
+#endif
+	for (i = 0; flag_ptr[i]; i++) {
+		int set = batch_stream_flags & (1 << i) ? 1 : 0;
 		if (*flag_ptr[i] != set) {
-			if (verbose) {
+			if (i == 9) {
+				rprintf(FERROR,
+					"%s specify the --iconv option to use this batch file.\n",
+					set ? "Please" : "Do not");
+				exit_cleanup(RERR_SYNTAX);
+			}
+			if (INFO_GTE(MISC, 1)) {
 				rprintf(FINFO,
 					"%sing the %s option to match the batchfile.\n",
 					set ? "Sett" : "Clear", flag_name[i]);
@@ -88,42 +154,65 @@ void read_stream_flags(int fd)
 		else if (xfer_dirs < 2)
 			xfer_dirs = 0;
 	}
+
+	if (tweaked_append)
+		append_mode = 1;
+	else if (tweaked_append_verify)
+		append_mode = 2;
 }
 
-static void write_arg(int fd, char *arg)
+static int write_arg(const char *arg)
 {
-	char *x, *s;
+	const char *x, *s;
+	int len, err = 0;
 
 	if (*arg == '-' && (x = strchr(arg, '=')) != NULL) {
-		write(fd, arg, x - arg + 1);
+		err |= write(batch_sh_fd, arg, x - arg + 1) != x - arg + 1;
 		arg += x - arg + 1;
 	}
 
 	if (strpbrk(arg, " \"'&;|[]()$#!*?^\\") != NULL) {
-		write(fd, "'", 1);
+		err |= write(batch_sh_fd, "'", 1) != 1;
 		for (s = arg; (x = strchr(s, '\'')) != NULL; s = x + 1) {
-			write(fd, s, x - s + 1);
-			write(fd, "'", 1);
+			err |= write(batch_sh_fd, s, x - s + 1) != x - s + 1;
+			err |= write(batch_sh_fd, "'", 1) != 1;
 		}
-		write(fd, s, strlen(s));
-		write(fd, "'", 1);
-		return;
+		len = strlen(s);
+		err |= write(batch_sh_fd, s, len) != len;
+		err |= write(batch_sh_fd, "'", 1) != 1;
+		return err;
 	}
 
-	write(fd, arg, strlen(arg));
+	len = strlen(arg);
+	err |= write(batch_sh_fd, arg, len) != len;
+
+	return err;
+}
+
+/* Writes out a space and then an option (or other string) with an optional "=" + arg suffix. */
+static int write_opt(const char *opt, const char *arg)
+{
+	int len = strlen(opt);
+	int err = write(batch_sh_fd, " ", 1) != 1;
+	err = write(batch_sh_fd, opt, len) != len ? 1 : 0; 
+	if (arg) {
+		err |= write(batch_sh_fd, "=", 1) != 1;
+		err |= write_arg(arg);
+	}
+	return err;
 }
 
 static void write_filter_rules(int fd)
 {
-	struct filter_struct *ent;
+	filter_rule *ent;
 
 	write_sbuf(fd, " <<'#E#'\n");
 	for (ent = filter_list.head; ent; ent = ent->next) {
 		unsigned int plen;
-		char *p = get_rule_prefix(ent->match_flags, "- ", 0, &plen);
+		char *p = get_rule_prefix(ent, "- ", 0, &plen);
 		write_buf(fd, p, plen);
 		write_sbuf(fd, ent->pattern);
-		if (ent->match_flags & MATCHFLG_DIRECTORY)
+		if (ent->rflags & FILTRULE_DIRECTORY)
 			write_byte(fd, '/');
 		write_byte(fd, eol_nulls ? 0 : '\n');
 	}
@@ -132,40 +221,66 @@ static void write_filter_rules(int fd)
 	write_sbuf(fd, "#E#");
 }
 
+/* This sets batch_fd and (for --write-batch) batch_sh_fd. */
+void open_batch_files(void)
+{
+	if (write_batch) {
+		char filename[MAXPATHLEN];
+
+		stringjoin(filename, sizeof filename, batch_name, ".sh", NULL);
+
+		batch_sh_fd = do_open(filename, O_WRONLY | O_CREAT | O_TRUNC, S_IRUSR | S_IWUSR | S_IXUSR);
+		if (batch_sh_fd < 0) {
+			rsyserr(FERROR, errno, "Batch file %s open error", full_fname(filename));
+			exit_cleanup(RERR_FILESELECT);
+		}
+
+		batch_fd = do_open(batch_name, O_WRONLY | O_CREAT | O_TRUNC, S_IRUSR | S_IWUSR);
+	} else if (strcmp(batch_name, "-") == 0)
+		batch_fd = STDIN_FILENO;
+	else
+		batch_fd = do_open(batch_name, O_RDONLY, S_IRUSR | S_IWUSR);
+
+	if (batch_fd < 0) {
+		rsyserr(FERROR, errno, "Batch file %s open error", full_fname(batch_name));
+		exit_cleanup(RERR_FILEIO);
+	}
+}
+
 /* This routine tries to write out an equivalent --read-batch command
  * given the user's --write-batch args.  However, it doesn't really
  * understand most of the options, so it uses some overly simple
  * heuristics to munge the command line into something that will
  * (hopefully) work. */
-void write_batch_shell_file(int argc, char *argv[], int file_arg_cnt)
+void write_batch_shell_file(void)
 {
-	int fd, i;
-	char *p, filename[MAXPATHLEN];
-
-	stringjoin(filename, sizeof filename,
-		   batch_name, ".sh", NULL);
-	fd = do_open(filename, O_WRONLY | O_CREAT | O_TRUNC,
-		     S_IRUSR | S_IWUSR | S_IEXEC);
-	if (fd < 0) {
-		rsyserr(FERROR, errno, "Batch file %s open error",
-			safe_fname(filename));
-		exit_cleanup(1);
-	}
+	int i, j, len, err = 0;
+	char *p, *p2;
 
 	/* Write argvs info to BATCH.sh file */
-	write_arg(fd, argv[0]);
+	err |= write_arg(raw_argv[0]);
 	if (filter_list.head) {
 		if (protocol_version >= 29)
-			write_sbuf(fd, " --filter=._-");
+			err |= write_opt("--filter", "._-");
 		else
-			write_sbuf(fd, " --exclude-from=-");
+			err |= write_opt("--exclude-from", "-");
 	}
-	for (i = 1; i < argc - file_arg_cnt; i++) {
-		p = argv[i];
+
+	/* Elide the filename args from the option list, but scan for them in reverse. */
+	for (i = raw_argc-1, j = cooked_argc-1; i > 0 && j >= 0; i--) {
+		if (strcmp(raw_argv[i], cooked_argv[j]) == 0) {
+			raw_argv[i] = NULL;
+			j--;
+		}
+	}
+
+	for (i = 1; i < raw_argc; i++) {
+		if (!(p = raw_argv[i]))
+			continue;
 		if (strncmp(p, "--files-from", 12) == 0
-		    || strncmp(p, "--filter", 8) == 0
-		    || strncmp(p, "--include", 9) == 0
-		    || strncmp(p, "--exclude", 9) == 0) {
+		 || strncmp(p, "--filter", 8) == 0
+		 || strncmp(p, "--include", 9) == 0
+		 || strncmp(p, "--exclude", 9) == 0) {
 			if (strchr(p, '=') == NULL)
 				i++;
 			continue;
@@ -174,61 +289,24 @@ void write_batch_shell_file(int argc, char *argv[], int file_arg_cnt)
 			i++;
 			continue;
 		}
-		write(fd, " ", 1);
-		if (strncmp(p, "--write-batch", 13) == 0) {
-			write(fd, "--read-batch", 12);
-			if (p[13] == '=') {
-				write(fd, "=", 1);
-				write_arg(fd, p + 14);
-			}
-		} else
-			write_arg(fd, p);
+		if (strncmp(p, "--write-batch", len = 13) == 0
+		 || strncmp(p, "--only-write-batch", len = 18) == 0)
+			err |= write_opt("--read-batch", p[len] == '=' ? p + len + 1 : NULL);
+		else {
+			err |= write(batch_sh_fd, " ", 1) != 1;
+			err |= write_arg(p);
+		}
 	}
-	if (!(p = check_for_hostspec(argv[argc - 1], &p, &i)))
-		p = argv[argc - 1];
-	write(fd, " ${1:-", 6);
-	write_arg(fd, p);
-	write_byte(fd, '}');
+	if (!(p = check_for_hostspec(cooked_argv[cooked_argc - 1], &p2, &i)))
+		p = cooked_argv[cooked_argc - 1];
+	err |= write_opt("${1:-", NULL);
+	err |= write_arg(p);
+	err |= write(batch_sh_fd, "}", 1) != 1;
 	if (filter_list.head)
-		write_filter_rules(fd);
-	if (write(fd, "\n", 1) != 1 || close(fd) < 0) {
-		rsyserr(FERROR, errno, "Batch file %s write error",
-			safe_fname(filename));
-		exit_cleanup(1);
+		write_filter_rules(batch_sh_fd);
+	if (write(batch_sh_fd, "\n", 1) != 1 || close(batch_sh_fd) < 0 || err) {
+		rsyserr(FERROR, errno, "Batch file %s.sh write error", batch_name);
+		exit_cleanup(RERR_FILEIO);
 	}
-}
-
-void show_flist(int index, struct file_struct **fptr)
-{
-	/*  for debugging    show_flist(flist->count, flist->files * */
-
-	int i;
-	for (i = 0; i < index; i++) {
-		rprintf(FINFO, "flist->flags=%#x\n", fptr[i]->flags);
-		rprintf(FINFO, "flist->modtime=%#lx\n",
-			(long unsigned) fptr[i]->modtime);
-		rprintf(FINFO, "flist->length=%.0f\n",
-			(double) fptr[i]->length);
-		rprintf(FINFO, "flist->mode=%#o\n", (int) fptr[i]->mode);
-		rprintf(FINFO, "flist->basename=%s\n",
-			safe_fname(fptr[i]->basename));
-		if (fptr[i]->dirname) {
-			rprintf(FINFO, "flist->dirname=%s\n",
-				safe_fname(fptr[i]->dirname));
-		}
-		if (am_sender && fptr[i]->dir.root) {
-			rprintf(FINFO, "flist->dir.root=%s\n",
-				safe_fname(fptr[i]->dir.root));
-		}
-	}
-}
-
-/* for debugging */
-void show_argvs(int argc, char *argv[])
-{
-	int i;
-
-	rprintf(FINFO, "BATCH.C:show_argvs,argc=%d\n", argc);
-	for (i = 0; i < argc; i++)
-		rprintf(FINFO, "i=%d,argv[i]=%s\n", i, safe_fname(argv[i]));
+	batch_sh_fd = -1;
 }

byteorder.h

@@ -1,27 +1,28 @@
-/* 
-   simple byteorder handling
-   Copyright (C) Andrew Tridgell 1992-1995
-   
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
+/*
+ * Simple byteorder handling.
+ *
+ * Copyright (C) 1992-1995 Andrew Tridgell
+ * Copyright (C) 2007-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
 #undef CAREFUL_ALIGNMENT
 
-/* we know that the x86 can handle misalignment and has the "right" 
-   byteorder */
-#ifdef __i386__
+/* We know that the x86 can handle misalignment and has the same
+ * byte order (LSB-first) as the 32-bit numbers we transmit. */
+#if defined __i386__ || defined __i486__ || defined __i586__ || defined __i686__ || __amd64
 #define CAREFUL_ALIGNMENT 0
 #endif
 
@@ -30,25 +31,102 @@
 #endif
 
 #define CVAL(buf,pos) (((unsigned char *)(buf))[pos])
-#define PVAL(buf,pos) ((unsigned)CVAL(buf,pos))
-#define SCVAL(buf,pos,val) (CVAL(buf,pos) = (val))
-
+#define UVAL(buf,pos) ((uint32)CVAL(buf,pos))
 
 #if CAREFUL_ALIGNMENT
-#define SVAL(buf,pos) (PVAL(buf,pos)|PVAL(buf,(pos)+1)<<8)
-#define IVAL(buf,pos) (SVAL(buf,pos)|SVAL(buf,(pos)+2)<<16)
-#define SSVALX(buf,pos,val) (CVAL(buf,pos)=(val)&0xFF,CVAL(buf,pos+1)=(val)>>8)
-#define SIVALX(buf,pos,val) (SSVALX(buf,pos,val&0xFFFF),SSVALX(buf,pos+2,val>>16))
-#define SIVAL(buf,pos,val) SIVALX((buf),(pos),((uint32)(val)))
-#else
-/* this handles things for architectures like the 386 that can handle
-   alignment errors */
-/*
-   WARNING: This section is dependent on the length of int32
-   being correct. set CAREFUL_ALIGNMENT if it is not.
-*/
-#define IVAL(buf,pos) (*(uint32 *)((char *)(buf) + (pos)))
-#define SIVAL(buf,pos,val) IVAL(buf,pos)=((uint32)(val))
-#endif
 
+static inline uint32
+IVALu(const uchar *buf, int pos)
+{
+	return UVAL(buf, pos)
+	     | UVAL(buf, pos + 1) << 8
+	     | UVAL(buf, pos + 2) << 16
+	     | UVAL(buf, pos + 3) << 24;
+}
+
+static inline void
+SIVALu(uchar *buf, int pos, uint32 val)
+{
+	CVAL(buf, pos)     = val;
+	CVAL(buf, pos + 1) = val >> 8;
+	CVAL(buf, pos + 2) = val >> 16;
+	CVAL(buf, pos + 3) = val >> 24;
+}
+
+static inline int64
+IVAL64(const char *buf, int pos)
+{
+	return IVALu((uchar*)buf, pos) | (int64)IVALu((uchar*)buf, pos + 4) << 32;
+}
+
+static inline void
+SIVAL64(char *buf, int pos, int64 val)
+{
+	SIVALu((uchar*)buf, pos, val);
+	SIVALu((uchar*)buf, pos + 4, val >> 32);
+}
+
+#else /* !CAREFUL_ALIGNMENT */
+
+/* This handles things for architectures like the 386 that can handle alignment errors.
+ * WARNING: This section is dependent on the length of an int32 (and thus a uint32)
+ * being correct (4 bytes)!  Set CAREFUL_ALIGNMENT if it is not. */
+
+static inline uint32
+IVALu(const uchar *buf, int pos)
+{
+	union {
+		const uchar *b;
+		const uint32 *num;
+	} u;
+	u.b = buf + pos;
+	return *u.num;
+}
+
+static inline void
+SIVALu(uchar *buf, int pos, uint32 val)
+{
+	union {
+		uchar *b;
+		uint32 *num;
+	} u;
+	u.b = buf + pos;
+	*u.num = val;
+}
+
+static inline int64
+IVAL64(const char *buf, int pos)
+{
+	union {
+		const char *b;
+		const int64 *num;
+	} u;
+	u.b = buf + pos;
+	return *u.num;
+}
+
+static inline void
+SIVAL64(char *buf, int pos, int64 val)
+{
+	union {
+		char *b;
+		int64 *num;
+	} u;
+	u.b = buf + pos;
+	*u.num = val;
+}
+
+#endif /* !CAREFUL_ALIGNMENT */
+
+static inline uint32
+IVAL(const char *buf, int pos)
+{
+	return IVALu((uchar*)buf, pos);
+}
+
+static inline void
+SIVAL(char *buf, int pos, uint32 val)
+{
+	SIVALu((uchar*)buf, pos, val);
+}
 

case_N.h

@@ -0,0 +1,92 @@
+/*
+ * Allow an arbitrary sequence of case labels.
+ *
+ * Copyright (C) 2006-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
+
+/* This is included multiple times, once for every segment in a switch statement.
+ * This produces the next "case N:" statement in sequence. */
+
+#if !defined CASE_N_STATE_0
+#define CASE_N_STATE_0
+	case 0:
+#elif !defined CASE_N_STATE_1
+#define CASE_N_STATE_1
+	/* FALLTHROUGH */
+	case 1:
+#elif !defined CASE_N_STATE_2
+#define CASE_N_STATE_2
+	/* FALLTHROUGH */
+	case 2:
+#elif !defined CASE_N_STATE_3
+#define CASE_N_STATE_3
+	/* FALLTHROUGH */
+	case 3:
+#elif !defined CASE_N_STATE_4
+#define CASE_N_STATE_4
+	/* FALLTHROUGH */
+	case 4:
+#elif !defined CASE_N_STATE_5
+#define CASE_N_STATE_5
+	/* FALLTHROUGH */
+	case 5:
+#elif !defined CASE_N_STATE_6
+#define CASE_N_STATE_6
+	/* FALLTHROUGH */
+	case 6:
+#elif !defined CASE_N_STATE_7
+#define CASE_N_STATE_7
+	/* FALLTHROUGH */
+	case 7:
+#elif !defined CASE_N_STATE_8
+#define CASE_N_STATE_8
+	/* FALLTHROUGH */
+	case 8:
+#elif !defined CASE_N_STATE_9
+#define CASE_N_STATE_9
+	/* FALLTHROUGH */
+	case 9:
+#elif !defined CASE_N_STATE_10
+#define CASE_N_STATE_10
+	/* FALLTHROUGH */
+	case 10:
+#elif !defined CASE_N_STATE_11
+#define CASE_N_STATE_11
+	/* FALLTHROUGH */
+	case 11:
+#elif !defined CASE_N_STATE_12
+#define CASE_N_STATE_12
+	/* FALLTHROUGH */
+	case 12:
+#elif !defined CASE_N_STATE_13
+#define CASE_N_STATE_13
+	/* FALLTHROUGH */
+	case 13:
+#elif !defined CASE_N_STATE_14
+#define CASE_N_STATE_14
+	/* FALLTHROUGH */
+	case 14:
+#elif !defined CASE_N_STATE_15
+#define CASE_N_STATE_15
+	/* FALLTHROUGH */
+	case 15:
+#elif !defined CASE_N_STATE_16
+#define CASE_N_STATE_16
+	/* FALLTHROUGH */
+	case 16:
+#else
+#error Need to add more case statements!
+#endif

checksum.c

@@ -1,143 +1,510 @@
 /*
-   Copyright (C) Andrew Tridgell 1996
-   Copyright (C) Paul Mackerras 1996
-
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
+ * Routines to support checksumming of bytes.
+ *
+ * Copyright (C) 1996 Andrew Tridgell
+ * Copyright (C) 1996 Paul Mackerras
+ * Copyright (C) 2004-2020 Wayne Davison
+ *
+ * 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 3 of the License, or
+ * (at your option) any later version.
+ *
+ * In addition, as a special exception, the copyright holders give
+ * permission to dynamically link rsync with the OpenSSL and xxhash
+ * libraries when those libraries are being distributed in compliance
+ * with their license terms, and to distribute a dynamically linked
+ * combination of rsync and these libraries.  This is also considered
+ * to be covered under the GPL's System Libraries exception.
+ *
+ * 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, visit the http://fsf.org website.
+ */
 
 #include "rsync.h"
 
-int csum_length=2; /* initial value */
-
-#define CSUM_CHUNK 64
+#ifdef SUPPORT_XXHASH
+#include "xxhash.h"
+# if XXH_VERSION_NUMBER >= 800
+#  define SUPPORT_XXH3 1
+# endif
+#endif
 
+extern int am_server;
+extern int whole_file;
 extern int checksum_seed;
 extern int protocol_version;
+extern int proper_seed_order;
+extern const char *checksum_choice;
 
+struct name_num_obj valid_checksums = {
+	"checksum", NULL, NULL, 0, 0, {
+#ifdef SUPPORT_XXH3
+		{ CSUM_XXH3_128, "xxh128", NULL },
+		{ CSUM_XXH3_64, "xxh3", NULL },
+#endif
+#ifdef SUPPORT_XXHASH
+		{ CSUM_XXH64, "xxh64", NULL },
+		{ CSUM_XXH64, "xxhash", NULL },
+#endif
+		{ CSUM_MD5, "md5", NULL },
+		{ CSUM_MD4, "md4", NULL },
+		{ CSUM_NONE, "none", NULL },
+		{ 0, NULL, NULL }
+	}
+};
+
+int xfersum_type = 0; /* used for the file transfer checksums */
+int checksum_type = 0; /* used for the pre-transfer (--checksum) checksums */
+
+int parse_csum_name(const char *name, int len)
+{
+	struct name_num_item *nni;
+
+	if (len < 0 && name)
+		len = strlen(name);
+
+	if (!name || (len == 4 && strncasecmp(name, "auto", 4) == 0)) {
+		if (protocol_version >= 30)
+			return CSUM_MD5;
+		if (protocol_version >= 27)
+			return CSUM_MD4_OLD;
+		if (protocol_version >= 21)
+			return CSUM_MD4_BUSTED;
+		return CSUM_MD4_ARCHAIC;
+	}
+
+	nni = get_nni_by_name(&valid_checksums, name, len);
+
+	if (!nni) {
+		rprintf(FERROR, "unknown checksum name: %s\n", name);
+		exit_cleanup(RERR_UNSUPPORTED);
+	}
+
+	return nni->num;
+}
+
+static const char *checksum_name(int num)
+{
+	struct name_num_item *nni = get_nni_by_num(&valid_checksums, num);
+
+	return nni ? nni->name : num < CSUM_MD4 ? "md4" : "UNKNOWN";
+}
+
+void parse_checksum_choice(int final_call)
+{
+	if (valid_checksums.negotiated_name)
+		xfersum_type = checksum_type = valid_checksums.negotiated_num;
+	else {
+		char *cp = checksum_choice ? strchr(checksum_choice, ',') : NULL;
+		if (cp) {
+			xfersum_type = parse_csum_name(checksum_choice, cp - checksum_choice);
+			checksum_type = parse_csum_name(cp+1, -1);
+		} else
+			xfersum_type = checksum_type = parse_csum_name(checksum_choice, -1);
+		if (am_server && checksum_choice)
+			validate_choice_vs_env(NSTR_CHECKSUM, xfersum_type, checksum_type);
+	}
+
+	if (xfersum_type == CSUM_NONE)
+		whole_file = 1;
+
+	/* Snag the checksum name for both write_batch's option output & the following debug output. */
+	if (valid_checksums.negotiated_name)
+		checksum_choice = valid_checksums.negotiated_name;
+	else if (checksum_choice == NULL)
+		checksum_choice = checksum_name(xfersum_type);
+
+	if (final_call && DEBUG_GTE(NSTR, am_server ? 3 : 1)) {
+		rprintf(FINFO, "%s%s checksum: %s\n",
+			am_server ? "Server" : "Client",
+			valid_checksums.negotiated_name ? " negotiated" : "",
+			checksum_choice);
+	}
+}
+
+int csum_len_for_type(int cst, BOOL flist_csum)
+{
+	switch (cst) {
+	  case CSUM_NONE:
+		return 1;
+	  case CSUM_MD4_ARCHAIC:
+		/* The oldest checksum code is rather weird: the file-list code only sent
+		 * 2-byte checksums, but all other checksums were full MD4 length. */
+		return flist_csum ? 2 : MD4_DIGEST_LEN;
+	  case CSUM_MD4:
+	  case CSUM_MD4_OLD:
+	  case CSUM_MD4_BUSTED:
+		return MD4_DIGEST_LEN;
+	  case CSUM_MD5:
+		return MD5_DIGEST_LEN;
+	  case CSUM_XXH64:
+	  case CSUM_XXH3_64:
+		return 64/8;
+	  case CSUM_XXH3_128:
+		return 128/8;
+	  default: /* paranoia to prevent missing case values */
+		exit_cleanup(RERR_UNSUPPORTED);
+	}
+	return 0;
+}
+
+/* Returns 0 if the checksum is not canonical (i.e. it includes a seed value).
+ * Returns 1 if the public sum order matches our internal sum order.
+ * Returns -1 if the public sum order is the reverse of our internal sum order.
+ */
+int canonical_checksum(int csum_type)
+{
+	switch (csum_type) {
+	  case CSUM_NONE:
+	  case CSUM_MD4_ARCHAIC:
+	  case CSUM_MD4_OLD:
+	  case CSUM_MD4_BUSTED:
+		break;
+	  case CSUM_MD4:
+	  case CSUM_MD5:
+		return -1;
+	  case CSUM_XXH64:
+	  case CSUM_XXH3_64:
+	  case CSUM_XXH3_128:
+		return 1;
+	  default: /* paranoia to prevent missing case values */
+		exit_cleanup(RERR_UNSUPPORTED);
+	}
+	return 0;
+}
+
+#ifndef HAVE_SIMD /* See simd-checksum-*.cpp. */
 /*
-  a simple 32 bit checksum that can be upadted from either end
+  a simple 32 bit checksum that can be updated from either end
   (inspired by Mark Adler's Adler-32 checksum)
   */
 uint32 get_checksum1(char *buf1, int32 len)
 {
-    int32 i;
-    uint32 s1, s2;
-    schar *buf = (schar *)buf1;
+	int32 i;
+	uint32 s1, s2;
+	schar *buf = (schar *)buf1;
 
-    s1 = s2 = 0;
-    for (i = 0; i < (len-4); i+=4) {
-	s2 += 4*(s1 + buf[i]) + 3*buf[i+1] + 2*buf[i+2] + buf[i+3] +
-	  10*CHAR_OFFSET;
-	s1 += (buf[i+0] + buf[i+1] + buf[i+2] + buf[i+3] + 4*CHAR_OFFSET);
-    }
-    for (; i < len; i++) {
-	s1 += (buf[i]+CHAR_OFFSET); s2 += s1;
-    }
-    return (s1 & 0xffff) + (s2 << 16);
+	s1 = s2 = 0;
+	for (i = 0; i < (len-4); i+=4) {
+		s2 += 4*(s1 + buf[i]) + 3*buf[i+1] + 2*buf[i+2] + buf[i+3] + 10*CHAR_OFFSET;
+		s1 += (buf[i+0] + buf[i+1] + buf[i+2] + buf[i+3] + 4*CHAR_OFFSET);
+	}
+	for (; i < len; i++) {
+		s1 += (buf[i]+CHAR_OFFSET); s2 += s1;
+	}
+	return (s1 & 0xffff) + (s2 << 16);
 }
-
+#endif
 
 void get_checksum2(char *buf, int32 len, char *sum)
 {
-	int32 i;
-	static char *buf1;
-	static int32 len1;
-	struct mdfour m;
+	switch (xfersum_type) {
+#ifdef SUPPORT_XXHASH
+	  case CSUM_XXH64:
+		SIVAL64(sum, 0, XXH64(buf, len, checksum_seed));
+		break;
+#endif
+#ifdef SUPPORT_XXH3
+	  case CSUM_XXH3_64:
+		SIVAL64(sum, 0, XXH3_64bits_withSeed(buf, len, checksum_seed));
+		break;
+	  case CSUM_XXH3_128: {
+		XXH128_hash_t digest = XXH3_128bits_withSeed(buf, len, checksum_seed);
+		SIVAL64(sum, 0, digest.low64);
+		SIVAL64(sum, 8, digest.high64);
+		break;
+	  }
+#endif
+	  case CSUM_MD5: {
+		MD5_CTX m5;
+		uchar seedbuf[4];
+		MD5_Init(&m5);
+		if (proper_seed_order) {
+			if (checksum_seed) {
+				SIVALu(seedbuf, 0, checksum_seed);
+				MD5_Update(&m5, seedbuf, 4);
+			}
+			MD5_Update(&m5, (uchar *)buf, len);
+		} else {
+			MD5_Update(&m5, (uchar *)buf, len);
+			if (checksum_seed) {
+				SIVALu(seedbuf, 0, checksum_seed);
+				MD5_Update(&m5, seedbuf, 4);
+			}
+		}
+		MD5_Final((uchar *)sum, &m5);
+		break;
+	  }
+	  case CSUM_MD4:
+#ifdef USE_OPENSSL
+	  {
+		MD4_CTX m4;
+		MD4_Init(&m4);
+		MD4_Update(&m4, (uchar *)buf, len);
+		if (checksum_seed) {
+			uchar seedbuf[4];
+			SIVALu(seedbuf, 0, checksum_seed);
+			MD4_Update(&m4, seedbuf, 4);
+		}
+		MD4_Final((uchar *)sum, &m4);
+		break;
+	  }
+#endif
+	  case CSUM_MD4_OLD:
+	  case CSUM_MD4_BUSTED:
+	  case CSUM_MD4_ARCHAIC: {
+		md_context m;
+		int32 i;
+		static char *buf1;
+		static int32 len1;
 
-	if (len > len1) {
-		if (buf1)
-			free(buf1);
-		buf1 = new_array(char, len+4);
-		len1 = len;
-		if (!buf1)
-			out_of_memory("get_checksum2");
+		mdfour_begin(&m);
+
+		if (len > len1) {
+			if (buf1)
+				free(buf1);
+			buf1 = new_array(char, len+4);
+			len1 = len;
+		}
+
+		memcpy(buf1, buf, len);
+		if (checksum_seed) {
+			SIVAL(buf1,len,checksum_seed);
+			len += 4;
+		}
+
+		for (i = 0; i + CSUM_CHUNK <= len; i += CSUM_CHUNK)
+			mdfour_update(&m, (uchar *)(buf1+i), CSUM_CHUNK);
+
+		/*
+		 * Prior to version 27 an incorrect MD4 checksum was computed
+		 * by failing to call mdfour_tail() for block sizes that
+		 * are multiples of 64.  This is fixed by calling mdfour_update()
+		 * even when there are no more bytes.
+		 */
+		if (len - i > 0 || xfersum_type > CSUM_MD4_BUSTED)
+			mdfour_update(&m, (uchar *)(buf1+i), len-i);
+
+		mdfour_result(&m, (uchar *)sum);
+		break;
+	  }
+	  default: /* paranoia to prevent missing case values */
+		exit_cleanup(RERR_UNSUPPORTED);
 	}
-
-	mdfour_begin(&m);
-
-	memcpy(buf1,buf,len);
-	if (checksum_seed) {
-		SIVAL(buf1,len,checksum_seed);
-		len += 4;
-	}
-
-	for(i = 0; i + CSUM_CHUNK <= len; i += CSUM_CHUNK) {
-		mdfour_update(&m, (uchar *)(buf1+i), CSUM_CHUNK);
-	}
-	/*
-	 * Prior to version 27 an incorrect MD4 checksum was computed
-	 * by failing to call mdfour_tail() for block sizes that
-	 * are multiples of 64.  This is fixed by calling mdfour_update()
-	 * even when there are no more bytes.
-	 */
-	if (len - i > 0 || protocol_version >= 27) {
-		mdfour_update(&m, (uchar *)(buf1+i), (len-i));
-	}
-
-	mdfour_result(&m, (uchar *)sum);
 }
 
-
-void file_checksum(char *fname,char *sum,OFF_T size)
+void file_checksum(const char *fname, const STRUCT_STAT *st_p, char *sum)
 {
-	OFF_T i;
 	struct map_struct *buf;
+	OFF_T i, len = st_p->st_size;
+	int32 remainder;
 	int fd;
-	OFF_T len = size;
-	struct mdfour m;
 
-	memset(sum,0,MD4_SUM_LENGTH);
+	memset(sum, 0, MAX_DIGEST_LEN);
 
 	fd = do_open(fname, O_RDONLY, 0);
 	if (fd == -1)
 		return;
 
-	buf = map_file(fd, size, MAX_MAP_SIZE, CSUM_CHUNK);
+	buf = map_file(fd, len, MAX_MAP_SIZE, CHUNK_SIZE);
 
-	mdfour_begin(&m);
+	switch (checksum_type) {
+#ifdef SUPPORT_XXHASH
+	  case CSUM_XXH64: {
+		static XXH64_state_t* state = NULL;
+		if (!state && !(state = XXH64_createState()))
+			out_of_memory("file_checksum");
 
-	for(i = 0; i + CSUM_CHUNK <= len; i += CSUM_CHUNK) {
-		mdfour_update(&m, (uchar *)map_ptr(buf, i, CSUM_CHUNK),
-			      CSUM_CHUNK);
+		XXH64_reset(state, 0);
+
+		for (i = 0; i + CHUNK_SIZE <= len; i += CHUNK_SIZE)
+			XXH64_update(state, (uchar *)map_ptr(buf, i, CHUNK_SIZE), CHUNK_SIZE);
+
+		remainder = (int32)(len - i);
+		if (remainder > 0)
+			XXH64_update(state, (uchar *)map_ptr(buf, i, remainder), remainder);
+
+		SIVAL64(sum, 0, XXH64_digest(state));
+		break;
+	  }
+#endif
+#ifdef SUPPORT_XXH3
+	  case CSUM_XXH3_64: {
+		static XXH3_state_t* state = NULL;
+		if (!state && !(state = XXH3_createState()))
+			out_of_memory("file_checksum");
+
+		XXH3_64bits_reset(state);
+
+		for (i = 0; i + CHUNK_SIZE <= len; i += CHUNK_SIZE)
+			XXH3_64bits_update(state, (uchar *)map_ptr(buf, i, CHUNK_SIZE), CHUNK_SIZE);
+
+		remainder = (int32)(len - i);
+		if (remainder > 0)
+			XXH3_64bits_update(state, (uchar *)map_ptr(buf, i, remainder), remainder);
+
+		SIVAL64(sum, 0, XXH3_64bits_digest(state));
+		break;
+	  }
+	  case CSUM_XXH3_128: {
+		XXH128_hash_t digest;
+		static XXH3_state_t* state = NULL;
+		if (!state && !(state = XXH3_createState()))
+			out_of_memory("file_checksum");
+
+		XXH3_128bits_reset(state);
+
+		for (i = 0; i + CHUNK_SIZE <= len; i += CHUNK_SIZE)
+			XXH3_128bits_update(state, (uchar *)map_ptr(buf, i, CHUNK_SIZE), CHUNK_SIZE);
+
+		remainder = (int32)(len - i);
+		if (remainder > 0)
+			XXH3_128bits_update(state, (uchar *)map_ptr(buf, i, remainder), remainder);
+
+		digest = XXH3_128bits_digest(state);
+		SIVAL64(sum, 0, digest.low64);
+		SIVAL64(sum, 8, digest.high64);
+		break;
+	  }
+#endif
+	  case CSUM_MD5: {
+		MD5_CTX m5;
+
+		MD5_Init(&m5);
+
+		for (i = 0; i + CHUNK_SIZE <= len; i += CHUNK_SIZE)
+			MD5_Update(&m5, (uchar *)map_ptr(buf, i, CHUNK_SIZE), CHUNK_SIZE);
+
+		remainder = (int32)(len - i);
+		if (remainder > 0)
+			MD5_Update(&m5, (uchar *)map_ptr(buf, i, remainder), remainder);
+
+		MD5_Final((uchar *)sum, &m5);
+		break;
+	  }
+	  case CSUM_MD4:
+#ifdef USE_OPENSSL
+	  {
+		MD4_CTX m4;
+
+		MD4_Init(&m4);
+
+		for (i = 0; i + CHUNK_SIZE <= len; i += CHUNK_SIZE)
+			MD4_Update(&m4, (uchar *)map_ptr(buf, i, CHUNK_SIZE), CHUNK_SIZE);
+
+		remainder = (int32)(len - i);
+		if (remainder > 0)
+			MD4_Update(&m4, (uchar *)map_ptr(buf, i, remainder), remainder);
+
+		MD4_Final((uchar *)sum, &m4);
+		break;
+	  }
+#endif
+	  case CSUM_MD4_OLD:
+	  case CSUM_MD4_BUSTED:
+	  case CSUM_MD4_ARCHAIC: {
+		md_context m;
+
+		mdfour_begin(&m);
+
+		for (i = 0; i + CHUNK_SIZE <= len; i += CHUNK_SIZE)
+			mdfour_update(&m, (uchar *)map_ptr(buf, i, CHUNK_SIZE), CHUNK_SIZE);
+
+		/* Prior to version 27 an incorrect MD4 checksum was computed
+		 * by failing to call mdfour_tail() for block sizes that
+		 * are multiples of 64.  This is fixed by calling mdfour_update()
+		 * even when there are no more bytes. */
+		remainder = (int32)(len - i);
+		if (remainder > 0 || checksum_type > CSUM_MD4_BUSTED)
+			mdfour_update(&m, (uchar *)map_ptr(buf, i, remainder), remainder);
+
+		mdfour_result(&m, (uchar *)sum);
+		break;
+	  }
+	  default:
+		rprintf(FERROR, "Invalid checksum-choice for --checksum: %s (%d)\n",
+			checksum_name(checksum_type), checksum_type);
+		exit_cleanup(RERR_UNSUPPORTED);
 	}
 
-	/* Prior to version 27 an incorrect MD4 checksum was computed
-	 * by failing to call mdfour_tail() for block sizes that
-	 * are multiples of 64.  This is fixed by calling mdfour_update()
-	 * even when there are no more bytes. */
-	if (len - i > 0 || protocol_version >= 27)
-		mdfour_update(&m, (uchar *)map_ptr(buf, i, len-i), len-i);
-
-	mdfour_result(&m, (uchar *)sum);
-
 	close(fd);
 	unmap_file(buf);
 }
 
-
 static int32 sumresidue;
-static char sumrbuf[CSUM_CHUNK];
-static struct mdfour md;
+static union {
+	md_context md;
+#ifdef USE_OPENSSL
+	MD4_CTX m4;
+#endif
+	MD5_CTX m5;
+} ctx;
+#ifdef SUPPORT_XXHASH
+static XXH64_state_t* xxh64_state;
+#endif
+#ifdef SUPPORT_XXH3
+static XXH3_state_t* xxh3_state;
+#endif
+static int cursum_type;
 
-void sum_init(int seed)
+void sum_init(int csum_type, int seed)
 {
 	char s[4];
-	mdfour_begin(&md);
-	sumresidue = 0;
-	SIVAL(s, 0, seed);
-	sum_update(s, 4);
+
+	if (csum_type < 0)
+		csum_type = parse_csum_name(NULL, 0);
+	cursum_type = csum_type;
+
+	switch (csum_type) {
+#ifdef SUPPORT_XXHASH
+	  case CSUM_XXH64:
+		if (!xxh64_state && !(xxh64_state = XXH64_createState()))
+			out_of_memory("sum_init");
+		XXH64_reset(xxh64_state, 0);
+		break;
+#endif
+#ifdef SUPPORT_XXH3
+	  case CSUM_XXH3_64:
+		if (!xxh3_state && !(xxh3_state = XXH3_createState()))
+			out_of_memory("sum_init");
+		XXH3_64bits_reset(xxh3_state);
+		break;
+	  case CSUM_XXH3_128:
+		if (!xxh3_state && !(xxh3_state = XXH3_createState()))
+			out_of_memory("sum_init");
+		XXH3_128bits_reset(xxh3_state);
+		break;
+#endif
+	  case CSUM_MD5:
+		MD5_Init(&ctx.m5);
+		break;
+	  case CSUM_MD4:
+#ifdef USE_OPENSSL
+		MD4_Init(&ctx.m4);
+#else
+		mdfour_begin(&ctx.md);
+		sumresidue = 0;
+#endif
+		break;
+	  case CSUM_MD4_OLD:
+	  case CSUM_MD4_BUSTED:
+	  case CSUM_MD4_ARCHAIC:
+		mdfour_begin(&ctx.md);
+		sumresidue = 0;
+		SIVAL(s, 0, seed);
+		sum_update(s, 4);
+		break;
+	  case CSUM_NONE:
+		break;
+	  default: /* paranoia to prevent missing case values */
+		exit_cleanup(RERR_UNSUPPORTED);
+	}
 }
 
 /**
@@ -148,37 +515,111 @@ void sum_init(int seed)
  * @todo Perhaps get rid of md and just pass in the address each time.
  * Very slightly clearer and slower.
  **/
-void sum_update(char *p, int32 len)
+void sum_update(const char *p, int32 len)
 {
-	if (len + sumresidue < CSUM_CHUNK) {
-		memcpy(sumrbuf + sumresidue, p, len);
-		sumresidue += len;
-		return;
-	}
+	switch (cursum_type) {
+#ifdef SUPPORT_XXHASH
+	  case CSUM_XXH64:
+		XXH64_update(xxh64_state, p, len);
+		break;
+#endif
+#ifdef SUPPORT_XXH3
+	  case CSUM_XXH3_64:
+		XXH3_64bits_update(xxh3_state, p, len);
+		break;
+	  case CSUM_XXH3_128:
+		XXH3_128bits_update(xxh3_state, p, len);
+		break;
+#endif
+	  case CSUM_MD5:
+		MD5_Update(&ctx.m5, (uchar *)p, len);
+		break;
+	  case CSUM_MD4:
+#ifdef USE_OPENSSL
+		MD4_Update(&ctx.m4, (uchar *)p, len);
+		break;
+#endif
+	  case CSUM_MD4_OLD:
+	  case CSUM_MD4_BUSTED:
+	  case CSUM_MD4_ARCHAIC:
+		if (len + sumresidue < CSUM_CHUNK) {
+			memcpy(ctx.md.buffer + sumresidue, p, len);
+			sumresidue += len;
+			break;
+		}
 
-	if (sumresidue) {
-		int32 i = CSUM_CHUNK - sumresidue;
-		memcpy(sumrbuf + sumresidue, p, i);
-		mdfour_update(&md, (uchar *)sumrbuf, CSUM_CHUNK);
-		len -= i;
-		p += i;
-	}
+		if (sumresidue) {
+			int32 i = CSUM_CHUNK - sumresidue;
+			memcpy(ctx.md.buffer + sumresidue, p, i);
+			mdfour_update(&ctx.md, (uchar *)ctx.md.buffer, CSUM_CHUNK);
+			len -= i;
+			p += i;
+		}
 
-	while (len >= CSUM_CHUNK) {
-		mdfour_update(&md, (uchar *)p, CSUM_CHUNK);
-		len -= CSUM_CHUNK;
-		p += CSUM_CHUNK;
-	}
+		while (len >= CSUM_CHUNK) {
+			mdfour_update(&ctx.md, (uchar *)p, CSUM_CHUNK);
+			len -= CSUM_CHUNK;
+			p += CSUM_CHUNK;
+		}
 
-	sumresidue = len;
-	if (sumresidue)
-		memcpy(sumrbuf, p, sumresidue);
+		sumresidue = len;
+		if (sumresidue)
+			memcpy(ctx.md.buffer, p, sumresidue);
+		break;
+	  case CSUM_NONE:
+		break;
+	  default: /* paranoia to prevent missing case values */
+		exit_cleanup(RERR_UNSUPPORTED);
+	}
 }
 
-void sum_end(char *sum)
+/* NOTE: all the callers of sum_end() pass in a pointer to a buffer that is
+ * MAX_DIGEST_LEN in size, so even if the csum-len is shorter that that (i.e.
+ * CSUM_MD4_ARCHAIC), we don't have to worry about limiting the data we write
+ * into the "sum" buffer. */
+int sum_end(char *sum)
 {
-	if (sumresidue || protocol_version >= 27)
-		mdfour_update(&md, (uchar *)sumrbuf, sumresidue);
+	switch (cursum_type) {
+#ifdef SUPPORT_XXHASH
+	  case CSUM_XXH64:
+		SIVAL64(sum, 0, XXH64_digest(xxh64_state));
+		break;
+#endif
+#ifdef SUPPORT_XXH3
+	  case CSUM_XXH3_64:
+		SIVAL64(sum, 0, XXH3_64bits_digest(xxh3_state));
+		break;
+	  case CSUM_XXH3_128: {
+		XXH128_hash_t digest = XXH3_128bits_digest(xxh3_state);
+		SIVAL64(sum, 0, digest.low64);
+		SIVAL64(sum, 8, digest.high64);
+		break;
+	  }
+#endif
+	  case CSUM_MD5:
+		MD5_Final((uchar *)sum, &ctx.m5);
+		break;
+	  case CSUM_MD4:
+#ifdef USE_OPENSSL
+		MD4_Final((uchar *)sum, &ctx.m4);
+		break;
+#endif
+	  case CSUM_MD4_OLD:
+		mdfour_update(&ctx.md, (uchar *)ctx.md.buffer, sumresidue);
+		mdfour_result(&ctx.md, (uchar *)sum);
+		break;
+	  case CSUM_MD4_BUSTED:
+	  case CSUM_MD4_ARCHAIC:
+		if (sumresidue)
+			mdfour_update(&ctx.md, (uchar *)ctx.md.buffer, sumresidue);
+		mdfour_result(&ctx.md, (uchar *)sum);
+		break;
+	  case CSUM_NONE:
+		*sum = '\0';
+		break;
+	  default: /* paranoia to prevent missing case values */
+		exit_cleanup(RERR_UNSUPPORTED);
+	}
 
-	mdfour_result(&md, (uchar *)sum);
+	return csum_len_for_type(cursum_type, 0);
 }

chmod.c

@@ -0,0 +1,249 @@
+/*
+ * Implement the core of the --chmod option.
+ *
+ * Copyright (C) 2002 Scott Howard
+ * Copyright (C) 2005-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
+
+#include "rsync.h"
+#include "itypes.h"
+
+extern mode_t orig_umask;
+
+#define FLAG_X_KEEP (1<<0)
+#define FLAG_DIRS_ONLY (1<<1)
+#define FLAG_FILES_ONLY (1<<2)
+
+struct chmod_mode_struct {
+	struct chmod_mode_struct *next;
+	int ModeAND, ModeOR;
+	char flags;
+};
+
+#define CHMOD_ADD 1
+#define CHMOD_SUB 2
+#define CHMOD_EQ  3
+#define CHMOD_SET 4
+
+#define STATE_ERROR 0
+#define STATE_1ST_HALF 1
+#define STATE_2ND_HALF 2
+#define STATE_OCTAL_NUM 3
+
+/* Parse a chmod-style argument, and break it down into one or more AND/OR
+ * pairs in a linked list.  We return a pointer to new items on success
+ * (appending the items to the specified list), or NULL on error. */
+struct chmod_mode_struct *parse_chmod(const char *modestr,
+				      struct chmod_mode_struct **root_mode_ptr)
+{
+	int state = STATE_1ST_HALF;
+	int where = 0, what = 0, op = 0, topbits = 0, topoct = 0, flags = 0;
+	struct chmod_mode_struct *first_mode = NULL, *curr_mode = NULL,
+				 *prev_mode = NULL;
+
+	while (state != STATE_ERROR) {
+		if (!*modestr || *modestr == ',') {
+			int bits;
+
+			if (!op) {
+				state = STATE_ERROR;
+				break;
+			}
+			prev_mode = curr_mode;
+			curr_mode = new_array(struct chmod_mode_struct, 1);
+			if (prev_mode)
+				prev_mode->next = curr_mode;
+			else
+				first_mode = curr_mode;
+			curr_mode->next = NULL;
+
+			if (where)
+				bits = where * what;
+			else {
+				where = 0111;
+				bits = (where * what) & ~orig_umask;
+			}
+
+			switch (op) {
+			case CHMOD_ADD:
+				curr_mode->ModeAND = CHMOD_BITS;
+				curr_mode->ModeOR  = bits + topoct;
+				break;
+			case CHMOD_SUB:
+				curr_mode->ModeAND = CHMOD_BITS - bits - topoct;
+				curr_mode->ModeOR  = 0;
+				break;
+			case CHMOD_EQ:
+				curr_mode->ModeAND = CHMOD_BITS - (where * 7) - (topoct ? topbits : 0);
+				curr_mode->ModeOR  = bits + topoct;
+				break;
+			case CHMOD_SET:
+				curr_mode->ModeAND = 0;
+				curr_mode->ModeOR  = bits;
+				break;
+			}
+
+			curr_mode->flags = flags;
+
+			if (!*modestr)
+				break;
+			modestr++;
+
+			state = STATE_1ST_HALF;
+			where = what = op = topoct = topbits = flags = 0;
+		}
+
+		switch (state) {
+		case STATE_1ST_HALF:
+			switch (*modestr) {
+			case 'D':
+				if (flags & FLAG_FILES_ONLY)
+					state = STATE_ERROR;
+				flags |= FLAG_DIRS_ONLY;
+				break;
+			case 'F':
+				if (flags & FLAG_DIRS_ONLY)
+					state = STATE_ERROR;
+				flags |= FLAG_FILES_ONLY;
+				break;
+			case 'u':
+				where |= 0100;
+				topbits |= 04000;
+				break;
+			case 'g':
+				where |= 0010;
+				topbits |= 02000;
+				break;
+			case 'o':
+				where |= 0001;
+				break;
+			case 'a':
+				where |= 0111;
+				break;
+			case '+':
+				op = CHMOD_ADD;
+				state = STATE_2ND_HALF;
+				break;
+			case '-':
+				op = CHMOD_SUB;
+				state = STATE_2ND_HALF;
+				break;
+			case '=':
+				op = CHMOD_EQ;
+				state = STATE_2ND_HALF;
+				break;
+			default:
+				if (isDigit(modestr) && *modestr < '8' && !where) {
+					op = CHMOD_SET;
+					state =  STATE_OCTAL_NUM;
+					where = 1;
+					what = *modestr - '0';
+				} else
+					state = STATE_ERROR;
+				break;
+			}
+			break;
+		case STATE_2ND_HALF:
+			switch (*modestr) {
+			case 'r':
+				what |= 4;
+				break;
+			case 'w':
+				what |= 2;
+				break;
+			case 'X':
+				flags |= FLAG_X_KEEP;
+				/* FALL THROUGH */
+			case 'x':
+				what |= 1;
+				break;
+			case 's':
+				if (topbits)
+					topoct |= topbits;
+				else
+					topoct = 04000;
+				break;
+			case 't':
+				topoct |= 01000;
+				break;
+			default:
+				state = STATE_ERROR;
+				break;
+			}
+			break;
+		case STATE_OCTAL_NUM:
+			if (isDigit(modestr) && *modestr < '8') {
+				what = what*8 + *modestr - '0';
+				if (what > CHMOD_BITS)
+					state = STATE_ERROR;
+			} else
+				state = STATE_ERROR;
+			break;
+		}
+		modestr++;
+	}
+
+	if (state == STATE_ERROR) {
+		free_chmod_mode(first_mode);
+		return NULL;
+	}
+
+	if (!(curr_mode = *root_mode_ptr))
+		*root_mode_ptr = first_mode;
+	else {
+		while (curr_mode->next)
+			curr_mode = curr_mode->next;
+		curr_mode->next = first_mode;
+	}
+
+	return first_mode;
+}
+
+
+/* Takes an existing file permission and a list of AND/OR changes, and
+ * create a new permissions. */
+int tweak_mode(int mode, struct chmod_mode_struct *chmod_modes)
+{
+	int IsX = mode & 0111;
+	int NonPerm = mode & ~CHMOD_BITS;
+
+	for ( ; chmod_modes; chmod_modes = chmod_modes->next) {
+		if ((chmod_modes->flags & FLAG_DIRS_ONLY) && !S_ISDIR(NonPerm))
+			continue;
+		if ((chmod_modes->flags & FLAG_FILES_ONLY) && S_ISDIR(NonPerm))
+			continue;
+		mode &= chmod_modes->ModeAND;
+		if ((chmod_modes->flags & FLAG_X_KEEP) && !IsX && !S_ISDIR(NonPerm))
+			mode |= chmod_modes->ModeOR & ~0111;
+		else
+			mode |= chmod_modes->ModeOR;
+	}
+
+	return mode | NonPerm;
+}
+
+/* Free the linked list created by parse_chmod. */
+int free_chmod_mode(struct chmod_mode_struct *chmod_modes)
+{
+	struct chmod_mode_struct *next;
+
+	while (chmod_modes) {
+		next = chmod_modes->next;
+		free(chmod_modes);
+		chmod_modes = next;
+	}
+	return 0;
+}

cleanup.c

@@ -1,30 +1,47 @@
-/* -*- c-file-style: "linux" -*-
-
-   Copyright (C) 1996-2000 by Andrew Tridgell
-   Copyright (C) Paul Mackerras 1996
-   Copyright (C) 2002 by Martin Pool
-
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
+/*
+ * End-of-run cleanup routines.
+ *
+ * Copyright (C) 1996-2000 Andrew Tridgell
+ * Copyright (C) 1996 Paul Mackerras
+ * Copyright (C) 2002 Martin Pool
+ * Copyright (C) 2003-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
 #include "rsync.h"
 
+extern int dry_run;
+extern int am_server;
+extern int am_daemon;
+extern int am_receiver;
+extern int am_sender;
 extern int io_error;
 extern int keep_partial;
-extern int log_got_error;
+extern int got_xfer_error;
+extern int protocol_version;
+extern int output_needs_newline;
 extern char *partial_dir;
+extern char *logfile_name;
+
+int called_from_signal_handler = 0;
+BOOL shutting_down = False;
+BOOL flush_ok_after_signal = False;
+
+#ifdef HAVE_SIGACTION
+static struct sigaction sigact;
+#endif
 
 /**
  * Close all open sockets and files, allowing a (somewhat) graceful
@@ -70,10 +87,10 @@ void close_all(void)
  **/
 int cleanup_got_literal = 0;
 
-static char *cleanup_fname;
-static char *cleanup_new_fname;
+static const char *cleanup_fname;
+static const char *cleanup_new_fname;
 static struct file_struct *cleanup_file;
-static int cleanup_fd_r, cleanup_fd_w;
+static int cleanup_fd_r = -1, cleanup_fd_w = -1;
 static pid_t cleanup_pid = 0;
 
 pid_t cleanup_child_pid = -1;
@@ -83,91 +100,197 @@ pid_t cleanup_child_pid = -1;
  *
  * @param code one of the RERR_* codes from errcode.h.
  **/
-void _exit_cleanup(int code, const char *file, int line)
+NORETURN void _exit_cleanup(int code, const char *file, int line)
 {
-	int ocode = code;
-	static int inside_cleanup = 0;
+	static int switch_step = 0;
+	static int exit_code = 0, exit_line = 0;
+	static const char *exit_file = NULL;
+	static int first_code = 0;
 
-	if (inside_cleanup > 10) {
-		/* prevent the occasional infinite recursion */
-		return;
-	}
-	inside_cleanup++;
+	SIGACTION(SIGUSR1, SIG_IGN);
+	SIGACTION(SIGUSR2, SIG_IGN);
 
-	signal(SIGUSR1, SIG_IGN);
-	signal(SIGUSR2, SIG_IGN);
-
-	if (verbose > 3) {
-		rprintf(FINFO,"_exit_cleanup(code=%d, file=%s, line=%d): entered\n",
-			code, safe_fname(file), line);
+	if (!exit_code) { /* Preserve first error exit info when recursing. */
+		exit_code = code;
+		exit_file = file;
+		exit_line = line < 0 ? -line : line;
 	}
 
-	if (cleanup_child_pid != -1) {
-		int status;
-		if (waitpid(cleanup_child_pid, &status, WNOHANG) == cleanup_child_pid) {
-			status = WEXITSTATUS(status);
-			if (status > code)
-				code = status;
+	/* If this is the exit at the end of the run, the server side
+	 * should not attempt to output a message (see log_exit()). */
+	if (am_server && code == 0)
+		am_server = 2;
+
+	/* Some of our actions might cause a recursive call back here, so we
+	 * keep track of where we are in the cleanup and never repeat a step. */
+	switch (switch_step) {
+#include "case_N.h" /* case 0: */
+		switch_step++;
+
+		first_code = code;
+
+		if (output_needs_newline) {
+			fputc('\n', stdout);
+			output_needs_newline = 0;
 		}
-	}
 
-	if (cleanup_got_literal && cleanup_fname && keep_partial
-	    && handle_partial_dir(cleanup_new_fname, PDIR_CREATE)) {
-		char *fname = cleanup_fname;
-		cleanup_fname = NULL;
-		if (cleanup_fd_r != -1)
-			close(cleanup_fd_r);
-		if (cleanup_fd_w != -1) {
-			flush_write_file(cleanup_fd_w);
-			close(cleanup_fd_w);
+		if (DEBUG_GTE(EXIT, 2)) {
+			rprintf(FINFO,
+				"[%s] _exit_cleanup(code=%d, file=%s, line=%d): entered\n",
+				who_am_i(), code, file, line);
 		}
-		finish_transfer(cleanup_new_fname, fname, cleanup_file, 0,
-				!partial_dir);
-	}
-	io_flush(FULL_FLUSH);
-	if (cleanup_fname)
-		do_unlink(cleanup_fname);
-	if (code)
-		kill_all(SIGUSR1);
-	if (cleanup_pid && cleanup_pid == getpid()) {
-		char *pidf = lp_pid_file();
-		if (pidf && *pidf)
-			unlink(lp_pid_file());
+
+#include "case_N.h"
+		switch_step++;
+
+		if (cleanup_child_pid != -1) {
+			int status;
+			int pid = wait_process(cleanup_child_pid, &status, WNOHANG);
+			if (pid == cleanup_child_pid) {
+				status = WEXITSTATUS(status);
+				if (status > exit_code)
+					exit_code = status;
+			}
+		}
+
+#include "case_N.h"
+		switch_step++;
+
+		if (cleanup_got_literal && (cleanup_fname || cleanup_fd_w != -1)) {
+			if (cleanup_fd_r != -1) {
+				close(cleanup_fd_r);
+				cleanup_fd_r = -1;
+			}
+			if (cleanup_fd_w != -1) {
+				flush_write_file(cleanup_fd_w);
+				close(cleanup_fd_w);
+				cleanup_fd_w = -1;
+			}
+			if (cleanup_fname && cleanup_new_fname && keep_partial
+			 && handle_partial_dir(cleanup_new_fname, PDIR_CREATE)) {
+				int tweak_modtime = 0;
+				const char *fname = cleanup_fname;
+				cleanup_fname = NULL;
+				if (!partial_dir) {
+					/* We don't want to leave a partial file with a modern time or it
+					 * could be skipped via --update.  Setting the time to something
+					 * really old also helps it to stand out as unfinished in an ls. */
+					tweak_modtime = 1;
+					cleanup_file->modtime = 0;
+				}
+				finish_transfer(cleanup_new_fname, fname, NULL, NULL,
+						cleanup_file, tweak_modtime, !partial_dir);
+			}
+		}
+
+#include "case_N.h"
+		switch_step++;
+
+		if (flush_ok_after_signal) {
+			flush_ok_after_signal = False;
+			if (code == RERR_SIGNAL)
+				io_flush(FULL_FLUSH);
+		}
+		if (!exit_code && !code)
+			io_flush(FULL_FLUSH);
+
+#include "case_N.h"
+		switch_step++;
+
+		if (cleanup_fname)
+			do_unlink(cleanup_fname);
+		if (exit_code)
+			kill_all(SIGUSR1);
+		if (cleanup_pid && cleanup_pid == getpid()) {
+			char *pidf = lp_pid_file();
+			if (pidf && *pidf)
+				unlink(lp_pid_file());
+		}
+
+		if (exit_code == 0) {
+			if (code)
+				exit_code = code;
+			if (io_error & IOERR_DEL_LIMIT)
+				exit_code = RERR_DEL_LIMIT;
+			if (io_error & IOERR_VANISHED)
+				exit_code = RERR_VANISHED;
+			if (io_error & IOERR_GENERAL || got_xfer_error)
+				exit_code = RERR_PARTIAL;
+		}
+
+		/* If line < 0, this exit is after a MSG_ERROR_EXIT event, so
+		 * we don't want to output a duplicate error. */
+		if ((exit_code && line > 0)
+		 || am_daemon || (logfile_name && (am_server || !INFO_GTE(STATS, 1)))) {
+#ifdef HAVE_USLEEP /* A tiny delay just in case both sender & receiver are sending a msg at the same time. */
+			if (am_server && exit_code)
+				usleep(50);
+#endif
+			log_exit(exit_code, exit_file, exit_line);
+		}
+
+#include "case_N.h"
+		switch_step++;
+
+		if (DEBUG_GTE(EXIT, 1)) {
+			rprintf(FINFO,
+				"[%s] _exit_cleanup(code=%d, file=%s, line=%d): "
+				"about to call exit(%d)%s\n",
+				who_am_i(), first_code, exit_file, exit_line, exit_code,
+				dry_run ? " (DRY RUN)" : "");
+		}
+
+#include "case_N.h"
+		switch_step++;
+
+		if (exit_code && exit_code != RERR_SOCKETIO && exit_code != RERR_STREAMIO && exit_code != RERR_SIGNAL1
+		 && exit_code != RERR_TIMEOUT && !shutting_down) {
+			if (protocol_version >= 31 || am_receiver) {
+				if (line > 0) {
+					if (DEBUG_GTE(EXIT, 3)) {
+						rprintf(FINFO, "[%s] sending MSG_ERROR_EXIT with exit_code %d\n",
+							who_am_i(), exit_code);
+					}
+					send_msg_int(MSG_ERROR_EXIT, exit_code);
+				}
+				if (!am_sender)
+					io_flush(MSG_FLUSH); /* Be sure to send all messages */
+				noop_io_until_death();
+			}
+			else if (!am_sender)
+				io_flush(MSG_FLUSH); /* Be sure to send all messages */
+		}
+
+#include "case_N.h"
+		switch_step++;
+
+		if (am_server && exit_code)
+			msleep(100);
+		close_all();
+
+		/* FALLTHROUGH */
+	default:
+		break;
 	}
 
-	if (code == 0) {
-		if (io_error & IOERR_DEL_LIMIT)
-			code = RERR_DEL_LIMIT;
-		if (io_error & IOERR_VANISHED)
-			code = RERR_VANISHED;
-		if (io_error & IOERR_GENERAL || log_got_error)
-			code = RERR_PARTIAL;
-	}
-
-	if (code)
-		log_exit(code, file, line);
-
-	if (verbose > 2) {
-		rprintf(FINFO,"_exit_cleanup(code=%d, file=%s, line=%d): about to call exit(%d)\n",
-			ocode, safe_fname(file), line, code);
-	}
-
-	close_all();
-	exit(code);
+	if (called_from_signal_handler)
+		_exit(exit_code);
+	exit(exit_code);
 }
 
 void cleanup_disable(void)
 {
-	cleanup_fname = NULL;
+	cleanup_fname = cleanup_new_fname = NULL;
+	cleanup_fd_r = cleanup_fd_w = -1;
 	cleanup_got_literal = 0;
 }
 
 
-void cleanup_set(char *fnametmp, char *fname, struct file_struct *file,
+void cleanup_set(const char *fnametmp, const char *fname, struct file_struct *file,
 		 int fd_r, int fd_w)
 {
 	cleanup_fname = fnametmp;
-	cleanup_new_fname = fname;
+	cleanup_new_fname = fname; /* can be NULL on a partial-dir failure */
 	cleanup_file = file;
 	cleanup_fd_r = fd_r;
 	cleanup_fd_w = fd_w;

clientname.c

@@ -1,82 +1,86 @@
-/* -*- c-file-style: "linux" -*-
-
-   rsync -- fast file replication program
-
-   Copyright (C) 1992-2001 by Andrew Tridgell <tridge@samba.org>
-   Copyright (C) 2001, 2002 by Martin Pool <mbp@samba.org>
-
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
-
-/**
- * @file clientname.c
- *
+/*
  * Functions for looking up the remote name or addr of a socket.
  *
+ * Copyright (C) 1992-2001 Andrew Tridgell <tridge@samba.org>
+ * Copyright (C) 2001, 2002 Martin Pool <mbp@samba.org>
+ * Copyright (C) 2002-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
+
+/*
  * This file is now converted to use the new-style getaddrinfo()
  * interface, which supports IPv6 but is also supported on recent
  * IPv4-only machines.  On systems that don't have that interface, we
  * emulate it using the KAME implementation.
- **/
+ */
 
 #include "rsync.h"
+#include "itypes.h"
+
+extern int am_daemon;
 
 static const char default_name[] = "UNKNOWN";
-extern int am_daemon;
-extern int am_server;
+static const char proxyv2sig[] = "\r\n\r\n\0\r\nQUIT\n";
 
+static char ipaddr_buf[100];
 
-/**
- * Return the IP addr of the client as a string
- **/
+#define PROXY_V2_SIG_SIZE ((int)sizeof proxyv2sig - 1)
+#define PROXY_V2_HEADER_SIZE (PROXY_V2_SIG_SIZE + 1 + 1 + 2)
+
+#define CMD_LOCAL 0
+#define CMD_PROXY 1
+
+#define PROXY_FAM_TCPv4 0x11
+#define PROXY_FAM_TCPv6 0x21
+
+#define GET_SOCKADDR_FAMILY(ss) ((struct sockaddr*)ss)->sa_family
+
+static void client_sockaddr(int fd, struct sockaddr_storage *ss, socklen_t *ss_len);
+static int check_name(const char *ipaddr, const struct sockaddr_storage *ss, char *name_buf, size_t name_buf_size);
+static int valid_ipaddr(const char *s);
+
+/* Return the IP addr of the client as a string. */
 char *client_addr(int fd)
 {
-	static char addr_buf[100];
-	static int initialised;
 	struct sockaddr_storage ss;
 	socklen_t length = sizeof ss;
-	char *ssh_info, *p;
 
-	if (initialised)
-		return addr_buf;
+	if (*ipaddr_buf)
+		return ipaddr_buf;
 
-	initialised = 1;
-
-	if (am_server) {	/* daemon over --rsh mode */
-		strcpy(addr_buf, "0.0.0.0");
-		if ((ssh_info = getenv("SSH_CONNECTION")) != NULL
-		    || (ssh_info = getenv("SSH_CLIENT")) != NULL
-		    || (ssh_info = getenv("SSH2_CLIENT")) != NULL) {
-			strlcpy(addr_buf, ssh_info, sizeof addr_buf);
+	if (am_daemon < 0) {	/* daemon over --rsh mode */
+		char *env_str;
+		strlcpy(ipaddr_buf, "0.0.0.0", sizeof ipaddr_buf);
+		if ((env_str = getenv("REMOTE_HOST")) != NULL
+		 || (env_str = getenv("SSH_CONNECTION")) != NULL
+		 || (env_str = getenv("SSH_CLIENT")) != NULL
+		 || (env_str = getenv("SSH2_CLIENT")) != NULL) {
+			char *p;
+			strlcpy(ipaddr_buf, env_str, sizeof ipaddr_buf);
 			/* Truncate the value to just the IP address. */
-			if ((p = strchr(addr_buf, ' ')) != NULL)
+			if ((p = strchr(ipaddr_buf, ' ')) != NULL)
 				*p = '\0';
 		}
-	} else {
-		client_sockaddr(fd, &ss, &length);
-		getnameinfo((struct sockaddr *)&ss, length,
-			    addr_buf, sizeof addr_buf, NULL, 0, NI_NUMERICHOST);
+		if (valid_ipaddr(ipaddr_buf))
+			return ipaddr_buf;
 	}
 
-	return addr_buf;
-}
+	client_sockaddr(fd, &ss, &length);
+	getnameinfo((struct sockaddr *)&ss, length, ipaddr_buf, sizeof ipaddr_buf, NULL, 0, NI_NUMERICHOST);
 
-
-static int get_sockaddr_family(const struct sockaddr_storage *ss)
-{
-	return ((struct sockaddr *) ss)->sa_family;
+	return ipaddr_buf;
 }
 
 
@@ -93,66 +97,216 @@ static int get_sockaddr_family(const struct sockaddr_storage *ss)
  * After translation from sockaddr to name we do a forward lookup to
  * make sure nobody is spoofing PTR records.
  **/
-char *client_name(int fd)
+char *client_name(const char *ipaddr)
 {
 	static char name_buf[100];
-	static char port_buf[100];
-	static int initialised;
+	char port_buf[100];
 	struct sockaddr_storage ss;
 	socklen_t ss_len;
+	struct addrinfo hint, *answer;
+	int err;
 
-	if (initialised)
+	if (*name_buf)
 		return name_buf;
 
-	strcpy(name_buf, default_name);
-	initialised = 1;
+	strlcpy(name_buf, default_name, sizeof name_buf);
+
+	if (strcmp(ipaddr, "0.0.0.0") == 0)
+		return name_buf;
 
 	memset(&ss, 0, sizeof ss);
-
-	if (am_server) {	/* daemon over --rsh mode */
-		char *addr = client_addr(fd);
-		struct addrinfo hint, *answer;
-		int err;
-
-		memset(&hint, 0, sizeof hint);
+	memset(&hint, 0, sizeof hint);
 
 #ifdef AI_NUMERICHOST
-		hint.ai_flags = AI_NUMERICHOST;
+	hint.ai_flags = AI_NUMERICHOST;
 #endif
-		hint.ai_socktype = SOCK_STREAM;
+	hint.ai_socktype = SOCK_STREAM;
 
-		if ((err = getaddrinfo(addr, NULL, &hint, &answer)) != 0) {
-			rprintf(FLOG, "malformed address %s: %s\n",
-			        addr, gai_strerror(err));
-			return name_buf;
-		}
-
-		switch (answer->ai_family) {
-		case AF_INET:
-			ss_len = sizeof (struct sockaddr_in);
-			memcpy(&ss, answer->ai_addr, ss_len);
-			break;
-#ifdef INET6
-		case AF_INET6:
-			ss_len = sizeof (struct sockaddr_in6);
-			memcpy(&ss, answer->ai_addr, ss_len);
-			break;
-#endif
-		}
-		freeaddrinfo(answer);
-	} else {
-		ss_len = sizeof ss;
-		client_sockaddr(fd, &ss, &ss_len);
+	if ((err = getaddrinfo(ipaddr, NULL, &hint, &answer)) != 0) {
+		rprintf(FLOG, "malformed address %s: %s\n", ipaddr, gai_strerror(err));
+		return name_buf;
 	}
 
-	if (lookup_name(fd, &ss, ss_len, name_buf, sizeof name_buf,
-			port_buf, sizeof port_buf) == 0)
-		check_name(fd, &ss, name_buf);
+	switch (answer->ai_family) {
+	case AF_INET:
+		ss_len = sizeof (struct sockaddr_in);
+		memcpy(&ss, answer->ai_addr, ss_len);
+		break;
+#ifdef INET6
+	case AF_INET6:
+		ss_len = sizeof (struct sockaddr_in6);
+		memcpy(&ss, answer->ai_addr, ss_len);
+		break;
+#endif
+	default:
+		assert(0);
+	}
+	freeaddrinfo(answer);
+
+	/* reverse lookup */
+	err = getnameinfo((struct sockaddr*)&ss, ss_len, name_buf, sizeof name_buf,
+			  port_buf, sizeof port_buf, NI_NAMEREQD | NI_NUMERICSERV);
+	if (err) {
+		strlcpy(name_buf, default_name, sizeof name_buf);
+		rprintf(FLOG, "name lookup failed for %s: %s\n", ipaddr, gai_strerror(err));
+	} else
+		check_name(ipaddr, &ss, name_buf, sizeof name_buf);
 
 	return name_buf;
 }
 
 
+/* Try to read an proxy protocol header (V1 or V2). Returns 1 on success or 0 on failure. */
+int read_proxy_protocol_header(int fd)
+{
+	union {
+		struct {
+			char line[108];
+		} v1;
+		struct {
+			char sig[PROXY_V2_SIG_SIZE];
+			char ver_cmd;
+			char fam;
+			char len[2];
+			union {
+				struct {
+					char src_addr[4];
+					char dst_addr[4];
+					char src_port[2];
+					char dst_port[2];
+				} ip4;
+				struct {
+					char src_addr[16];
+					char dst_addr[16];
+					char src_port[2];
+					char dst_port[2];
+				} ip6;
+				struct {
+					char src_addr[108];
+					char dst_addr[108];
+				} unx;
+			} addr;
+		} v2;
+	} hdr;
+
+	read_buf(fd, (char*)&hdr, PROXY_V2_SIG_SIZE);
+
+	if (memcmp(hdr.v2.sig, proxyv2sig, PROXY_V2_SIG_SIZE) == 0) { /* Proxy V2 */
+		int ver, cmd, size;
+
+		read_buf(fd, (char*)&hdr + PROXY_V2_SIG_SIZE, PROXY_V2_HEADER_SIZE - PROXY_V2_SIG_SIZE);
+
+		ver = (hdr.v2.ver_cmd & 0xf0) >> 4;
+		cmd = (hdr.v2.ver_cmd & 0x0f);
+		size = (hdr.v2.len[0] << 8) + hdr.v2.len[1];
+
+		if (ver != 2 || size + PROXY_V2_HEADER_SIZE > (int)sizeof hdr)
+			return 0;
+
+		/* Grab all the remaining data in the binary request. */
+		read_buf(fd, (char*)&hdr + PROXY_V2_HEADER_SIZE, size);
+
+		switch (cmd) {
+		case CMD_PROXY:
+			switch (hdr.v2.fam) {
+			case PROXY_FAM_TCPv4:
+				if (size != sizeof hdr.v2.addr.ip4)
+					return 0;
+				inet_ntop(AF_INET, hdr.v2.addr.ip4.src_addr, ipaddr_buf, sizeof ipaddr_buf);
+				return valid_ipaddr(ipaddr_buf);
+			case PROXY_FAM_TCPv6:
+				if (size != sizeof hdr.v2.addr.ip6)
+					return 0;
+				inet_ntop(AF_INET6, hdr.v2.addr.ip6.src_addr, ipaddr_buf, sizeof ipaddr_buf);
+				return valid_ipaddr(ipaddr_buf);
+			default:
+				break;
+			}
+			/* For an unsupported protocol we'll ignore the proxy data (leaving ipaddr_buf unset)
+			 * and accept the connection, which will get handled as a normal socket addr. */
+			return 1;
+		case CMD_LOCAL:
+			return 1;
+		default:
+			break;
+		}
+
+		return 0;
+	}
+
+	if (memcmp(hdr.v1.line, "PROXY", 5) == 0) { /* Proxy V1 */
+		char *endc, *sp, *p = hdr.v1.line + PROXY_V2_SIG_SIZE;
+		int port_chk;
+
+		*p = '\0';
+		if (!strchr(hdr.v1.line, '\n')) {
+			while (1) {
+				read_buf(fd, p, 1);
+				if (*p++ == '\n')
+					break;
+				if (p - hdr.v1.line >= (int)sizeof hdr.v1.line - 1)
+					return 0;
+			}
+			*p = '\0';
+		}
+
+		endc = strchr(hdr.v1.line, '\r');
+		if (!endc || endc[1] != '\n' || endc[2])
+			return 0;
+		*endc = '\0';
+
+		p = hdr.v1.line + 5;
+
+		if (!isSpace(p++))
+			return 0;
+		if (strncmp(p, "TCP4", 4) == 0)
+			p += 4;
+		else if (strncmp(p, "TCP6", 4) == 0)
+			p += 4;
+		else if (strncmp(p, "UNKNOWN", 7) == 0)
+			return 1;
+		else
+			return 0;
+
+		if (!isSpace(p++))
+			return 0;
+
+		if ((sp = strchr(p, ' ')) == NULL)
+			return 0;
+		*sp = '\0';
+		if (!valid_ipaddr(p))
+			return 0;
+		strlcpy(ipaddr_buf, p, sizeof ipaddr_buf); /* It will always fit when valid. */
+
+		p = sp + 1;
+		if ((sp = strchr(p, ' ')) == NULL)
+			return 0;
+		*sp = '\0';
+		if (!valid_ipaddr(p))
+			return 0;
+		/* Ignore destination address. */
+
+		p = sp + 1;
+		if ((sp = strchr(p, ' ')) == NULL)
+			return 0;
+		*sp = '\0';
+		port_chk = strtol(p, &endc, 10);
+		if (*endc || port_chk == 0)
+			return 0;
+		/* Ignore source port. */
+
+		p = sp + 1;
+		port_chk = strtol(p, &endc, 10);
+		if (*endc || port_chk == 0)
+			return 0;
+		/* Ignore destination port. */
+
+		return 1;
+	}
+
+	return 0;
+}
+
 
 /**
  * Get the sockaddr for the client.
@@ -160,9 +314,7 @@ char *client_name(int fd)
  * If it comes in as an ipv4 address mapped into IPv6 format then we
  * convert it back to a regular IPv4.
  **/
-void client_sockaddr(int fd,
-		     struct sockaddr_storage *ss,
-		     socklen_t *ss_len)
+static void client_sockaddr(int fd, struct sockaddr_storage *ss, socklen_t *ss_len)
 {
 	memset(ss, 0, sizeof *ss);
 
@@ -173,8 +325,8 @@ void client_sockaddr(int fd,
 	}
 
 #ifdef INET6
-	if (get_sockaddr_family(ss) == AF_INET6 &&
-	    IN6_IS_ADDR_V4MAPPED(&((struct sockaddr_in6 *)ss)->sin6_addr)) {
+	if (GET_SOCKADDR_FAMILY(ss) == AF_INET6
+	 && IN6_IS_ADDR_V4MAPPED(&((struct sockaddr_in6 *)ss)->sin6_addr)) {
 		/* OK, so ss is in the IPv6 family, but it is really
 		 * an IPv4 address: something like
 		 * "::ffff:10.130.1.2".  If we use it as-is, then the
@@ -197,51 +349,20 @@ void client_sockaddr(int fd,
 		/* There is a macro to extract the mapped part
 		 * (IN6_V4MAPPED_TO_SINADDR ?), but it does not seem
 		 * to be present in the Linux headers. */
-		memcpy(&sin->sin_addr, &sin6.sin6_addr.s6_addr[12],
-		    sizeof sin->sin_addr);
+		memcpy(&sin->sin_addr, &sin6.sin6_addr.s6_addr[12], sizeof sin->sin_addr);
 	}
 #endif
 }
 
 
-/**
- * Look up a name from @p ss into @p name_buf.
- *
- * @param fd file descriptor for client socket.
- **/
-int lookup_name(int fd, const struct sockaddr_storage *ss,
-		socklen_t ss_len,
-		char *name_buf, size_t name_buf_len,
-		char *port_buf, size_t port_buf_len)
-{
-	int name_err;
-
-	/* reverse lookup */
-	name_err = getnameinfo((struct sockaddr *) ss, ss_len,
-			       name_buf, name_buf_len,
-			       port_buf, port_buf_len,
-			       NI_NAMEREQD | NI_NUMERICSERV);
-	if (name_err != 0) {
-		strcpy(name_buf, default_name);
-		rprintf(FLOG, "name lookup failed for %s: %s\n",
-			client_addr(fd), gai_strerror(name_err));
-		return name_err;
-	}
-
-	return 0;
-}
-
-
-
 /**
  * Compare an addrinfo from the resolver to a sockinfo.
  *
  * Like strcmp, returns 0 for identical.
  **/
-int compare_addrinfo_sockaddr(const struct addrinfo *ai,
-			      const struct sockaddr_storage *ss)
+static int compare_addrinfo_sockaddr(const struct addrinfo *ai, const struct sockaddr_storage *ss)
 {
-	int ss_family = get_sockaddr_family(ss);
+	int ss_family = GET_SOCKADDR_FAMILY(ss);
 	const char fn[] = "compare_addrinfo_sockaddr";
 
 	if (ai->ai_family != ss_family) {
@@ -257,8 +378,7 @@ int compare_addrinfo_sockaddr(const struct addrinfo *ai,
 		sin1 = (const struct sockaddr_in *) ss;
 		sin2 = (const struct sockaddr_in *) ai->ai_addr;
 
-		return memcmp(&sin1->sin_addr, &sin2->sin_addr,
-			      sizeof sin1->sin_addr);
+		return memcmp(&sin1->sin_addr, &sin2->sin_addr, sizeof sin1->sin_addr);
 	}
 
 #ifdef INET6
@@ -268,14 +388,13 @@ int compare_addrinfo_sockaddr(const struct addrinfo *ai,
 		sin1 = (const struct sockaddr_in6 *) ss;
 		sin2 = (const struct sockaddr_in6 *) ai->ai_addr;
 
-		if (ai->ai_addrlen < sizeof (struct sockaddr_in6)) {
+		if (ai->ai_addrlen < (int)sizeof (struct sockaddr_in6)) {
 			rprintf(FLOG, "%s: too short sockaddr_in6; length=%d\n",
-				fn, ai->ai_addrlen);
+				fn, (int)ai->ai_addrlen);
 			return 1;
 		}
 
-		if (memcmp(&sin1->sin6_addr, &sin2->sin6_addr,
-			   sizeof sin1->sin6_addr))
+		if (memcmp(&sin1->sin6_addr, &sin2->sin6_addr, sizeof sin1->sin6_addr))
 			return 1;
 
 #ifdef HAVE_SOCKADDR_IN6_SCOPE_ID
@@ -301,13 +420,11 @@ int compare_addrinfo_sockaddr(const struct addrinfo *ai,
  * because it doesn't seem that it could be spoofed in any way, and
  * getaddrinfo on random service names seems to cause problems on AIX.
  **/
-int check_name(int fd,
-	       const struct sockaddr_storage *ss,
-	       char *name_buf)
+static int check_name(const char *ipaddr, const struct sockaddr_storage *ss, char *name_buf, size_t name_buf_size)
 {
 	struct addrinfo hints, *res, *res0;
 	int error;
-	int ss_family = get_sockaddr_family(ss);
+	int ss_family = GET_SOCKADDR_FAMILY(ss);
 
 	memset(&hints, 0, sizeof hints);
 	hints.ai_family = ss_family;
@@ -317,11 +434,10 @@ int check_name(int fd,
 	if (error) {
 		rprintf(FLOG, "forward name lookup for %s failed: %s\n",
 			name_buf, gai_strerror(error));
-		strcpy(name_buf, default_name);
+		strlcpy(name_buf, default_name, name_buf_size);
 		return error;
 	}
 
-
 	/* Given all these results, we expect that one of them will be
 	 * the same as ss.  The comparison is a bit complicated. */
 	for (res = res0; res; res = res->ai_next) {
@@ -334,15 +450,79 @@ int check_name(int fd,
 		 * address that was the same as ss. */
 		rprintf(FLOG, "no known address for \"%s\": "
 			"spoofed address?\n", name_buf);
-		strcpy(name_buf, default_name);
+		strlcpy(name_buf, default_name, name_buf_size);
 	} else if (res == NULL) {
 		/* We hit the end of the list without finding an
 		 * address that was the same as ss. */
 		rprintf(FLOG, "%s is not a known address for \"%s\": "
-			"spoofed address?\n", client_addr(fd), name_buf);
-		strcpy(name_buf, default_name);
+			"spoofed address?\n", ipaddr, name_buf);
+		strlcpy(name_buf, default_name, name_buf_size);
 	}
 
 	freeaddrinfo(res0);
 	return 0;
 }
+
+/* Returns 1 for a valid IPv4 or IPv6 addr, or 0 for a bad one. */
+static int valid_ipaddr(const char *s)
+{
+	int i;
+
+	if (strchr(s, ':') != NULL) { /* Only IPv6 has a colon. */
+		int count, saw_double_colon = 0;
+		int ipv4_at_end = 0;
+
+		if (*s == ':') { /* A colon at the start must be a :: */
+			if (*++s != ':')
+				return 0;
+			saw_double_colon = 1;
+			s++;
+		}
+
+		for (count = 0; count < 8; count++) {
+			if (!*s)
+				return saw_double_colon;
+
+			if (strchr(s, ':') == NULL && strchr(s, '.') != NULL) {
+				if ((!saw_double_colon && count != 6) || (saw_double_colon && count > 6))
+					return 0;
+				ipv4_at_end = 1;
+				break;
+			}
+
+			if (!isHexDigit(s++)) /* Need 1-4 hex digits */
+				return 0;
+			if (isHexDigit(s) && isHexDigit(++s) && isHexDigit(++s) && isHexDigit(++s))
+				return 0;
+
+			if (*s == ':') {
+				if (!*++s)
+					return 0;
+				if (*s == ':') {
+					if (saw_double_colon)
+						return 0;
+					saw_double_colon = 1;
+					s++;
+				}
+			}
+		}
+
+		if (!ipv4_at_end)
+			return !*s;
+	}
+
+	/* IPv4 */
+	for (i = 0; i < 4; i++) {
+		long n;
+		char *end;
+
+		if (i && *s++ != '.')
+			return 0;
+		n = strtol(s, &end, 10);
+		if (n > 255 || n < 0 || end <= s || end > s+3)
+			return 0;
+		s = end;
+	}
+
+	return !*s;
+}

cmdormsg

@@ -0,0 +1,11 @@
+#!/bin/sh
+
+srcdir=`dirname $0`
+opt="$1"
+shift
+
+echo "$*"
+if ! "${@}"; then
+    echo "If you can't fix the issue, re-run $srcdir/configure with --$opt."
+    exit 1
+fi

compat.c

@@ -1,46 +1,577 @@
-/* 
-   Copyright (C) Andrew Tridgell 1996
-   Copyright (C) Paul Mackerras 1996
-   
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
-
-/**
- * @file compat.c
- *
+/*
  * Compatibility routines for older rsync protocol versions.
- **/
+ *
+ * Copyright (C) Andrew Tridgell 1996
+ * Copyright (C) Paul Mackerras 1996
+ * Copyright (C) 2004-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
 #include "rsync.h"
+#include "itypes.h"
 
-int remote_protocol = 0;
-
-extern int verbose;
 extern int am_server;
 extern int am_sender;
+extern int local_server;
 extern int inplace;
+extern int recurse;
+extern int use_qsort;
+extern int allow_inc_recurse;
+extern int preallocate_files;
+extern int append_mode;
 extern int fuzzy_basis;
 extern int read_batch;
+extern int write_batch;
+extern int delay_updates;
 extern int checksum_seed;
 extern int basis_dir_cnt;
+extern int prune_empty_dirs;
 extern int protocol_version;
-extern char *dest_option;
+extern int protect_args;
+extern int preserve_uid;
+extern int preserve_gid;
+extern int preserve_atimes;
+extern int preserve_acls;
+extern int preserve_xattrs;
+extern int xfer_flags_as_varint;
+extern int need_messages_from_generator;
+extern int delete_mode, delete_before, delete_during, delete_after;
+extern int do_compression;
+extern int do_compression_level;
+extern char *shell_cmd;
+extern char *partial_dir;
+extern char *files_from;
+extern char *filesfrom_host;
+extern const char *checksum_choice;
+extern const char *compress_choice;
+extern filter_rule_list filter_list;
+extern int need_unsorted_flist;
+#ifdef ICONV_OPTION
+extern iconv_t ic_send, ic_recv;
+extern char *iconv_opt;
+#endif
+extern struct name_num_obj valid_checksums;
+
+int remote_protocol = 0;
+int file_extra_cnt = 0; /* count of file-list extras that everyone gets */
+int inc_recurse = 0;
+int compat_flags = 0;
+int use_safe_inc_flist = 0;
+int want_xattr_optim = 0;
+int proper_seed_order = 0;
+int inplace_partial = 0;
+int do_negotiated_strings = 0;
+
+/* These index values are for the file-list's extra-attribute array. */
+int pathname_ndx, depth_ndx, atimes_ndx, uid_ndx, gid_ndx, acls_ndx, xattrs_ndx, unsort_ndx;
+
+int receiver_symlink_times = 0; /* receiver can set the time on a symlink */
+int sender_symlink_iconv = 0;	/* sender should convert symlink content */
+
+#ifdef ICONV_OPTION
+int filesfrom_convert = 0;
+#endif
+
+#define MAX_NSTR_STRLEN 256
+
+struct name_num_obj valid_compressions = {
+	"compress", NULL, NULL, 0, 0, {
+#ifdef SUPPORT_ZSTD
+		{ CPRES_ZSTD, "zstd", NULL },
+#endif
+#ifdef SUPPORT_LZ4
+		{ CPRES_LZ4, "lz4", NULL },
+#endif
+		{ CPRES_ZLIBX, "zlibx", NULL },
+		{ CPRES_ZLIB, "zlib", NULL },
+		{ CPRES_NONE, "none", NULL },
+		{ 0, NULL, NULL }
+	}
+};
+
+#define CF_INC_RECURSE	 (1<<0)
+#define CF_SYMLINK_TIMES (1<<1)
+#define CF_SYMLINK_ICONV (1<<2)
+#define CF_SAFE_FLIST	 (1<<3)
+#define CF_AVOID_XATTR_OPTIM (1<<4)
+#define CF_CHKSUM_SEED_FIX (1<<5)
+#define CF_INPLACE_PARTIAL_DIR (1<<6)
+#define CF_VARINT_FLIST_FLAGS (1<<7)
+
+static const char *client_info;
+
+/* The server makes sure that if either side only supports a pre-release
+ * version of a protocol, that both sides must speak a compatible version
+ * of that protocol for it to be advertised as available. */
+static void check_sub_protocol(void)
+{
+	char *dot;
+	int their_protocol, their_sub;
+#if SUBPROTOCOL_VERSION != 0
+	int our_sub = protocol_version < PROTOCOL_VERSION ? 0 : SUBPROTOCOL_VERSION;
+#else
+	int our_sub = 0;
+#endif
+
+	/* client_info starts with VER.SUB string if client is a pre-release. */
+	if (!(their_protocol = atoi(client_info))
+	 || !(dot = strchr(client_info, '.'))
+	 || !(their_sub = atoi(dot+1))) {
+#if SUBPROTOCOL_VERSION != 0
+		if (our_sub)
+			protocol_version--;
+#endif
+		return;
+	}
+
+	if (their_protocol < protocol_version) {
+		if (their_sub)
+			protocol_version = their_protocol - 1;
+		return;
+	}
+
+	if (their_protocol > protocol_version)
+		their_sub = 0; /* 0 == final version of older protocol */
+	if (their_sub != our_sub)
+		protocol_version--;
+}
+
+void set_allow_inc_recurse(void)
+{
+	client_info = shell_cmd ? shell_cmd : "";
+
+	if (!recurse || use_qsort)
+		allow_inc_recurse = 0;
+	else if (!am_sender
+	 && (delete_before || delete_after
+	  || delay_updates || prune_empty_dirs))
+		allow_inc_recurse = 0;
+	else if (am_server && !local_server
+	 && (strchr(client_info, 'i') == NULL))
+		allow_inc_recurse = 0;
+}
+
+void parse_compress_choice(int final_call)
+{
+	if (valid_compressions.negotiated_name)
+		do_compression = valid_compressions.negotiated_num;
+	else if (compress_choice) {
+		struct name_num_item *nni = get_nni_by_name(&valid_compressions, compress_choice, -1);
+		if (!nni) {
+			rprintf(FERROR, "unknown compress name: %s\n", compress_choice);
+			exit_cleanup(RERR_UNSUPPORTED);
+		}
+		do_compression = nni->num;
+		if (am_server)
+			validate_choice_vs_env(NSTR_COMPRESS, do_compression, -1);
+	} else if (do_compression)
+		do_compression = CPRES_ZLIB;
+	else
+		do_compression = CPRES_NONE;
+
+	if (do_compression != CPRES_NONE && final_call)
+		init_compression_level(); /* There's a chance this might turn compression off! */
+
+	if (do_compression == CPRES_NONE)
+		compress_choice = NULL;
+
+	/* Snag the compression name for both write_batch's option output & the following debug output. */
+	if (valid_compressions.negotiated_name)
+		compress_choice = valid_compressions.negotiated_name;
+	else if (compress_choice == NULL) {
+		struct name_num_item *nni = get_nni_by_num(&valid_compressions, do_compression);
+		compress_choice = nni ? nni->name : "UNKNOWN";
+	}
+
+	if (final_call && DEBUG_GTE(NSTR, am_server ? 3 : 1)
+	 && (do_compression != CPRES_NONE || do_compression_level != CLVL_NOT_SPECIFIED)) {
+		rprintf(FINFO, "%s%s compress: %s (level %d)\n",
+			am_server ? "Server" : "Client",
+			valid_compressions.negotiated_name ? " negotiated" : "",
+			compress_choice, do_compression_level);
+	}
+}
+
+struct name_num_item *get_nni_by_name(struct name_num_obj *nno, const char *name, int len)
+{
+	struct name_num_item *nni;
+
+	if (len < 0)
+		len = strlen(name);
+
+	for (nni = nno->list; nni->name; nni++) {
+		if (strncasecmp(name, nni->name, len) == 0 && nni->name[len] == '\0')
+			return nni;
+	}
+
+	return NULL;
+}
+
+struct name_num_item *get_nni_by_num(struct name_num_obj *nno, int num)
+{
+	struct name_num_item *nni;
+
+	for (nni = nno->list; nni->name; nni++) {
+		if (num == nni->num)
+			return nni;
+	}
+
+	return NULL;
+}
+
+static void init_nno_saw(struct name_num_obj *nno, int val)
+{
+	struct name_num_item *nni;
+	int cnt;
+
+	if (!nno->saw_len) {
+		for (nni = nno->list; nni->name; nni++) {
+			if (nni->num >= nno->saw_len)
+				nno->saw_len = nni->num + 1;
+		}
+	}
+
+	if (!nno->saw) {
+		nno->saw = new_array0(uchar, nno->saw_len);
+
+		/* We'll take this opportunity to make sure that the main_name values are set right. */
+		for (cnt = 1, nni = nno->list; nni->name; nni++, cnt++) {
+			if (nno->saw[nni->num])
+				nni->main_name = nno->list[nno->saw[nni->num]-1].name;
+			else
+				nno->saw[nni->num] = cnt;
+		}
+	}
+
+	memset(nno->saw, val, nno->saw_len);
+}
+
+/* Simplify the user-provided string so that it contains valid names without any duplicates.
+ * It also sets the "saw" flags to a 1-relative count of which name was seen first. */
+static int parse_nni_str(struct name_num_obj *nno, const char *from, char *tobuf, int tobuf_len)
+{
+	char *to = tobuf, *tok = NULL;
+	int saw_tok = 0, cnt = 0;
+
+	while (1) {
+		int at_space = isSpace(from);
+		char ch = *from++;
+		if (ch == '&')
+			ch = '\0';
+		if (!ch || at_space) {
+			if (tok) {
+				struct name_num_item *nni = get_nni_by_name(nno, tok, to - tok);
+				if (nni && !nno->saw[nni->num]) {
+					nno->saw[nni->num] = ++cnt;
+					if (nni->main_name) {
+						to = tok + strlcpy(tok, nni->main_name, tobuf_len - (tok - tobuf));
+						if (to - tobuf >= tobuf_len) {
+							to = tok - 1;
+							break;
+						}
+					}
+				} else
+					to = tok - (tok != tobuf);
+				saw_tok = 1;
+				tok = NULL;
+			}
+			if (!ch)
+				break;
+			continue;
+		}
+		if (!tok) {
+			if (to != tobuf)
+				*to++ = ' ';
+			tok = to;
+		}
+		if (to - tobuf >= tobuf_len - 1) {
+			to = tok - (tok != tobuf);
+			break;
+		}
+		*to++ = ch;
+	}
+	*to = '\0';
+
+	if (saw_tok && to == tobuf)
+		return strlcpy(tobuf, "INVALID", MAX_NSTR_STRLEN);
+
+	return to - tobuf;
+}
+
+/* This routine is always called with a tmpbuf of MAX_NSTR_STRLEN length, but the
+ * buffer may be pre-populated with a "len" length string to use OR a len of -1
+ * to tell us to read a string from the fd. */
+static void recv_negotiate_str(int f_in, struct name_num_obj *nno, char *tmpbuf, int len)
+{
+	struct name_num_item *ret = NULL;
+
+	if (len < 0)
+		len = read_vstring(f_in, tmpbuf, MAX_NSTR_STRLEN);
+
+	if (DEBUG_GTE(NSTR, am_server ? 3 : 2)) {
+		if (am_server)
+			rprintf(FINFO, "Client %s list (on server): %s\n", nno->type, tmpbuf);
+		else
+			rprintf(FINFO, "Server %s list (on client): %s\n", nno->type, tmpbuf);
+	}
+
+	if (len > 0) {
+		struct name_num_item *nni;
+		int best = nno->saw_len; /* We want best == 1 from the client list, so start with a big number. */
+		char *space, *tok = tmpbuf;
+		while (tok) {
+			while (*tok == ' ') tok++; /* Should be unneeded... */
+			if (!*tok)
+				break;
+			if ((space = strchr(tok, ' ')) != NULL)
+				*space = '\0';
+			nni = get_nni_by_name(nno, tok, -1);
+			if (space) {
+				*space = ' ';
+				tok = space + 1;
+			} else
+				tok = NULL;
+			if (!nni || !nno->saw[nni->num] || best <= nno->saw[nni->num])
+				continue;
+			ret = nni;
+			best = nno->saw[nni->num];
+			if (best == 1 || am_server) /* The server side stops at the first acceptable client choice */
+				break;
+		}
+		if (ret) {
+			free(nno->saw);
+			nno->saw = NULL;
+			nno->negotiated_name = ret->main_name ? ret->main_name : ret->name;
+			nno->negotiated_num = ret->num;
+			return;
+		}
+	}
+
+	if (!am_server || !do_negotiated_strings) {
+		char *cp = tmpbuf;
+		int j;
+		rprintf(FERROR, "Failed to negotiate a %s choice.\n", nno->type);
+		rprintf(FERROR, "%s list: %s\n", am_server ? "Client" : "Server", tmpbuf);
+		/* Recreate our original list from the saw values. This can't overflow our huge
+		 * buffer because we don't have enough valid entries to get anywhere close. */
+		for (j = 1, *cp = '\0'; j <= nno->saw_len; j++) {
+			struct name_num_item *nni;
+			for (nni = nno->list; nni->name; nni++) {
+				if (nno->saw[nni->num] == j) {
+					*cp++ = ' ';
+					cp += strlcpy(cp, nni->name, MAX_NSTR_STRLEN - (cp - tmpbuf));
+					break;
+				}
+			}
+		}
+		if (!*tmpbuf)
+			strlcpy(cp, " INVALID", MAX_NSTR_STRLEN);
+		rprintf(FERROR, "%s list:%s\n", am_server ? "Server" : "Client", tmpbuf);
+	}
+
+	exit_cleanup(RERR_UNSUPPORTED);
+}
+
+static const char *getenv_nstr(int ntype)
+{
+	const char *env_str = getenv(ntype == NSTR_COMPRESS ? "RSYNC_COMPRESS_LIST" : "RSYNC_CHECKSUM_LIST");
+
+	/* When writing a batch file, we always negotiate an old-style choice. */
+	if (write_batch) 
+		env_str = ntype == NSTR_COMPRESS ? "zlib" : protocol_version >= 30 ? "md5" : "md4";
+
+	if (am_server && env_str) {
+		char *cp = strchr(env_str, '&');
+		if (cp)
+			env_str = cp + 1;
+	}
+
+	return env_str;
+}
+
+void validate_choice_vs_env(int ntype, int num1, int num2)
+{
+	struct name_num_obj *nno = ntype == NSTR_COMPRESS ? &valid_compressions : &valid_checksums;
+	const char *list_str = getenv_nstr(ntype);
+	char tmpbuf[MAX_NSTR_STRLEN];
+
+	if (!list_str)
+		return;
+
+	while (isSpace(list_str)) list_str++;
+
+	if (!*list_str)
+		return;
+
+	init_nno_saw(nno, 0);
+	parse_nni_str(nno, list_str, tmpbuf, MAX_NSTR_STRLEN);
+
+	if (ntype == NSTR_CHECKSUM) /* If "md4" is in the env list, all the old MD4 choices are OK too. */
+		nno->saw[CSUM_MD4_ARCHAIC] = nno->saw[CSUM_MD4_BUSTED] = nno->saw[CSUM_MD4_OLD] = nno->saw[CSUM_MD4];
+
+	if (!nno->saw[num1] || (num2 >= 0 && !nno->saw[num2])) {
+		rprintf(FERROR, "Your --%s-choice value (%s) was refused by the server.\n", 
+			ntype == NSTR_COMPRESS ? "compress" : "checksum",
+			ntype == NSTR_COMPRESS ? compress_choice : checksum_choice);
+		exit_cleanup(RERR_UNSUPPORTED);
+	}
+
+	free(nno->saw);
+	nno->saw = NULL;
+}
+
+/* The saw buffer is initialized and used to store ordinal values from 1 to N
+ * for the order of the args in the array.  If dup_markup == '\0', duplicates
+ * are removed otherwise the char is prefixed to the duplicate term and, if it
+ * is an opening paren/bracket/brace, the matching closing char is suffixed.
+ * "none" is removed on the client side unless dup_markup != '\0'. */
+int get_default_nno_list(struct name_num_obj *nno, char *to_buf, int to_buf_len, char dup_markup)
+{
+	struct name_num_item *nni;
+	int len = 0, cnt = 0;
+	char delim = '\0', post_delim;
+
+	switch (dup_markup) {
+	case '(': post_delim = ')'; break;
+	case '[': post_delim = ']'; break;
+	case '{': post_delim = '}'; break;
+	default: post_delim = '\0'; break;
+	}
+
+	init_nno_saw(nno, 0);
+
+	for (nni = nno->list, len = 0; nni->name; nni++) {
+		if (nni->main_name) {
+			if (!dup_markup)
+				continue;
+			delim = dup_markup;
+		}
+		if (nni->num == 0 && !am_server && !dup_markup)
+			continue;
+		if (len)
+			to_buf[len++]= ' ';
+		if (delim) {
+			to_buf[len++]= delim;
+			delim = post_delim;
+		}
+		len += strlcpy(to_buf+len, nni->name, to_buf_len - len);
+		if (len >= to_buf_len - 3)
+			exit_cleanup(RERR_UNSUPPORTED); /* IMPOSSIBLE... */
+		if (delim) {
+			to_buf[len++]= delim;
+			delim = '\0';
+		}
+		nno->saw[nni->num] = ++cnt;
+	}
+
+	return len;
+}
+
+static void send_negotiate_str(int f_out, struct name_num_obj *nno, int ntype)
+{
+	char tmpbuf[MAX_NSTR_STRLEN];
+	const char *list_str = getenv_nstr(ntype);
+	int len;
+
+	if (list_str && *list_str) {
+		init_nno_saw(nno, 0);
+		len = parse_nni_str(nno, list_str, tmpbuf, MAX_NSTR_STRLEN);
+		list_str = tmpbuf;
+	} else
+		list_str = NULL;
+
+	if (!list_str || !*list_str)
+		len = get_default_nno_list(nno, tmpbuf, MAX_NSTR_STRLEN, '\0');
+
+	if (DEBUG_GTE(NSTR, am_server ? 3 : 2)) {
+		if (am_server)
+			rprintf(FINFO, "Server %s list (on server): %s\n", nno->type, tmpbuf);
+		else
+			rprintf(FINFO, "Client %s list (on client): %s\n", nno->type, tmpbuf);
+	}
+
+	/* Each side sends their list of valid names to the other side and then both sides
+	 * pick the first name in the client's list that is also in the server's list. */
+	if (do_negotiated_strings)
+		write_vstring(f_out, tmpbuf, len);
+}
+
+static void negotiate_the_strings(int f_in, int f_out)
+{
+	/* We send all the negotiation strings before we start to read them to help avoid a slow startup. */
+
+	if (!checksum_choice)
+		send_negotiate_str(f_out, &valid_checksums, NSTR_CHECKSUM);
+
+	if (do_compression && !compress_choice)
+		send_negotiate_str(f_out, &valid_compressions, NSTR_COMPRESS);
+
+	if (valid_checksums.saw) {
+		char tmpbuf[MAX_NSTR_STRLEN];
+		int len;
+		if (do_negotiated_strings)
+			len = -1;
+		else
+			len = strlcpy(tmpbuf, protocol_version >= 30 ? "md5" : "md4", MAX_NSTR_STRLEN);
+		recv_negotiate_str(f_in, &valid_checksums, tmpbuf, len);
+	}
+
+	if (valid_compressions.saw) {
+		char tmpbuf[MAX_NSTR_STRLEN];
+		int len;
+		if (do_negotiated_strings)
+			len = -1;
+		else
+			len = strlcpy(tmpbuf, "zlib", MAX_NSTR_STRLEN);
+		recv_negotiate_str(f_in, &valid_compressions, tmpbuf, len);
+	}
+
+	/* If the other side is too old to negotiate, the above steps just made sure that
+	 * the env didn't disallow the old algorithm. Mark things as non-negotiated. */
+	if (!do_negotiated_strings)
+		valid_checksums.negotiated_name = valid_compressions.negotiated_name = NULL;
+}
 
 void setup_protocol(int f_out,int f_in)
 {
+	assert(file_extra_cnt == 0);
+	assert(EXTRA64_CNT == 2 || EXTRA64_CNT == 1);
+
+	/* All int64 values must be set first so that they are guaranteed to be
+	 * aligned for direct int64-pointer memory access. */
+	if (preserve_atimes)
+		atimes_ndx = (file_extra_cnt += EXTRA64_CNT);
+	if (am_sender) /* This is most likely in the in64 union as well. */
+		pathname_ndx = (file_extra_cnt += PTR_EXTRA_CNT);
+	else
+		depth_ndx = ++file_extra_cnt;
+	if (preserve_uid)
+		uid_ndx = ++file_extra_cnt;
+	if (preserve_gid)
+		gid_ndx = ++file_extra_cnt;
+	if (preserve_acls && !am_sender)
+		acls_ndx = ++file_extra_cnt;
+	if (preserve_xattrs)
+		xattrs_ndx = ++file_extra_cnt;
+
+	if (am_server)
+		set_allow_inc_recurse();
+
 	if (remote_protocol == 0) {
+		if (am_server && !local_server)
+			check_sub_protocol();
 		if (!read_batch)
 			write_int(f_out, protocol_version);
 		remote_protocol = read_int(f_in);
@@ -48,12 +579,12 @@ void setup_protocol(int f_out,int f_in)
 			protocol_version = remote_protocol;
 	}
 	if (read_batch && remote_protocol > protocol_version) {
-	        rprintf(FERROR, "The protocol version in the batch file is too new (%d > %d).\n",
+		rprintf(FERROR, "The protocol version in the batch file is too new (%d > %d).\n",
 			remote_protocol, protocol_version);
 		exit_cleanup(RERR_PROTOCOL);
 	}
 
-	if (verbose > 3) {
+	if (DEBUG_GTE(PROTO, 1)) {
 		rprintf(FINFO, "(%s) Protocol versions: remote=%d, negotiated=%d\n",
 			am_server? "Server" : "Client", remote_protocol, protocol_version);
 	}
@@ -77,33 +608,181 @@ void setup_protocol(int f_out,int f_in)
 			PROTOCOL_VERSION, am_server? "Server" : "Client");
 		exit_cleanup(RERR_PROTOCOL);
 	}
+	if (read_batch)
+		check_batch_flags();
 
-	if (fuzzy_basis && protocol_version < 29) {
-		rprintf(FERROR,
-			"--fuzzy requires protocol 29 or higher (negotiated %d).\n",
-			protocol_version);
-		exit_cleanup(RERR_PROTOCOL);
+#ifndef SUPPORT_PREALLOCATION
+	if (preallocate_files && !am_sender) {
+		rprintf(FERROR, "preallocation is not supported on this %s\n",
+			am_server ? "Server" : "Client");
+		exit_cleanup(RERR_SYNTAX);
+	}
+#endif
+
+	if (protocol_version < 30) {
+		if (append_mode == 1)
+			append_mode = 2;
+		if (preserve_acls && !local_server) {
+			rprintf(FERROR,
+				"--acls requires protocol 30 or higher"
+				" (negotiated %d).\n",
+				protocol_version);
+			exit_cleanup(RERR_PROTOCOL);
+		}
+		if (preserve_xattrs && !local_server) {
+			rprintf(FERROR,
+				"--xattrs requires protocol 30 or higher"
+				" (negotiated %d).\n",
+				protocol_version);
+			exit_cleanup(RERR_PROTOCOL);
+		}
 	}
 
-	if (basis_dir_cnt && inplace && protocol_version < 29) {
-		rprintf(FERROR,
-			"%s with --inplace requires protocol 29 or higher (negotiated %d).\n",
-			dest_option, protocol_version);
-		exit_cleanup(RERR_PROTOCOL);
+	if (delete_mode && !(delete_before+delete_during+delete_after)) {
+		if (protocol_version < 30)
+			delete_before = 1;
+		else
+			delete_during = 1;
 	}
 
-	if (basis_dir_cnt > 1 && protocol_version < 29) {
-		rprintf(FERROR,
-			"Multiple %s options requires protocol 29 or higher (negotiated %d).\n",
-			dest_option, protocol_version);
-		exit_cleanup(RERR_PROTOCOL);
+	if (protocol_version < 29) {
+		if (fuzzy_basis) {
+			rprintf(FERROR,
+				"--fuzzy requires protocol 29 or higher"
+				" (negotiated %d).\n",
+				protocol_version);
+			exit_cleanup(RERR_PROTOCOL);
+		}
+
+		if (basis_dir_cnt && inplace) {
+			rprintf(FERROR,
+				"%s with --inplace requires protocol 29 or higher"
+				" (negotiated %d).\n",
+				alt_dest_opt(0), protocol_version);
+			exit_cleanup(RERR_PROTOCOL);
+		}
+
+		if (basis_dir_cnt > 1) {
+			rprintf(FERROR,
+				"Using more than one %s option requires protocol"
+				" 29 or higher (negotiated %d).\n",
+				alt_dest_opt(0), protocol_version);
+			exit_cleanup(RERR_PROTOCOL);
+		}
+
+		if (prune_empty_dirs) {
+			rprintf(FERROR,
+				"--prune-empty-dirs requires protocol 29 or higher"
+				" (negotiated %d).\n",
+				protocol_version);
+			exit_cleanup(RERR_PROTOCOL);
+		}
+	} else if (protocol_version >= 30) {
+		if (am_server) {
+			compat_flags = allow_inc_recurse ? CF_INC_RECURSE : 0;
+#ifdef CAN_SET_SYMLINK_TIMES
+			compat_flags |= CF_SYMLINK_TIMES;
+#endif
+#ifdef ICONV_OPTION
+			compat_flags |= CF_SYMLINK_ICONV;
+#endif
+			if (local_server || strchr(client_info, 'f') != NULL)
+				compat_flags |= CF_SAFE_FLIST;
+			if (local_server || strchr(client_info, 'x') != NULL)
+				compat_flags |= CF_AVOID_XATTR_OPTIM;
+			if (local_server || strchr(client_info, 'C') != NULL)
+				compat_flags |= CF_CHKSUM_SEED_FIX;
+			if (local_server || strchr(client_info, 'I') != NULL)
+				compat_flags |= CF_INPLACE_PARTIAL_DIR;
+			if (local_server || strchr(client_info, 'v') != NULL) {
+				do_negotiated_strings = 1;
+				compat_flags |= CF_VARINT_FLIST_FLAGS;
+			}
+			if (strchr(client_info, 'V') != NULL) { /* Support a pre-release 'V' that got superseded */
+				if (!write_batch)
+					compat_flags |= CF_VARINT_FLIST_FLAGS;
+				write_byte(f_out, compat_flags);
+			} else
+				write_varint(f_out, compat_flags);
+		} else { /* read_varint() is compatible with the older write_byte() when the 0x80 bit isn't on. */
+			compat_flags = read_varint(f_in);
+			if  (compat_flags & CF_VARINT_FLIST_FLAGS)
+				do_negotiated_strings = 1;
+		}
+		/* The inc_recurse var MUST be set to 0 or 1. */
+		inc_recurse = compat_flags & CF_INC_RECURSE ? 1 : 0;
+		want_xattr_optim = protocol_version >= 31 && !(compat_flags & CF_AVOID_XATTR_OPTIM);
+		proper_seed_order = compat_flags & CF_CHKSUM_SEED_FIX ? 1 : 0;
+		xfer_flags_as_varint = compat_flags & CF_VARINT_FLIST_FLAGS ? 1 : 0;
+		if (am_sender) {
+			receiver_symlink_times = am_server
+			    ? strchr(client_info, 'L') != NULL
+			    : !!(compat_flags & CF_SYMLINK_TIMES);
+		}
+#ifdef CAN_SET_SYMLINK_TIMES
+		else
+			receiver_symlink_times = 1;
+#endif
+#ifdef ICONV_OPTION
+		sender_symlink_iconv = iconv_opt && (am_server
+		    ? local_server || strchr(client_info, 's') != NULL
+		    : !!(compat_flags & CF_SYMLINK_ICONV));
+#endif
+		if (inc_recurse && !allow_inc_recurse) {
+			/* This should only be able to happen in a batch. */
+			fprintf(stderr,
+				"Incompatible options specified for inc-recursive %s.\n",
+				read_batch ? "batch file" : "connection");
+			exit_cleanup(RERR_SYNTAX);
+		}
+		use_safe_inc_flist = (compat_flags & CF_SAFE_FLIST) || protocol_version >= 31;
+		need_messages_from_generator = 1;
+		if (compat_flags & CF_INPLACE_PARTIAL_DIR)
+			inplace_partial = 1;
+#ifdef CAN_SET_SYMLINK_TIMES
+	} else if (!am_sender) {
+		receiver_symlink_times = 1;
+#endif
 	}
 
+	if (read_batch)
+		do_negotiated_strings = 0;
+
+	if (need_unsorted_flist && (!am_sender || inc_recurse))
+		unsort_ndx = ++file_extra_cnt;
+
+	if (partial_dir && *partial_dir != '/' && (!am_server || local_server)) {
+		int rflags = FILTRULE_NO_PREFIXES | FILTRULE_DIRECTORY;
+		if (!am_sender || protocol_version >= 30)
+			rflags |= FILTRULE_PERISHABLE;
+		parse_filter_str(&filter_list, partial_dir, rule_template(rflags), 0);
+	}
+
+
+#ifdef ICONV_OPTION
+	if (protect_args && files_from) {
+		if (am_sender)
+			filesfrom_convert = filesfrom_host && ic_send != (iconv_t)-1;
+		else
+			filesfrom_convert = !filesfrom_host && ic_recv != (iconv_t)-1;
+	}
+#endif
+
+	negotiate_the_strings(f_in, f_out);
+
 	if (am_server) {
 		if (!checksum_seed)
-			checksum_seed = time(NULL);
+			checksum_seed = time(NULL) ^ (getpid() << 6);
 		write_int(f_out, checksum_seed);
 	} else {
 		checksum_seed = read_int(f_in);
 	}
+
+	parse_checksum_choice(1); /* Sets checksum_type & xfersum_type */
+	parse_compress_choice(1); /* Sets do_compression */
+
+	if (write_batch && !am_server)
+		write_batch_shell_file();
+
+	init_flist();
 }

configure

@@ -0,0 +1,27 @@
+#!/bin/sh -e
+# This configure script ensures that the configure.sh script exists, and
+# if not, it tries to fetch rsync's generated files or build them.  We
+# then transfer control to the configure.sh script to do the real work.
+
+dir=`dirname $0`
+if test x"$dir" = x; then
+    dir=.
+fi
+
+if test "$dir" = '.'; then
+    branch=`packaging/prep-auto-dir` || exit 1
+    if test x"$branch" != x; then
+	cd build || exit 1
+	dir=..
+    fi
+fi
+
+if test ! -f configure.sh; then
+    if ! "$dir/prepare-source" build; then
+	echo 'Failed to build configure.sh and/or config.h.in -- giving up.' >&2
+	rm -f configure.sh
+	exit 1
+    fi
+fi
+
+exec ./configure.sh --srcdir="$dir" "${@}"

configure.in

@@ -1,736 +0,0 @@
-dnl Process this file with autoconf to produce a configure script.
-
-AC_INIT()
-AC_CONFIG_SRCDIR([byteorder.h])
-AC_CONFIG_HEADER(config.h)
-AC_PREREQ(2.59)
-
-RSYNC_VERSION=2.6.4
-AC_SUBST(RSYNC_VERSION)
-AC_MSG_NOTICE([Configuring rsync $RSYNC_VERSION])
-
-AC_DEFINE_UNQUOTED(RSYNC_VERSION, ["$RSYNC_VERSION"], [rsync release version])
-
-LDFLAGS=${LDFLAGS-""}
-
-AC_CANONICAL_TARGET([])
-
-dnl Checks for programs.
-AC_PROG_CC
-AC_PROG_CPP
-AC_PROG_EGREP
-AC_PROG_INSTALL
-AC_PROG_CC_STDC
-AC_SUBST(SHELL)
-
-AC_DEFINE([_GNU_SOURCE], 1,
-          [Define _GNU_SOURCE so that we get all necessary prototypes])
-
-if test "x$ac_cv_prog_cc_stdc" = xno
-then
-	AC_MSG_WARN([rsync requires an ANSI C compiler and you don't seem to have one])
-fi
-
-# We must decide this before testing the compiler.
-
-# Please allow this to default to yes, so that your users have more
-# chance of getting a useful stack trace if problems occur.
-
-AC_MSG_CHECKING([whether to include debugging symbols])
-AC_ARG_ENABLE(debug,
-	AC_HELP_STRING([--enable-debug],
-		[including debugging symbols and features (default yes)]),
-		[], [])
-
-if test x"$enable_debug" = x"no"
-then
-    AC_MSG_RESULT(no)
-    CFLAGS=${CFLAGS-"-O"}
-else
-    AC_MSG_RESULT([yes])
-    # leave CFLAGS alone; AC_PROG_CC will try to include -g if it can
-    dnl AC_DEFINE(DEBUG, 1, [Define to turn on debugging code that may slow normal operation])
-    dnl CFLAGS=${CFLAGS-"-g"}
-fi
-
-
-
-
-AC_ARG_ENABLE(profile,
-	AC_HELP_STRING([--enable-profile],
-		[turn on CPU profiling (default no)],
-		[], []))
-if test x"$enable_profile" = xyes
-then
-	CFLAGS="$CFLAGS -pg"
-fi
-
-
-# Specifically, this turns on panic_action handling.
-AC_ARG_ENABLE(maintainer-mode,
-	AC_HELP_STRING([--enable-maintainer-mode],
-		[turn on extra debug features],
-		[], []))
-if test x"$enable_maintainer_mode" = xyes
-then
-	CFLAGS="$CFLAGS -DMAINTAINER_MODE"
-fi
-
-
-# This is needed for our included version of popt.  Kind of silly, but
-# I don't want our version too far out of sync.
-CFLAGS="$CFLAGS -DHAVE_CONFIG_H"
-
-# If GCC, turn on warnings.
-if test x"$GCC" = x"yes"
-then
-	CFLAGS="$CFLAGS -Wall -W"
-fi
-
-AC_ARG_WITH(included-popt,
-        [  --with-included-popt    use bundled popt library, not from system])
-
-AC_ARG_WITH(rsync-path,
-	[  --with-rsync-path=PATH  set default --rsync-path to PATH (default: rsync)],
-	[ RSYNC_PATH="$with_rsync_path" ],
-	[ RSYNC_PATH="rsync" ])
-
-AC_DEFINE_UNQUOTED(RSYNC_PATH, "$RSYNC_PATH", [location of rsync on remote machine])
-
-AC_ARG_WITH(rsyncd-conf,
-	AC_HELP_STRING([--with-rsyncd-conf=PATH], [set configuration file for rsync server to PATH (default: /etc/rsyncd.conf)]),
-	[ if test ! -z "$with_rsyncd_conf" ; then
-		case $with_rsyncd_conf in
-			yes|no)
-				RSYNCD_SYSCONF="/etc/rsyncd.conf"
-				;;
-			/*)
-				RSYNCD_SYSCONF="$with_rsyncd_conf"
-				;;
-			*)
-                                AC_MSG_ERROR(You must specify an absolute path to --with-rsyncd-conf=PATH)
-				;;
-		esac
-	else
-		RSYNCD_SYSCONF="/etc/rsyncd.conf"
-	fi ],
-	[ RSYNCD_SYSCONF="/etc/rsyncd.conf" ])
-
-AC_DEFINE_UNQUOTED(RSYNCD_SYSCONF, "$RSYNCD_SYSCONF", [location of configuration file for rsync server])
-
-AC_ARG_WITH(rsh,
-	AC_HELP_STRING([--with-rsh=CMD], [set remote shell command to CMD (default: ssh)]))
-
-AC_CHECK_PROG(HAVE_REMSH, remsh, 1, 0)
-AC_DEFINE_UNQUOTED(HAVE_REMSH, $HAVE_REMSH, [remote shell is remsh not rsh])
-
-if test x"$with_rsh" != x
-then
-	RSYNC_RSH="$with_rsh"
-else
-	RSYNC_RSH="ssh"
-fi
-
-AC_DEFINE_UNQUOTED(RSYNC_RSH, "$RSYNC_RSH", [default -e command])
-
-# arrgh. libc in the current debian stable screws up the largefile
-# stuff, getting byte range locking wrong
-AC_CACHE_CHECK([for broken largefile support],rsync_cv_HAVE_BROKEN_LARGEFILE,[
-AC_TRY_RUN([
-#define _FILE_OFFSET_BITS 64
-#include <stdio.h>
-#include <fcntl.h>
-#include <sys/types.h>
-#include <sys/wait.h>
-
-int main(void)
-{
-	struct flock lock;
-	int status;
-	char tpl[32] = "/tmp/locktest.XXXXXX";
-	int fd = mkstemp(tpl);
-	if (fd < 0) {
-		strcpy(tpl, "conftest.dat");
-		fd = open(tpl, O_CREAT|O_RDWR, 0600);
-	}
-
-	lock.l_type = F_WRLCK;
-	lock.l_whence = SEEK_SET;
-	lock.l_start = 0;
-	lock.l_len = 1;
-	lock.l_pid = 0;
-	fcntl(fd,F_SETLK,&lock);
-	if (fork() == 0) {
-		lock.l_start = 1;
-		_exit(fcntl(fd,F_SETLK,&lock) == 0);
-	}
-	wait(&status);
-	unlink(tpl);
-	exit(WEXITSTATUS(status));
-}
-],
-rsync_cv_HAVE_BROKEN_LARGEFILE=yes,rsync_cv_HAVE_BROKEN_LARGEFILE=no,rsync_cv_HAVE_BROKEN_LARGEFILE=cross)])
-if test x"$rsync_cv_HAVE_BROKEN_LARGEFILE" != x"yes"; then
-   AC_SYS_LARGEFILE
-fi
-
-ipv6type=unknown
-ipv6lib=none
-ipv6trylibc=yes
-
-AC_ARG_ENABLE(ipv6,
-	AC_HELP_STRING([--disable-ipv6], [don't even try to use IPv6]))
-
-if test "x$enable_ipv6" != xno
-then
-	AC_MSG_CHECKING([ipv6 stack type])
-	for i in inria kame linux-glibc linux-inet6 toshiba v6d zeta; do
-		case $i in
-		inria)
-			# http://www.kame.net/
-			AC_EGREP_CPP(yes, [
-#include <netinet/in.h>
-#ifdef IPV6_INRIA_VERSION
-yes
-#endif],
-				[ipv6type=$i;
-				AC_DEFINE(INET6, 1, [true if you have IPv6])
-				])
-			;;
-		kame)
-			# http://www.kame.net/
-			AC_EGREP_CPP(yes, [
-#include <netinet/in.h>
-#ifdef __KAME__
-yes
-#endif],
-				[ipv6type=$i;
-				AC_DEFINE(INET6, 1, [true if you have IPv6])])
-			;;
-		linux-glibc)
-			# http://www.v6.linux.or.jp/
-			AC_EGREP_CPP(yes, [
-#include <features.h>
-#if defined(__GLIBC__) && __GLIBC__ >= 2 && __GLIBC_MINOR__ >= 1
-yes
-#endif],
-				[ipv6type=$i;
-AC_DEFINE(INET6, 1, [true if you have IPv6])])
-			;;
-		linux-inet6)
-			# http://www.v6.linux.or.jp/
-			if test -d /usr/inet6 -o -f /usr/inet6/lib/libinet6.a; then
-				ipv6type=$i
-				ipv6lib=inet6
-				ipv6libdir=/usr/inet6/lib
-				ipv6trylibc=yes;
-				AC_DEFINE(INET6, 1, [true if you have IPv6])
-				CFLAGS="-I/usr/inet6/include $CFLAGS"
-			fi
-			;;
-		toshiba)
-			AC_EGREP_CPP(yes, [
-#include <sys/param.h>
-#ifdef _TOSHIBA_INET6
-yes
-#endif],
-				[ipv6type=$i;
-				ipv6lib=inet6;
-				ipv6libdir=/usr/local/v6/lib;
-				AC_DEFINE(INET6, 1, [true if you have IPv6])])
-			;;
-		v6d)
-			AC_EGREP_CPP(yes, [
-#include </usr/local/v6/include/sys/v6config.h>
-#ifdef __V6D__
-yes
-#endif],
-				[ipv6type=$i;
-				ipv6lib=v6;
-				ipv6libdir=/usr/local/v6/lib;
-				AC_DEFINE(INET6, 1, [true if you have IPv6])])
-			;;
-		zeta)
-			AC_EGREP_CPP(yes, [
-#include <sys/param.h>
-#ifdef _ZETA_MINAMI_INET6
-yes
-#endif],
-				[ipv6type=$i;
-				ipv6lib=inet6;
-				ipv6libdir=/usr/local/v6/lib;
-				AC_DEFINE(INET6, 1, [true if you have IPv6])])
-			;;
-		esac
-		if test "$ipv6type" != "unknown"; then
-			break
-		fi
-	done
-	AC_MSG_RESULT($ipv6type)
-
-	AC_SEARCH_LIBS(getaddrinfo, inet6)
-fi
-
-AC_MSG_CHECKING([whether to call shutdown on all sockets])
-case $host_os in
-	*cygwin* ) AC_MSG_RESULT(yes)
-                   AC_DEFINE(SHUTDOWN_ALL_SOCKETS, 1,
-			    [Define to 1 if sockets need to be shutdown])
-		   ;;
-	       * ) AC_MSG_RESULT(no);;
-esac
-
-AC_C_BIGENDIAN
-AC_HEADER_DIRENT
-AC_HEADER_TIME
-AC_HEADER_SYS_WAIT
-AC_CHECK_HEADERS(sys/fcntl.h sys/select.h fcntl.h sys/time.h sys/unistd.h \
-    unistd.h utime.h grp.h compat.h sys/param.h ctype.h sys/wait.h \
-    sys/ioctl.h sys/filio.h string.h stdlib.h sys/socket.h sys/mode.h \
-    sys/un.h glob.h mcheck.h arpa/inet.h arpa/nameser.h \
-    netdb.h malloc.h float.h)
-AC_HEADER_MAJOR
-
-AC_CHECK_SIZEOF(int)
-AC_CHECK_SIZEOF(long)
-AC_CHECK_SIZEOF(long long)
-AC_CHECK_SIZEOF(short)
-AC_CHECK_SIZEOF(off_t)
-AC_CHECK_SIZEOF(off64_t)
-
-AC_C_INLINE
-AC_C_LONG_DOUBLE
-
-AC_TYPE_SIGNAL
-AC_TYPE_UID_T
-AC_TYPE_MODE_T
-AC_TYPE_OFF_T
-AC_TYPE_SIZE_T
-AC_TYPE_PID_T
-AC_TYPE_GETGROUPS
-AC_CHECK_MEMBERS([struct stat.st_rdev])
-
-TYPE_SOCKLEN_T
-
-AC_CACHE_CHECK([for errno in errno.h],rsync_cv_errno, [
-    AC_TRY_COMPILE([#include <errno.h>],[int i = errno],
-	rsync_cv_errno=yes,rsync_cv_have_errno_decl=no)])
-if test x"$rsync_cv_errno" = x"yes"; then
-   AC_DEFINE(HAVE_ERRNO_DECL, 1, [Define to 1 if errno is declared in errno.h])
-fi
-
-# The following test taken from the cvs sources
-# If we can't find connect, try looking in -lsocket, -lnsl, and -linet.
-# These need checks to be before checks for any other functions that
-#    might be in the same libraries.
-# The Irix 5 libc.so has connect and gethostbyname, but Irix 5 also has
-# libsocket.so which has a bad implementation of gethostbyname (it
-# only looks in /etc/hosts), so we only look for -lsocket if we need
-# it.
-AC_CHECK_FUNCS(connect)
-if test x"$ac_cv_func_connect" = x"no"; then
-    case "$LIBS" in
-    *-lnsl*) ;;
-    *) AC_CHECK_LIB(nsl_s, printf) ;;
-    esac
-    case "$LIBS" in
-    *-lnsl*) ;;
-    *) AC_CHECK_LIB(nsl, printf) ;;
-    esac
-    case "$LIBS" in
-    *-lsocket*) ;;
-    *) AC_CHECK_LIB(socket, connect) ;;
-    esac
-    case "$LIBS" in
-    *-linet*) ;;
-    *) AC_CHECK_LIB(inet, connect) ;;
-    esac
-    dnl We can't just call AC_CHECK_FUNCS(connect) here, because the value
-    dnl has been cached.
-    if test x"$ac_cv_lib_socket_connect" = x"yes" ||
-       test x"$ac_cv_lib_inet_connect" = x"yes"; then
-        # ac_cv_func_connect=yes
-        # don't!  it would cause AC_CHECK_FUNC to succeed next time configure is run
-        AC_DEFINE(HAVE_CONNECT, 1, [Define to 1 if you have the "connect" function])
-    fi
-fi
-
-AC_CHECK_LIB(resolv, inet_ntop)
-
-dnl AC_MSG_NOTICE([Looking in libraries: $LIBS])
-
-AC_CHECK_FUNCS(inet_ntop, , [AC_LIBOBJ(lib/inet_ntop)])
-AC_CHECK_FUNCS(inet_pton, , [AC_LIBOBJ(lib/inet_pton)])
-
-# Irix 6.5 has getaddrinfo but not the corresponding defines, so use
-#   builtin getaddrinfo if one of the defines don't exist
-AC_CACHE_CHECK([whether defines needed by getaddrinfo exist],
-               rsync_cv_HAVE_GETADDR_DEFINES,[
-			AC_EGREP_CPP(yes, [
-			#include <sys/types.h>
-			#include <sys/socket.h>
-			#include <netdb.h>
-			#ifdef AI_PASSIVE
-			yes
-			#endif],
-			rsync_cv_HAVE_GETADDR_DEFINES=yes,
-			rsync_cv_HAVE_GETADDR_DEFINES=no)])
-if test x"$rsync_cv_HAVE_GETADDR_DEFINES" = x"yes"; then
-	# Tru64 UNIX has getaddrinfo() but has it renamed in libc as
-	# something else so we must include <netdb.h> to get the
-	# redefinition.
-	AC_CHECK_FUNCS(getaddrinfo, ,
-		[AC_MSG_CHECKING([for getaddrinfo by including <netdb.h>])
-		AC_TRY_LINK([#include <sys/types.h>
-		#include <sys/socket.h>
-		#include <netdb.h>],[getaddrinfo(NULL, NULL, NULL, NULL);],
-			[AC_MSG_RESULT([yes])
-			AC_DEFINE(HAVE_GETADDRINFO, 1,
-				[Define to 1 if you have the "getaddrinfo" function.])],
-			[AC_MSG_RESULT([no])
-			AC_LIBOBJ(lib/getaddrinfo)])])
-	AC_CHECK_FUNCS(getnameinfo, , [AC_LIBOBJ(lib/getnameinfo)])
-else
-	AC_LIBOBJ(lib/getaddrinfo)
-	AC_LIBOBJ(lib/getnameinfo)
-fi
-
-AC_CHECK_MEMBER([struct sockaddr.sa_len],
-		[ AC_DEFINE(HAVE_SOCKADDR_LEN, 1, [Do we have sockaddr.sa_len?]) ],
-		[],
-		[
-#include <sys/types.h>
-#include <sys/socket.h>
-])
-
-AC_CHECK_MEMBER([struct sockaddr_in.sin_len],
-		[ AC_DEFINE(HAVE_SOCKADDR_IN_LEN, 1, [Do we have sockaddr_in.sin_len?]) ],
-		[],
-		[
-#include <sys/types.h>
-#include <sys/socket.h>
-#include <netinet/in.h>
-])
-
-AC_CHECK_MEMBER([struct sockaddr_un.sun_len],
-		[ AC_DEFINE(HAVE_SOCKADDR_UN_LEN, 1, [Do we have sockaddr_un.sun_len?]) ],
-		[],
-		[
-#include <sys/types.h>
-#include <sys/socket.h>
-#include <netinet/in.h>
-])
-
-AC_MSG_CHECKING(struct sockaddr_storage)
-AC_TRY_COMPILE([#include <sys/types.h>
-#include <sys/socket.h>],
-[struct sockaddr_storage x;],
-	AC_MSG_RESULT(yes)
-	AC_DEFINE(HAVE_SOCKADDR_STORAGE, 1,
-		[Define to 1 if you have struct sockaddr_storage.] ),
-	AC_MSG_RESULT(no))
-
-AC_CHECK_MEMBER([struct sockaddr_in6.sin6_scope_id],
-		[ AC_DEFINE(HAVE_SOCKADDR_IN6_SCOPE_ID, 1, [Do we have sockaddr_in6.sin6_scope_id?]) ],
-		[],
-		[
-#include <sys/types.h>
-#include <sys/socket.h>
-#include <netinet/in.h>
-])
-
-AC_MSG_CHECKING(struct stat64)
-AC_TRY_COMPILE([#include <stdio.h>
-#if HAVE_SYS_TYPES_H
-# include <sys/types.h>
-#endif
-#if HAVE_SYS_STAT_H
-# include <sys/stat.h>
-#endif
-#if STDC_HEADERS
-# include <stdlib.h>
-# include <stddef.h>
-#else
-# if HAVE_STDLIB_H
-#  include <stdlib.h>
-# endif
-#endif
-],[struct stat64 st;],
-    AC_MSG_RESULT(yes)
-	AC_DEFINE(HAVE_STRUCT_STAT64,1,[Define to 1 if you have struct stat64.]),
-    AC_MSG_RESULT(no))
-
-# if we can't find strcasecmp, look in -lresolv (for Unixware at least)
-#
-AC_CHECK_FUNCS(strcasecmp)
-if test x"$ac_cv_func_strcasecmp" = x"no"; then
-    AC_CHECK_LIB(resolv, strcasecmp)
-fi
-
-dnl At the moment we don't test for a broken memcmp(), because all we
-dnl need to do is test for equality, not comparison, and it seems that
-dnl every platform has a memcmp that can do at least that.
-dnl AC_FUNC_MEMCMP
-
-AC_FUNC_UTIME_NULL
-AC_FUNC_ALLOCA
-AC_CHECK_FUNCS(waitpid wait4 getcwd strdup strerror chown chmod mknod mkfifo \
-    fchmod fstat strchr readlink link utime utimes strftime mtrace ftruncate \
-    memmove lchown vsnprintf snprintf vasprintf asprintf setsid glob strpbrk \
-    strlcat strlcpy strtol mallinfo getgroups setgroups geteuid getegid \
-    setmode open64 mkstemp64 va_copy __va_copy)
-
-AC_CHECK_FUNCS(getpgrp tcgetpgrp)
-if test $ac_cv_func_getpgrp = yes; then
-    AC_FUNC_GETPGRP
-fi
-
-# Determine whether chown follows symlinks (it should).
-AC_CACHE_CHECK([whether chown() dereferences symlinks],rsync_cv_chown_follows_symlink,[
-  AC_TRY_RUN([
-#if HAVE_UNISTD_H
-# include <unistd.h>
-#endif
-#include <stdlib.h>
-#include <errno.h>
-    main() {
-	char const *dangling_symlink = "conftest.dangle";
-	unlink(dangling_symlink);
-	if (symlink("conftest.no-such", dangling_symlink) < 0) abort();
-	if (chown(dangling_symlink, getuid(), getgid()) < 0 && errno == ENOENT) exit(0);
-	exit(1);
-    }],
-  rsync_cv_chown_follows_symlink=yes,rsync_cv_chown_follows_symlink=no,rsync_cv_chown_follows_symlink=yes)])
-if test $rsync_cv_chown_follows_symlink = no; then
-  AC_DEFINE(CHOWN_MODIFIES_SYMLINK, 1, [Define to 1 if chown modifies symlinks.])
-fi
-
-AC_CACHE_CHECK([for working socketpair],rsync_cv_HAVE_SOCKETPAIR,[
-AC_TRY_RUN([
-#include <sys/types.h>
-#include <sys/socket.h>
-
-main() {
-       int fd[2];
-       exit((socketpair(AF_UNIX, SOCK_STREAM, 0, fd) != -1) ? 0 : 1);
-}],
-rsync_cv_HAVE_SOCKETPAIR=yes,rsync_cv_HAVE_SOCKETPAIR=no,rsync_cv_HAVE_SOCKETPAIR=cross)])
-if test x"$rsync_cv_HAVE_SOCKETPAIR" = x"yes"; then
-    AC_DEFINE(HAVE_SOCKETPAIR, 1, [Define to 1 if you have the "socketpair" function])
-fi
-
-if test x"$with_included_popt" != x"yes"
-then
-    AC_CHECK_LIB(popt, poptGetContext, , [with_included_popt=yes])
-fi
-
-AC_MSG_CHECKING([whether to use included libpopt])
-if test x"$with_included_popt" = x"yes"
-then
-    AC_MSG_RESULT($srcdir/popt)
-    BUILD_POPT='$(popt_OBJS)'
-    CFLAGS="$CFLAGS -I$srcdir/popt"
-    if test x"$ALLOCA" != x
-    then
-	# this can be removed when/if we add an included alloca.c;
-	#  see autoconf documentation on AC_FUNC_ALLOCA
-	AC_MSG_WARN([included libpopt will use malloc, not alloca (which wastes a small amount of memory)])
-    fi
-else
-    AC_MSG_RESULT(no)
-fi
-
-AC_CACHE_CHECK([for unsigned char],rsync_cv_SIGNED_CHAR_OK,[
-AC_TRY_COMPILE([],[signed char *s = ""],
-rsync_cv_SIGNED_CHAR_OK=yes,rsync_cv_SIGNED_CHAR_OK=no)])
-if test x"$rsync_cv_SIGNED_CHAR_OK" = x"yes"; then
-    AC_DEFINE(SIGNED_CHAR_OK, 1, [Define to 1 if "signed char" is a valid type])
-fi
-
-AC_CACHE_CHECK([for broken readdir],rsync_cv_HAVE_BROKEN_READDIR,[
-AC_TRY_RUN([#include <sys/types.h>
-#include <dirent.h>
-main() { struct dirent *di; DIR *d = opendir("."); di = readdir(d);
-if (di && di->d_name[-2] == '.' && di->d_name[-1] == 0 &&
-di->d_name[0] == 0) exit(0); exit(1);} ],
-rsync_cv_HAVE_BROKEN_READDIR=yes,rsync_cv_HAVE_BROKEN_READDIR=no,rsync_cv_HAVE_BROKEN_READDIR=cross)])
-if test x"$rsync_cv_HAVE_BROKEN_READDIR" = x"yes"; then
-    AC_DEFINE(HAVE_BROKEN_READDIR, 1, [Define to 1 if readdir() is broken])
-fi
-
-AC_CACHE_CHECK([for utimbuf],rsync_cv_HAVE_UTIMBUF,[
-AC_TRY_COMPILE([#include <sys/types.h>
-#include <utime.h>],
-[struct utimbuf tbuf;  tbuf.actime = 0; tbuf.modtime = 1; exit(utime("foo.c",&tbuf));],
-rsync_cv_HAVE_UTIMBUF=yes,rsync_cv_HAVE_UTIMBUF=no)])
-if test x"$rsync_cv_HAVE_UTIMBUF" = x"yes"; then
-    AC_DEFINE(HAVE_UTIMBUF, 1, [Define to 1 if you have the "struct utimbuf" type])
-fi
-
-AC_CACHE_CHECK([if gettimeofday takes tz argument],rsync_cv_HAVE_GETTIMEOFDAY_TZ,[
-AC_TRY_COMPILE([#include <sys/time.h>
-#include <unistd.h>],
-[struct timeval tv; exit(gettimeofday(&tv, NULL));],
-rsync_cv_HAVE_GETTIMEOFDAY_TZ=yes,rsync_cv_HAVE_GETTIMEOFDAY_TZ=no)])
-if test x"$rsync_cv_HAVE_GETTIMEOFDAY_TZ" != x"no"; then
-    AC_DEFINE(HAVE_GETTIMEOFDAY_TZ, 1, [Define to 1 if gettimeofday() takes a time-zone arg])
-fi
-
-AC_CACHE_CHECK([for C99 vsnprintf],rsync_cv_HAVE_C99_VSNPRINTF,[
-AC_TRY_RUN([
-#include <sys/types.h>
-#include <stdarg.h>
-void foo(const char *format, ...) {
-       va_list ap;
-       int len;
-       char buf[5];
-
-       va_start(ap, format);
-       len = vsnprintf(0, 0, format, ap);
-       va_end(ap);
-       if (len != 5) exit(1);
-
-       if (snprintf(buf, 3, "hello") != 5 || strcmp(buf, "he") != 0) exit(1);
-
-       exit(0);
-}
-main() { foo("hello"); }
-],
-rsync_cv_HAVE_C99_VSNPRINTF=yes,rsync_cv_HAVE_C99_VSNPRINTF=no,rsync_cv_HAVE_C99_VSNPRINTF=cross)])
-if test x"$rsync_cv_HAVE_C99_VSNPRINTF" = x"yes"; then
-    AC_DEFINE(HAVE_C99_VSNPRINTF, 1, [Define to 1 if vsprintf has a C99-compatible return value])
-fi
-
-
-AC_CACHE_CHECK([for secure mkstemp],rsync_cv_HAVE_SECURE_MKSTEMP,[
-AC_TRY_RUN([#include <stdlib.h>
-#include <sys/types.h>
-#include <sys/stat.h>
-#include <unistd.h>
-main() {
-  struct stat st;
-  char tpl[20]="/tmp/test.XXXXXX";
-  int fd = mkstemp(tpl);
-  if (fd == -1) exit(1);
-  unlink(tpl);
-  if (fstat(fd, &st) != 0) exit(1);
-  if ((st.st_mode & 0777) != 0600) exit(1);
-  exit(0);
-}],
-rsync_cv_HAVE_SECURE_MKSTEMP=yes,
-rsync_cv_HAVE_SECURE_MKSTEMP=no,
-rsync_cv_HAVE_SECURE_MKSTEMP=cross)])
-if test x"$rsync_cv_HAVE_SECURE_MKSTEMP" = x"yes"; then
-    AC_DEFINE(HAVE_SECURE_MKSTEMP, 1, [Define to 1 if mkstemp() is available and works right])
-fi
-
-
-AC_CACHE_CHECK([for broken inet_ntoa],rsync_cv_REPLACE_INET_NTOA,[
-AC_TRY_RUN([
-#include <stdio.h>
-#include <sys/types.h>
-#include <netinet/in.h>
-#include <arpa/inet.h>
-main() { struct in_addr ip; ip.s_addr = 0x12345678;
-if (strcmp(inet_ntoa(ip),"18.52.86.120") &&
-    strcmp(inet_ntoa(ip),"120.86.52.18")) { exit(1); }
-exit(0);}],
-           rsync_cv_REPLACE_INET_NTOA=no,rsync_cv_REPLACE_INET_NTOA=yes,rsync_cv_REPLACE_INET_NTOA=cross)])
-if test x"$rsync_cv_REPLACE_INET_NTOA" = x"yes"; then
-    AC_DEFINE(REPLACE_INET_NTOA, 1, [Define to 1 if inet_ntoa() needs to be replaced])
-fi
-
-
-AC_CACHE_CHECK([for broken inet_aton],rsync_cv_REPLACE_INET_ATON,[
-AC_TRY_RUN([
-#include <stdio.h>
-#include <sys/types.h>
-#include <netinet/in.h>
-#include <arpa/inet.h>
-main() { struct in_addr ip;
-if (inet_aton("example", &ip) == 0) exit(0); exit(1);}],
-           rsync_cv_REPLACE_INET_ATON=no,rsync_cv_REPLACE_INET_ATON=yes,rsync_cv_REPLACE_INET_ATON=cross)])
-if test x"$rsync_cv_REPLACE_INET_ATON" = x"yes"; then
-    AC_DEFINE(REPLACE_INET_ATON, 1, [Define to 1 if inet_aton() needs to be replaced])
-fi
-
-AC_CACHE_CHECK([if mknod creates FIFOs],rsync_cv_MKNOD_CREATES_FIFOS,[
-AC_TRY_RUN([
-#include <stdio.h>
-#include <sys/stat.h>
-#include <errno.h>
-main() { int rc, ec; char *fn = "fifo-test";
-unlink(fn); rc = mknod(fn,S_IFIFO,0600); ec = errno; unlink(fn);
-if (rc) {printf("%d %d\n",rc,ec); return ec;}
-return 0;}],
-           rsync_cv_MKNOD_CREATES_FIFOS=yes,rsync_cv_MKNOD_CREATES_FIFOS=no,rsync_cv_MKNOD_CREATES_FIFOS=cross)])
-if test x"$rsync_cv_MKNOD_CREATES_FIFOS" = x"yes"; then
-    AC_DEFINE(MKNOD_CREATES_FIFOS, 1, [Define to 1 if mknod() can create FIFOs.])
-fi
-
-AC_CACHE_CHECK([if mknod creates sockets],rsync_cv_MKNOD_CREATES_SOCKETS,[
-AC_TRY_RUN([
-#include <stdio.h>
-#include <sys/stat.h>
-#include <errno.h>
-main() { int rc, ec; char *fn = "sock-test";
-unlink(fn); rc = mknod(fn,S_IFSOCK,0600); ec = errno; unlink(fn);
-if (rc) {printf("%d %d\n",rc,ec); return ec;}
-return 0;}],
-           rsync_cv_MKNOD_CREATES_SOCKETS=yes,rsync_cv_MKNOD_CREATES_SOCKETS=no,rsync_cv_MKNOD_CREATES_SOCKETS=cross)])
-if test x"$rsync_cv_MKNOD_CREATES_SOCKETS" = x"yes"; then
-    AC_DEFINE(MKNOD_CREATES_SOCKETS, 1, [Define to 1 if mknod() can create sockets.])
-fi
-
-#
-# The following test was mostly taken from the tcl/tk plus patches
-#
-AC_CACHE_CHECK([whether -c -o works],rsync_cv_DASHC_WORKS_WITH_DASHO,[
-rm -rf conftest*
-cat > conftest.$ac_ext <<EOF
-int main() { return 0; }
-EOF
-${CC-cc} -c -o conftest..o conftest.$ac_ext
-if test -f conftest..o; then
-    rsync_cv_DASHC_WORKS_WITH_DASHO=yes
-else
-    rsync_cv_DASHC_WORKS_WITH_DASHO=no
-fi
-rm -rf conftest*
-])
-if test x"$rsync_cv_DASHC_WORKS_WITH_DASHO" = x"yes"; then
-    OBJ_SAVE="#"
-    OBJ_RESTORE="#"
-    CC_SHOBJ_FLAG='-o $@'
-else
-    OBJ_SAVE='	@b=`basename $@ .o`;rm -f $$b.o.sav;if test -f $$b.o; then mv $$b.o $$b.o.sav;fi;'
-    OBJ_RESTORE='	@b=`basename $@ .o`;if test "$$b.o" != "$@"; then mv $$b.o $@; if test -f $$b.o.sav; then mv $$b.o.sav $$b.o; fi; fi'
-    CC_SHOBJ_FLAG=""
-fi
-
-AC_SUBST(OBJ_SAVE)
-AC_SUBST(OBJ_RESTORE)
-AC_SUBST(CC_SHOBJ_FLAG)
-AC_SUBST(BUILD_POPT)
-
-AC_CONFIG_FILES([Makefile lib/dummy zlib/dummy popt/dummy shconfig])
-AC_OUTPUT
-
-if test x"$with_rsh" = x; then
-	if test x"$HAVE_REMSH" = x1; then
-		rmsh1='remsh:'
-		rmsh2='=remsh'
-	else
-		rmsh1='rsh:  '
-		rmsh2='=rsh  '
-	fi
-	AC_MSG_RESULT()
-	AC_MSG_RESULT([    **********************************************************************])
-	AC_MSG_RESULT([    *   As of v2.6.0, the default remote shell is ssh instead of rsh!!   *])
-	AC_MSG_RESULT([    *   To use previous default of $rmsh1 ./configure --with-rsh$rmsh2   *])
-	AC_MSG_RESULT([    **********************************************************************])
-fi
-
-AC_MSG_RESULT()
-AC_MSG_RESULT([    rsync ${RSYNC_VERSION} configuration successful])
-AC_MSG_RESULT()

connection.c

@@ -1,49 +1,47 @@
-/* 
-   Copyright (C) Andrew Tridgell 1998
-   
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
+/*
+ * Support the max connections option.
+ *
+ * Copyright (C) 1998 Andrew Tridgell
+ * Copyright (C) 2006-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
-/* support the max connections option */
 #include "rsync.h"
 
-
-/****************************************************************************
-simple routine to do connection counting
-****************************************************************************/
-int claim_connection(char *fname,int max_connections)
+/* A simple routine to do connection counting.  This returns 1 on success
+ * and 0 on failure, with errno also being set if the open() failed (errno
+ * will be 0 if the lock request failed). */
+int claim_connection(char *fname, int max_connections)
 {
 	int fd, i;
 
-	if (max_connections <= 0)
+	if (max_connections == 0)
 		return 1;
 
-	fd = open(fname,O_RDWR|O_CREAT, 0600);
-
-	if (fd == -1) {
+	if ((fd = open(fname, O_RDWR|O_CREAT, 0600)) < 0)
 		return 0;
+
+	/* Find a free spot. */
+	for (i = 0; i < max_connections; i++) {
+		if (lock_range(fd, i*4, 4))
+			return 1;
 	}
 
-	/* find a free spot */
-	for (i=0;i<max_connections;i++) {
-		if (lock_range(fd, i*4, 4)) return 1;
-	}		
-
-	/* only interested in open failures */
-	errno = 0;
-
 	close(fd);
+
+	/* A lock failure needs to return an errno of 0. */
+	errno = 0;
 	return 0;
 }

csprotocol.txt

@@ -3,21 +3,22 @@ basically a summary of clientserver.c and authenticate.c.
 
  -- Martin Pool <mbp@samba.org>
 
-$Id$
-
-
-
 
 This is the protocol used for rsync --daemon; i.e. connections to port
 873 rather than invocations over a remote shell.
 
 When the server accepts a connection, it prints a greeting
 
-  @RSYNCD: <version>
+  @RSYNCD: <version>.<subprotocol>
 
-where <version> is the numeric version; currently 24.  It follows this
-with a free text message-of-the-day.  It expects to see a similar
-greeting back from the client.
+where <version> is the numeric version (see PROTOCOL_VERSION in rsync.h)
+'.' is a literal period, and <subprotocol> is the numeric subprotocol
+version (see SUBPROTOCOL_VERSION -- it will be 0 for final releases).
+Protocols prior to 30 only output <version> alone.  The daemon expects
+to see a similar greeting back from the client.  For protocols prior to
+30, an absent ".<subprotocol>" value is assumed to be 0.  For protocol
+30, an absent value is a fatal error.  The daemon then follows this line
+with a free-format text message-of-the-day (if any is defined).
 
 The server is now in the connected state.  The client can either send
 the command
@@ -75,8 +76,13 @@ stay tuned (or write it yourself!).
 ------------
 Protocol version changes
 
-25	 (2001-08-20, 2.4.7pre2) 
+30	(2007-10-04, 3.0.0pre1)
 
-	 Send an explicit "@RSYNC EXIT" command at the end of the
-	 module listing.  We never intentionally end the transmission
-	 by just closing the socket anymore.
+	The use of a ".<subprotocol>" number was added to
+	@RSYNCD: <version>.<subprotocol>
+
+25	(2001-08-20, 2.4.7pre2) 
+
+	Send an explicit "@RSYNC EXIT" command at the end of the
+	module listing.  We never intentionally end the transmission
+	by just closing the socket anymore.

define-from-md.awk

@@ -0,0 +1,41 @@
+#!/usr/bin/awk -f
+
+# The caller must pass args: -v hfile=NAME rsync.1.md
+
+BEGIN {
+    heading = "/* DO NOT EDIT THIS FILE!  It is auto-generated from a list of values in " ARGV[1] "! */"
+    if (hfile ~ /compress/) {
+	define = "#define DEFAULT_DONT_COMPRESS"
+	prefix = "*."
+    } else {
+	define = "#define DEFAULT_CVSIGNORE"
+	prefix = ""
+    }
+    value_list = ""
+}
+
+/^    > [^ ]+$/ {
+    gsub(/`/, "")
+    if (value_list != "") value_list = value_list " "
+    value_list = value_list prefix $2
+    next
+}
+
+value_list ~ /\.gz / && hfile ~ /compress/ {
+    exit
+}
+
+value_list ~ /SCCS / && hfile ~ /cvsignore/ {
+    exit
+}
+
+value_list = ""
+
+END {
+    if (value_list != "")
+	print heading "\n\n" define " \"" value_list "\"" > hfile
+    else {
+	print "Failed to find a value list in " ARGV[1] " for " hfile
+	exit 1
+    }
+}

delete.c

@@ -0,0 +1,240 @@
+/*
+ * Deletion routines used in rsync.
+ *
+ * Copyright (C) 1996-2000 Andrew Tridgell
+ * Copyright (C) 1996 Paul Mackerras
+ * Copyright (C) 2002 Martin Pool <mbp@samba.org>
+ * Copyright (C) 2003-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
+
+#include "rsync.h"
+
+extern int am_root;
+extern int make_backups;
+extern int max_delete;
+extern char *backup_dir;
+extern char *backup_suffix;
+extern int backup_suffix_len;
+extern struct stats stats;
+
+int ignore_perishable = 0;
+int non_perishable_cnt = 0;
+int skipped_deletes = 0;
+
+static inline int is_backup_file(char *fn)
+{
+	int k = strlen(fn) - backup_suffix_len;
+	return k > 0 && strcmp(fn+k, backup_suffix) == 0;
+}
+
+/* The directory is about to be deleted: if DEL_RECURSE is given, delete all
+ * its contents, otherwise just checks for content.  Returns DR_SUCCESS or
+ * DR_NOT_EMPTY.  Note that fname must point to a MAXPATHLEN buffer!  (The
+ * buffer is used for recursion, but returned unchanged.)
+ */
+static enum delret delete_dir_contents(char *fname, uint16 flags)
+{
+	struct file_list *dirlist;
+	enum delret ret;
+	unsigned remainder;
+	void *save_filters;
+	int j, dlen;
+	char *p;
+
+	if (DEBUG_GTE(DEL, 3)) {
+		rprintf(FINFO, "delete_dir_contents(%s) flags=%d\n",
+			fname, flags);
+	}
+
+	dlen = strlen(fname);
+	save_filters = push_local_filters(fname, dlen);
+
+	non_perishable_cnt = 0;
+	dirlist = get_dirlist(fname, dlen, 0);
+	ret = non_perishable_cnt ? DR_NOT_EMPTY : DR_SUCCESS;
+
+	if (!dirlist->used)
+		goto done;
+
+	if (!(flags & DEL_RECURSE)) {
+		ret = DR_NOT_EMPTY;
+		goto done;
+	}
+
+	p = fname + dlen;
+	if (dlen != 1 || *fname != '/')
+		*p++ = '/';
+	remainder = MAXPATHLEN - (p - fname);
+
+	/* We do our own recursion, so make delete_item() non-recursive. */
+	flags = (flags & ~(DEL_RECURSE|DEL_MAKE_ROOM|DEL_NO_UID_WRITE))
+	      | DEL_DIR_IS_EMPTY;
+
+	for (j = dirlist->used; j--; ) {
+		struct file_struct *fp = dirlist->files[j];
+
+		if (fp->flags & FLAG_MOUNT_DIR && S_ISDIR(fp->mode)) {
+			if (DEBUG_GTE(DEL, 1)) {
+				rprintf(FINFO,
+					"mount point, %s, pins parent directory\n",
+					f_name(fp, NULL));
+			}
+			ret = DR_NOT_EMPTY;
+			continue;
+		}
+
+		strlcpy(p, fp->basename, remainder);
+		if (!(fp->mode & S_IWUSR) && !am_root && fp->flags & FLAG_OWNED_BY_US)
+			do_chmod(fname, fp->mode | S_IWUSR);
+		/* Save stack by recursing to ourself directly. */
+		if (S_ISDIR(fp->mode)) {
+			if (delete_dir_contents(fname, flags | DEL_RECURSE) != DR_SUCCESS)
+				ret = DR_NOT_EMPTY;
+		}
+		if (delete_item(fname, fp->mode, flags) != DR_SUCCESS)
+			ret = DR_NOT_EMPTY;
+	}
+
+	fname[dlen] = '\0';
+
+  done:
+	flist_free(dirlist);
+	pop_local_filters(save_filters);
+
+	if (ret == DR_NOT_EMPTY) {
+		rprintf(FINFO, "cannot delete non-empty directory: %s\n",
+			fname);
+	}
+	return ret;
+}
+
+/* Delete a file or directory.  If DEL_RECURSE is set in the flags, this will
+ * delete recursively.
+ *
+ * Note that fbuf must point to a MAXPATHLEN buffer if the mode indicates it's
+ * a directory! (The buffer is used for recursion, but returned unchanged.)
+ */
+enum delret delete_item(char *fbuf, uint16 mode, uint16 flags)
+{
+	enum delret ret;
+	char *what;
+	int ok;
+
+	if (DEBUG_GTE(DEL, 2)) {
+		rprintf(FINFO, "delete_item(%s) mode=%o flags=%d\n",
+			fbuf, (int)mode, (int)flags);
+	}
+
+	if (flags & DEL_NO_UID_WRITE)
+		do_chmod(fbuf, mode | S_IWUSR);
+
+	if (S_ISDIR(mode) && !(flags & DEL_DIR_IS_EMPTY)) {
+		/* This only happens on the first call to delete_item() since
+		 * delete_dir_contents() always calls us w/DEL_DIR_IS_EMPTY. */
+		ignore_perishable = 1;
+		/* If DEL_RECURSE is not set, this just reports emptiness. */
+		ret = delete_dir_contents(fbuf, flags);
+		ignore_perishable = 0;
+		if (ret == DR_NOT_EMPTY || ret == DR_AT_LIMIT)
+			goto check_ret;
+		/* OK: try to delete the directory. */
+	}
+
+	if (!(flags & DEL_MAKE_ROOM) && max_delete >= 0 && stats.deleted_files >= max_delete) {
+		skipped_deletes++;
+		return DR_AT_LIMIT;
+	}
+
+	if (S_ISDIR(mode)) {
+		what = "rmdir";
+		ok = do_rmdir(fbuf) == 0;
+	} else {
+		if (make_backups > 0 && !(flags & DEL_FOR_BACKUP) && (backup_dir || !is_backup_file(fbuf))) {
+			what = "make_backup";
+			ok = make_backup(fbuf, True);
+			if (ok == 2) {
+				what = "unlink";
+				ok = robust_unlink(fbuf) == 0;
+			}
+		} else {
+			what = "unlink";
+			ok = robust_unlink(fbuf) == 0;
+		}
+	}
+
+	if (ok) {
+		if (!(flags & DEL_MAKE_ROOM)) {
+			log_delete(fbuf, mode);
+			stats.deleted_files++;
+			if (S_ISREG(mode)) {
+				/* Nothing more to count */
+			} else if (S_ISDIR(mode))
+				stats.deleted_dirs++;
+#ifdef SUPPORT_LINKS
+			else if (S_ISLNK(mode))
+				stats.deleted_symlinks++;
+#endif
+			else if (IS_DEVICE(mode))
+				stats.deleted_symlinks++;
+			else
+				stats.deleted_specials++;
+		}
+		ret = DR_SUCCESS;
+	} else {
+		if (S_ISDIR(mode) && errno == ENOTEMPTY) {
+			rprintf(FINFO, "cannot delete non-empty directory: %s\n",
+				fbuf);
+			ret = DR_NOT_EMPTY;
+		} else if (errno != ENOENT) {
+			rsyserr(FERROR_XFER, errno, "delete_file: %s(%s) failed",
+				what, fbuf);
+			ret = DR_FAILURE;
+		} else
+			ret = DR_SUCCESS;
+	}
+
+  check_ret:
+	if (ret != DR_SUCCESS && flags & DEL_MAKE_ROOM) {
+		const char *desc;
+		switch (flags & DEL_MAKE_ROOM) {
+		case DEL_FOR_FILE: desc = "regular file"; break;
+		case DEL_FOR_DIR: desc = "directory"; break;
+		case DEL_FOR_SYMLINK: desc = "symlink"; break;
+		case DEL_FOR_DEVICE: desc = "device file"; break;
+		case DEL_FOR_SPECIAL: desc = "special file"; break;
+		default: exit_cleanup(RERR_UNSUPPORTED); /* IMPOSSIBLE */
+		}
+		rprintf(FERROR_XFER, "could not make way for %s %s: %s\n",
+			flags & DEL_FOR_BACKUP ? "backup" : "new",
+			desc, fbuf);
+	}
+	return ret;
+}
+
+uint16 get_del_for_flag(uint16 mode)
+{
+	if (S_ISREG(mode))
+		return DEL_FOR_FILE;
+	if (S_ISDIR(mode))
+		return DEL_FOR_DIR;
+	if (S_ISLNK(mode))
+		return DEL_FOR_SYMLINK;
+	if (IS_DEVICE(mode))
+		return DEL_FOR_DEVICE;
+	if (IS_SPECIAL(mode))
+		return DEL_FOR_SPECIAL;
+	exit_cleanup(RERR_UNSUPPORTED); /* IMPOSSIBLE */
+}

doc/.cvsignore

@@ -1,2 +0,0 @@
-rsync.pdf
-rsync.ps

doc/rsync.sgml

@@ -126,7 +126,7 @@
       
       <para><option>-P</option>
 	Display a progress indicator while files are transferred.  This should
-	normally be ommitted if rsync is not run on a terminal.
+	normally be omitted if rsync is not run on a terminal.
       </para>
     </section>
 
@@ -348,4 +348,4 @@ running rsync giving the network directory.
     
     <para><ulink url="http://www.ccp14.ac.uk/ccp14admin/rsync/"></ulink></para>
   </appendix>
-</book>
+</book>

errcode.h

@@ -1,42 +1,43 @@
-/* -*- c-file-style: "linux"; -*-
-   
-   Copyright (C) 1998-2000 by Andrew Tridgell
-   
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
-
 /*
- * error codes returned by rsync.  If you change these, please also update the
- * string mappings in log.c and the EXIT VALUES in rsync.yo
+ * Error codes returned by rsync.
+ *
+ * Copyright (C) 1998-2000 Andrew Tridgell
+ * Copyright (C) 2003-2019 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
  */
 
+/* If you change these, please also update the string mappings in log.c and
+ * the EXIT VALUES in rsync.yo. */
+
 #define RERR_OK         0
 #define RERR_SYNTAX     1       /* syntax or usage error */
 #define RERR_PROTOCOL   2       /* protocol incompatibility */
 #define RERR_FILESELECT 3       /* errors selecting input/output files, dirs */
 #define RERR_UNSUPPORTED 4      /* requested action not supported */
 #define RERR_STARTCLIENT 5      /* error starting client-server protocol */
-#define RERR_LOG_FAILURE 6      /* daemon unable to append to log-file */
 
 #define RERR_SOCKETIO   10      /* error in socket IO */
 #define RERR_FILEIO     11      /* error in file IO */
 #define RERR_STREAMIO   12      /* error in rsync protocol data stream */
 #define RERR_MESSAGEIO  13      /* errors with program diagnostics */
 #define RERR_IPC        14      /* error in IPC code */
+#define RERR_CRASHED    15      /* sibling crashed */
+#define RERR_TERMINATED 16      /* sibling terminated abnormally */
 
-#define RERR_SIGNAL     20      /* status returned when sent SIGUSR1, SIGINT */
+#define RERR_SIGNAL1    19      /* status returned when sent SIGUSR1 */
+#define RERR_SIGNAL     20      /* status returned when sent SIGINT, SIGTERM, SIGHUP */
 #define RERR_WAITCHILD  21      /* some error returned by waitpid() */
 #define RERR_MALLOC     22      /* error allocating core memory buffers */
 #define RERR_PARTIAL    23      /* partial transfer */
@@ -44,6 +45,7 @@
 #define RERR_DEL_LIMIT  25      /* skipped some deletes due to --max-delete */
 
 #define RERR_TIMEOUT    30      /* timeout in data send/receive */
+#define RERR_CONTIMEOUT 35      /* timeout waiting for daemon connection */
 
 /* Although it doesn't seem to be specified anywhere,
  * ssh and the shell seem to return these values:

fileio.c

@@ -1,76 +1,123 @@
 /*
-   Copyright (C) Andrew Tridgell 1998
-   Copyright (C) 2002 by Martin Pool
+ * File IO utilities used in rsync.
+ *
+ * Copyright (C) 1998 Andrew Tridgell
+ * Copyright (C) 2002 Martin Pool
+ * Copyright (C) 2004-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
-
-/*
-  File IO utilities used in rsync
-  */
 #include "rsync.h"
+#include "inums.h"
+
+#ifndef ENODATA
+#define ENODATA EAGAIN
+#endif
+
+/* We want all reads to be aligned on 1K boundaries. */
+#define ALIGN_BOUNDARY 1024
+/* How far past the boundary is an offset? */
+#define ALIGNED_OVERSHOOT(oft) ((oft) & (ALIGN_BOUNDARY-1))
+/* Round up a length to the next boundary */
+#define ALIGNED_LENGTH(len) ((((len) - 1) | (ALIGN_BOUNDARY-1)) + 1)
 
 extern int sparse_files;
 
-static char last_byte;
-static int last_sparse;
+OFF_T preallocated_len = 0;
 
-int sparse_end(int f)
+static OFF_T sparse_seek = 0;
+static OFF_T sparse_past_write = 0;
+
+int sparse_end(int f, OFF_T size)
 {
-	if (last_sparse) {
-		do_lseek(f,-1,SEEK_CUR);
-		return (write(f,&last_byte,1) == 1 ? 0 : -1);
+	int ret;
+
+	sparse_past_write = 0;
+
+	if (!sparse_seek)
+		return 0;
+
+#ifdef HAVE_FTRUNCATE
+	ret = do_ftruncate(f, size);
+#else
+	if (do_lseek(f, sparse_seek-1, SEEK_CUR) != size-1)
+		ret = -1;
+	else {
+		do {
+			ret = write(f, "", 1);
+		} while (ret < 0 && errno == EINTR);
+
+		ret = ret <= 0 ? -1 : 0;
 	}
-	last_sparse = 0;
-	return 0;
+#endif
+
+	sparse_seek = 0;
+
+	return ret;
 }
 
-
-static int write_sparse(int f,char *buf,size_t len)
+/* Note that the offset is just the caller letting us know where
+ * the current file position is in the file. The use_seek arg tells
+ * us that we should seek over matching data instead of writing it. */
+static int write_sparse(int f, int use_seek, OFF_T offset, const char *buf, int len)
 {
-	size_t l1=0, l2=0;
+	int l1 = 0, l2 = 0;
 	int ret;
 
 	for (l1 = 0; l1 < len && buf[l1] == 0; l1++) {}
 	for (l2 = 0; l2 < len-l1 && buf[len-(l2+1)] == 0; l2++) {}
 
-	last_byte = buf[len-1];
-
-	if (l1 == len || l2 > 0)
-		last_sparse=1;
-
-	if (l1 > 0) {
-		do_lseek(f,l1,SEEK_CUR);
-	}
+	sparse_seek += l1;
 
 	if (l1 == len)
 		return len;
 
-	ret = write(f, buf + l1, len - (l1+l2));
-	if (ret == -1 || ret == 0)
-		return ret;
-	else if (ret != (int) (len - (l1+l2)))
-		return (l1+ret);
+	if (sparse_seek) {
+		if (sparse_past_write >= preallocated_len) {
+			if (do_lseek(f, sparse_seek, SEEK_CUR) < 0)
+				return -1;
+		} else if (do_punch_hole(f, sparse_past_write, sparse_seek) < 0) {
+			sparse_seek = 0;
+			return -1;
+		}
+	}
+	sparse_seek = l2;
+	sparse_past_write = offset + len - l2;
 
-	if (l2 > 0)
-		do_lseek(f,l2,SEEK_CUR);
+	if (use_seek) {
+		/* The in-place data already matches. */
+		if (do_lseek(f, len - (l1+l2), SEEK_CUR) < 0)
+			return -1;
+		return len;
+	}
+
+	while ((ret = write(f, buf + l1, len - (l1+l2))) <= 0) {
+		if (ret < 0 && errno == EINTR)
+			continue;
+		sparse_seek = 0;
+		return ret;
+	}
+
+	if (ret != (int)(len - (l1+l2))) {
+		sparse_seek = 0;
+		return l1+ret;
+	}
 
 	return len;
 }
 
-
 static char *wf_writeBuf;
 static size_t wf_writeBufSize;
 static size_t wf_writeBufCnt;
@@ -92,29 +139,26 @@ int flush_write_file(int f)
 	return ret;
 }
 
-
-/*
- * write_file does not allow incomplete writes.  It loops internally
- * until len bytes are written or errno is set.
- */
-int write_file(int f,char *buf,size_t len)
+/* write_file does not allow incomplete writes.  It loops internally
+ * until len bytes are written or errno is set.  Note that use_seek and
+ * offset are only used in sparse processing (see write_sparse()). */
+int write_file(int f, int use_seek, OFF_T offset, const char *buf, int len)
 {
 	int ret = 0;
 
 	while (len > 0) {
 		int r1;
-		if (sparse_files) {
+		if (sparse_files > 0) {
 			int len1 = MIN(len, SPARSE_WRITE_SIZE);
-			r1 = write_sparse(f, buf, len1);
+			r1 = write_sparse(f, use_seek, offset, buf, len1);
+			offset += r1;
 		} else {
 			if (!wf_writeBuf) {
 				wf_writeBufSize = WRITE_SIZE * 8;
 				wf_writeBufCnt  = 0;
 				wf_writeBuf = new_array(char, wf_writeBufSize);
-				if (!wf_writeBuf)
-					out_of_memory("write_file");
 			}
-			r1 = MIN(len, wf_writeBufSize - wf_writeBufCnt);
+			r1 = (int)MIN((size_t)len, wf_writeBufSize - wf_writeBufCnt);
 			if (r1) {
 				memcpy(wf_writeBuf + wf_writeBufCnt, buf, r1);
 				wf_writeBufCnt += r1;
@@ -138,26 +182,47 @@ int write_file(int f,char *buf,size_t len)
 	return ret;
 }
 
+/* An in-place update found identical data at an identical location. We either
+ * just seek past it, or (for an in-place sparse update), we give the data to
+ * the sparse processor with the use_seek flag set. */
+int skip_matched(int fd, OFF_T offset, const char *buf, int len)
+{
+	OFF_T pos;
+
+	if (sparse_files > 0) {
+		if (write_file(fd, 1, offset, buf, len) != len)
+			return -1;
+		return 0;
+	}
+
+	if (flush_write_file(fd) < 0)
+		return -1;
+
+	if ((pos = do_lseek(fd, len, SEEK_CUR)) != offset + len) {
+		rsyserr(FERROR_XFER, errno, "lseek returned %s, not %s",
+			big_num(pos), big_num(offset));
+		return -1;
+	}
+
+	return 0;
+}
 
 /* This provides functionality somewhat similar to mmap() but using read().
  * It gives sliding window access to a file.  mmap() is not used because of
  * the possibility of another program (such as a mailer) truncating the
  * file thus giving us a SIGBUS. */
-struct map_struct *map_file(int fd, OFF_T len, int32 read_size,
-			    int32 blk_size)
+struct map_struct *map_file(int fd, OFF_T len, int32 read_size, int32 blk_size)
 {
 	struct map_struct *map;
 
-	if (!(map = new(struct map_struct)))
-		out_of_memory("map_file");
+	map = new0(struct map_struct);
 
 	if (blk_size && (read_size % blk_size))
 		read_size += blk_size - (read_size % blk_size);
 
-	memset(map, 0, sizeof map[0]);
 	map->fd = fd;
 	map->file_size = len;
-	map->def_window_size = read_size;
+	map->def_window_size = ALIGNED_LENGTH(read_size);
 
 	return map;
 }
@@ -166,9 +231,8 @@ struct map_struct *map_file(int fd, OFF_T len, int32 read_size,
 /* slide the read window in the file */
 char *map_ptr(struct map_struct *map, OFF_T offset, int32 len)
 {
-	int32 nread;
 	OFF_T window_start, read_start;
-	int32 window_size, read_size, read_offset;
+	int32 window_size, read_size, read_offset, align_fudge;
 
 	if (len == 0)
 		return NULL;
@@ -183,28 +247,25 @@ char *map_ptr(struct map_struct *map, OFF_T offset, int32 len)
 		return map->p + (offset - map->p_offset);
 
 	/* nope, we are going to have to do a read. Work out our desired window */
-	window_start = offset;
+	align_fudge = (int32)ALIGNED_OVERSHOOT(offset);
+	window_start = offset - align_fudge;
 	window_size = map->def_window_size;
 	if (window_start + window_size > map->file_size)
-		window_size = map->file_size - window_start;
-	if (len > window_size)
-		window_size = len;
+		window_size = (int32)(map->file_size - window_start);
+	if (window_size < len + align_fudge)
+		window_size = ALIGNED_LENGTH(len + align_fudge);
 
 	/* make sure we have allocated enough memory for the window */
 	if (window_size > map->p_size) {
 		map->p = realloc_array(map->p, char, window_size);
-		if (!map->p)
-			out_of_memory("map_ptr");
 		map->p_size = window_size;
 	}
 
-	/* Now try to avoid re-reading any bytes by reusing any bytes
-	 * from the previous buffer. */
-	if (window_start >= map->p_offset &&
-	    window_start < map->p_offset + map->p_len &&
-	    window_start + window_size >= map->p_offset + map->p_len) {
+	/* Now try to avoid re-reading any bytes by reusing any bytes from the previous buffer. */
+	if (window_start >= map->p_offset && window_start < map->p_offset + map->p_len
+	 && window_start + window_size >= map->p_offset + map->p_len) {
 		read_start = map->p_offset + map->p_len;
-		read_offset = read_start - window_start;
+		read_offset = (int32)(read_start - window_start);
 		read_size = window_size - read_offset;
 		memmove(map->p, map->p + (map->p_len - read_offset), read_offset);
 	} else {
@@ -217,37 +278,37 @@ char *map_ptr(struct map_struct *map, OFF_T offset, int32 len)
 		rprintf(FERROR, "invalid read_size of %ld in map_ptr\n",
 			(long)read_size);
 		exit_cleanup(RERR_FILEIO);
-	} else {
-		if (map->p_fd_offset != read_start) {
-			OFF_T ret = do_lseek(map->fd, read_start, SEEK_SET);
-			if (ret != read_start) {
-				rsyserr(FERROR, errno,
-					"lseek returned %.0f, not %.0f",
-					(double)ret, (double)read_start);
-				exit_cleanup(RERR_FILEIO);
-			}
-			map->p_fd_offset = read_start;
-		}
-
-		if ((nread=read(map->fd,map->p + read_offset,read_size)) != read_size) {
-			if (nread < 0) {
-				nread = 0;
-				if (!map->status)
-					map->status = errno;
-			}
-			/* the best we can do is zero the buffer - the file
-			   has changed mid transfer! */
-			memset(map->p+read_offset+nread, 0, read_size - nread);
-		}
-		map->p_fd_offset += nread;
 	}
 
+	if (map->p_fd_offset != read_start) {
+		OFF_T ret = do_lseek(map->fd, read_start, SEEK_SET);
+		if (ret != read_start) {
+			rsyserr(FERROR, errno, "lseek returned %s, not %s",
+				big_num(ret), big_num(read_start));
+			exit_cleanup(RERR_FILEIO);
+		}
+		map->p_fd_offset = read_start;
+	}
 	map->p_offset = window_start;
 	map->p_len = window_size;
 
-	return map->p;
-}
+	while (read_size > 0) {
+		int32 nread = read(map->fd, map->p + read_offset, read_size);
+		if (nread <= 0) {
+			if (!map->status)
+				map->status = nread ? errno : ENODATA;
+			/* The best we can do is zero the buffer -- the file
+			 * has changed mid transfer! */
+			memset(map->p + read_offset, 0, read_size);
+			break;
+		}
+		map->p_fd_offset += nread;
+		read_offset += nread;
+		read_size -= nread;
+	}
 
+	return map->p + align_fudge;
+}
 
 int unmap_file(struct map_struct *map)
 {
@@ -258,7 +319,9 @@ int unmap_file(struct map_struct *map)
 		map->p = NULL;
 	}
 	ret = map->status;
-	memset(map, 0, sizeof map[0]);
+#if 0 /* I don't think we really need this. */
+	force_memzero(map, sizeof map[0]);
+#endif
 	free(map);
 
 	return ret;

getfsdev.c

@@ -15,8 +15,7 @@
 			fprintf(stderr, "Unable to stat `%s'\n", *argv);
 			exit(1);
 		}
-		printf("%ld/%ld\n", (long)major(st.st_dev),
-				    (long)minor(st.st_dev));
+		printf("%ld/%ld\n", (long)major(st.st_dev), (long)minor(st.st_dev));
 	}
 
 	return 0;

getgroups.c

@@ -1,33 +1,26 @@
 /*
- * Copyright (C) 2002 by Martin Pool
- * 
+ * Print out the gids of all groups for the current user.  This is like
+ * `id -G` on Linux, but it's too hard to find a portable equivalent.
+ *
+ * Copyright (C) 2002 Martin Pool
+ * Copyright (C) 2003-2020 Wayne Davison
+ *
  * 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.
- * 
+ * it under the terms of the GNU General Public License version 3 as
+ * published by the Free Software Foundation.
+ *
  * 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., 675 Mass Ave, Cambridge, MA 02139, USA.
- */
-
-/**
- * @file getgroups.c
  *
- * Print out the gids of all groups for the current user.  This is
- * like `id -G` on Linux, but it's too hard to find a portable
- * equivalent.
- **/
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, visit the http://fsf.org website.
+ */
 
 #include "rsync.h"
 
-int
-main(UNUSED(int argc), UNUSED(char *argv[]))
+ int main(UNUSED(int argc), UNUSED(char *argv[]))
 {
 	int n, i;
 	gid_t *list;
@@ -63,6 +56,6 @@ main(UNUSED(int argc), UNUSED(char *argv[]))
 	if (!gid_in_list)
 		printf("%lu", (unsigned long)gid);
 	printf("\n");
-		
+
 	return 0;
 }

hashtable.c

@@ -0,0 +1,498 @@
+/*
+ * Routines to provide a memory-efficient hashtable.
+ *
+ * Copyright (C) 2007-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
+
+#include "rsync.h"
+
+#define HASH_LOAD_LIMIT(size) ((size)*3/4)
+
+struct hashtable *hashtable_create(int size, int key64)
+{
+	int req = size;
+	struct hashtable *tbl;
+	int node_size = key64 ? sizeof (struct ht_int64_node)
+			      : sizeof (struct ht_int32_node);
+
+	/* Pick a power of 2 that can hold the requested size. */
+	if (size & (size-1) || size < 16) {
+		size = 16;
+		while (size < req)
+			size *= 2;
+	}
+
+	tbl = new(struct hashtable);
+	tbl->nodes = new_array0(char, size * node_size);
+	tbl->size = size;
+	tbl->entries = 0;
+	tbl->node_size = node_size;
+	tbl->key64 = key64 ? 1 : 0;
+
+	if (DEBUG_GTE(HASH, 1)) {
+		char buf[32];
+		if (req != size)
+			snprintf(buf, sizeof buf, "req: %d, ", req);
+		else
+			*buf = '\0';
+		rprintf(FINFO, "[%s] created hashtable %lx (%ssize: %d, keys: %d-bit)\n",
+			who_am_i(), (long)tbl, buf, size, key64 ? 64 : 32);
+	}
+
+	return tbl;
+}
+
+void hashtable_destroy(struct hashtable *tbl)
+{
+	if (DEBUG_GTE(HASH, 1)) {
+		rprintf(FINFO, "[%s] destroyed hashtable %lx (size: %d, keys: %d-bit)\n",
+			who_am_i(), (long)tbl, tbl->size, tbl->key64 ? 64 : 32);
+	}
+	free(tbl->nodes);
+	free(tbl);
+}
+
+/* Returns the node that holds the indicated key if it exists. When it does not
+ * exist, it returns either NULL (when data_when_new is NULL), or it returns a
+ * new node with its node->data set to the indicated value.
+ *
+ * If your code doesn't know the data value for a new node in advance (usually
+ * because it doesn't know if a node is new or not) you should pass in a unique
+ * (non-0) value that you can use to check if the returned node is new. You can
+ * then overwrite the data with any value you want (even 0) since it only needs
+ * to be different than whatever data_when_new value you use later on.
+ *
+ * This return is a void* just because it might be pointing at a ht_int32_node
+ * or a ht_int64_node, and that makes the caller's assignment a little easier. */
+void *hashtable_find(struct hashtable *tbl, int64 key, void *data_when_new)
+{
+	int key64 = tbl->key64;
+	struct ht_int32_node *node;
+	uint32 ndx;
+
+	if (key64 ? key == 0 : (int32)key == 0) {
+		rprintf(FERROR, "Internal hashtable error: illegal key supplied!\n");
+		exit_cleanup(RERR_MESSAGEIO);
+	}
+
+	if (data_when_new && tbl->entries > HASH_LOAD_LIMIT(tbl->size)) {
+		void *old_nodes = tbl->nodes;
+		int size = tbl->size * 2;
+		int i;
+
+		tbl->nodes = new_array0(char, size * tbl->node_size);
+		tbl->size = size;
+		tbl->entries = 0;
+
+		if (DEBUG_GTE(HASH, 1)) {
+			rprintf(FINFO, "[%s] growing hashtable %lx (size: %d, keys: %d-bit)\n",
+				who_am_i(), (long)tbl, size, key64 ? 64 : 32);
+		}
+
+		for (i = size / 2; i-- > 0; ) {
+			struct ht_int32_node *move_node = HT_NODE(tbl, old_nodes, i);
+			int64 move_key = HT_KEY(move_node, key64);
+			if (move_key == 0)
+				continue;
+			if (move_node->data)
+				hashtable_find(tbl, move_key, move_node->data);
+			else {
+				node = hashtable_find(tbl, move_key, "");
+				node->data = 0;
+			}
+		}
+
+		free(old_nodes);
+	}
+
+	if (!key64) {
+		/* Based on Jenkins One-at-a-time hash. */
+		uchar buf[4], *keyp = buf;
+		int i;
+
+		SIVALu(buf, 0, key);
+		for (ndx = 0, i = 0; i < 4; i++) {
+			ndx += keyp[i];
+			ndx += (ndx << 10);
+			ndx ^= (ndx >> 6);
+		}
+		ndx += (ndx << 3);
+		ndx ^= (ndx >> 11);
+		ndx += (ndx << 15);
+	} else {
+		/* Based on Jenkins hashword() from lookup3.c. */
+		uint32 a, b, c;
+
+		/* Set up the internal state */
+		a = b = c = 0xdeadbeef + (8 << 2);
+
+#define rot(x,k) (((x)<<(k)) ^ ((x)>>(32-(k))))
+#if SIZEOF_INT64 >= 8
+		b += (uint32)(key >> 32);
+#endif
+		a += (uint32)key;
+		c ^= b; c -= rot(b, 14);
+		a ^= c; a -= rot(c, 11);
+		b ^= a; b -= rot(a, 25);
+		c ^= b; c -= rot(b, 16);
+		a ^= c; a -= rot(c, 4);
+		b ^= a; b -= rot(a, 14);
+		c ^= b; c -= rot(b, 24);
+#undef rot
+		ndx = c;
+	}
+
+	/* If it already exists, return the node.  If we're not
+	 * allocating, return NULL if the key is not found. */
+	while (1) {
+		int64 nkey;
+
+		ndx &= tbl->size - 1;
+		node = HT_NODE(tbl, tbl->nodes, ndx);
+		nkey = HT_KEY(node, key64);
+
+		if (nkey == key)
+			return node;
+		if (nkey == 0) {
+			if (!data_when_new)
+				return NULL;
+			break;
+		}
+		ndx++;
+	}
+
+	/* Take over this empty spot and then return the node. */
+	if (key64)
+		((struct ht_int64_node*)node)->key = key;
+	else
+		node->key = (int32)key;
+	node->data = data_when_new;
+	tbl->entries++;
+	return node;
+}
+
+#ifndef WORDS_BIGENDIAN
+# define HASH_LITTLE_ENDIAN 1
+# define HASH_BIG_ENDIAN 0
+#else
+# define HASH_LITTLE_ENDIAN 0
+# define HASH_BIG_ENDIAN 1
+#endif
+
+/*
+ -------------------------------------------------------------------------------
+ lookup3.c, by Bob Jenkins, May 2006, Public Domain.
+
+ These are functions for producing 32-bit hashes for hash table lookup.
+ hash_word(), hashlittle(), hashlittle2(), hashbig(), mix(), and final()
+ are externally useful functions.  Routines to test the hash are included
+ if SELF_TEST is defined.  You can use this free for any purpose.  It's in
+ the public domain.  It has no warranty.
+
+ You probably want to use hashlittle().  hashlittle() and hashbig()
+ hash byte arrays.  hashlittle() is is faster than hashbig() on
+ little-endian machines.  Intel and AMD are little-endian machines.
+ On second thought, you probably want hashlittle2(), which is identical to
+ hashlittle() except it returns two 32-bit hashes for the price of one.
+ You could implement hashbig2() if you wanted but I haven't bothered here.
+
+ If you want to find a hash of, say, exactly 7 integers, do
+   a = i1;  b = i2;  c = i3;
+   mix(a,b,c);
+   a += i4; b += i5; c += i6;
+   mix(a,b,c);
+   a += i7;
+   final(a,b,c);
+ then use c as the hash value.  If you have a variable length array of
+ 4-byte integers to hash, use hash_word().  If you have a byte array (like
+ a character string), use hashlittle().  If you have several byte arrays, or
+ a mix of things, see the comments above hashlittle().
+
+ Why is this so big?  I read 12 bytes at a time into 3 4-byte integers,
+ then mix those integers.  This is fast (you can do a lot more thorough
+ mixing with 12*3 instructions on 3 integers than you can with 3 instructions
+ on 1 byte), but shoehorning those bytes into integers efficiently is messy.
+*/
+
+#define hashsize(n) ((uint32_t)1<<(n))
+#define hashmask(n) (hashsize(n)-1)
+#define rot(x,k) (((x)<<(k)) | ((x)>>(32-(k))))
+
+/*
+ -------------------------------------------------------------------------------
+ mix -- mix 3 32-bit values reversibly.
+
+ This is reversible, so any information in (a,b,c) before mix() is
+ still in (a,b,c) after mix().
+
+ If four pairs of (a,b,c) inputs are run through mix(), or through
+ mix() in reverse, there are at least 32 bits of the output that
+ are sometimes the same for one pair and different for another pair.
+ This was tested for:
+ * pairs that differed by one bit, by two bits, in any combination
+   of top bits of (a,b,c), or in any combination of bottom bits of
+   (a,b,c).
+ * "differ" is defined as +, -, ^, or ~^.  For + and -, I transformed
+   the output delta to a Gray code (a^(a>>1)) so a string of 1's (as
+   is commonly produced by subtraction) look like a single 1-bit
+   difference.
+ * the base values were pseudorandom, all zero but one bit set, or
+   all zero plus a counter that starts at zero.
+
+ Some k values for my "a-=c; a^=rot(c,k); c+=b;" arrangement that
+ satisfy this are
+     4  6  8 16 19  4
+     9 15  3 18 27 15
+    14  9  3  7 17  3
+ Well, "9 15 3 18 27 15" didn't quite get 32 bits diffing
+ for "differ" defined as + with a one-bit base and a two-bit delta.  I
+ used http://burtleburtle.net/bob/hash/avalanche.html to choose
+ the operations, constants, and arrangements of the variables.
+
+ This does not achieve avalanche.  There are input bits of (a,b,c)
+ that fail to affect some output bits of (a,b,c), especially of a.  The
+ most thoroughly mixed value is c, but it doesn't really even achieve
+ avalanche in c.
+
+ This allows some parallelism.  Read-after-writes are good at doubling
+ the number of bits affected, so the goal of mixing pulls in the opposite
+ direction as the goal of parallelism.  I did what I could.  Rotates
+ seem to cost as much as shifts on every machine I could lay my hands
+ on, and rotates are much kinder to the top and bottom bits, so I used
+ rotates.
+ -------------------------------------------------------------------------------
+*/
+#define mix(a,b,c) \
+{ \
+  a -= c;  a ^= rot(c, 4);  c += b; \
+  b -= a;  b ^= rot(a, 6);  a += c; \
+  c -= b;  c ^= rot(b, 8);  b += a; \
+  a -= c;  a ^= rot(c,16);  c += b; \
+  b -= a;  b ^= rot(a,19);  a += c; \
+  c -= b;  c ^= rot(b, 4);  b += a; \
+}
+
+/*
+ -------------------------------------------------------------------------------
+ final -- final mixing of 3 32-bit values (a,b,c) into c
+
+ Pairs of (a,b,c) values differing in only a few bits will usually
+ produce values of c that look totally different.  This was tested for
+ * pairs that differed by one bit, by two bits, in any combination
+   of top bits of (a,b,c), or in any combination of bottom bits of
+   (a,b,c).
+ * "differ" is defined as +, -, ^, or ~^.  For + and -, I transformed
+   the output delta to a Gray code (a^(a>>1)) so a string of 1's (as
+   is commonly produced by subtraction) look like a single 1-bit
+   difference.
+ * the base values were pseudorandom, all zero but one bit set, or
+   all zero plus a counter that starts at zero.
+
+ These constants passed:
+  14 11 25 16 4 14 24
+  12 14 25 16 4 14 24
+ and these came close:
+   4  8 15 26 3 22 24
+  10  8 15 26 3 22 24
+  11  8 15 26 3 22 24
+ -------------------------------------------------------------------------------
+*/
+#define final(a,b,c) \
+{ \
+  c ^= b; c -= rot(b,14); \
+  a ^= c; a -= rot(c,11); \
+  b ^= a; b -= rot(a,25); \
+  c ^= b; c -= rot(b,16); \
+  a ^= c; a -= rot(c,4);  \
+  b ^= a; b -= rot(a,14); \
+  c ^= b; c -= rot(b,24); \
+}
+
+
+/*
+ -------------------------------------------------------------------------------
+ hashlittle() -- hash a variable-length key into a 32-bit value
+   k       : the key (the unaligned variable-length array of bytes)
+   length  : the length of the key, counting by bytes
+   val2    : IN: can be any 4-byte value OUT: second 32 bit hash.
+ Returns a 32-bit value.  Every bit of the key affects every bit of
+ the return value.  Two keys differing by one or two bits will have
+ totally different hash values.  Note that the return value is better
+ mixed than val2, so use that first.
+
+ The best hash table sizes are powers of 2.  There is no need to do
+ mod a prime (mod is sooo slow!).  If you need less than 32 bits,
+ use a bitmask.  For example, if you need only 10 bits, do
+   h = (h & hashmask(10));
+ In which case, the hash table should have hashsize(10) elements.
+
+ If you are hashing n strings (uint8_t **)k, do it like this:
+   for (i=0, h=0; i<n; ++i) h = hashlittle( k[i], len[i], h);
+
+ By Bob Jenkins, 2006.  bob_jenkins@burtleburtle.net.  You may use this
+ code any way you wish, private, educational, or commercial.  It's free.
+
+ Use for hash table lookup, or anything where one collision in 2^^32 is
+ acceptable.  Do NOT use for cryptographic purposes.
+ -------------------------------------------------------------------------------
+*/
+
+uint32_t hashlittle(const void *key, size_t length)
+{
+  uint32_t a,b,c;                                          /* internal state */
+  union { const void *ptr; size_t i; } u;     /* needed for Mac Powerbook G4 */
+
+  /* Set up the internal state */
+  a = b = c = 0xdeadbeef + ((uint32_t)length);
+
+  u.ptr = key;
+  if (HASH_LITTLE_ENDIAN && ((u.i & 0x3) == 0)) {
+    const uint32_t *k = (const uint32_t *)key;         /* read 32-bit chunks */
+    const uint8_t  *k8;
+
+    /*------ all but last block: aligned reads and affect 32 bits of (a,b,c) */
+    while (length > 12)
+    {
+      a += k[0];
+      b += k[1];
+      c += k[2];
+      mix(a,b,c);
+      length -= 12;
+      k += 3;
+    }
+
+    /*----------------------------- handle the last (probably partial) block */
+    k8 = (const uint8_t *)k;
+    switch(length)
+    {
+    case 12: c+=k[2]; b+=k[1]; a+=k[0]; break;
+    case 11: c+=((uint32_t)k8[10])<<16;  /* fall through */
+    case 10: c+=((uint32_t)k8[9])<<8;    /* fall through */
+    case 9 : c+=k8[8];                   /* fall through */
+    case 8 : b+=k[1]; a+=k[0]; break;
+    case 7 : b+=((uint32_t)k8[6])<<16;   /* fall through */
+    case 6 : b+=((uint32_t)k8[5])<<8;    /* fall through */
+    case 5 : b+=k8[4];                   /* fall through */
+    case 4 : a+=k[0]; break;
+    case 3 : a+=((uint32_t)k8[2])<<16;   /* fall through */
+    case 2 : a+=((uint32_t)k8[1])<<8;    /* fall through */
+    case 1 : a+=k8[0]; break;
+    case 0 : return c;
+    }
+  } else if (HASH_LITTLE_ENDIAN && ((u.i & 0x1) == 0)) {
+    const uint16_t *k = (const uint16_t *)key;         /* read 16-bit chunks */
+    const uint8_t  *k8;
+
+    /*--------------- all but last block: aligned reads and different mixing */
+    while (length > 12)
+    {
+      a += k[0] + (((uint32_t)k[1])<<16);
+      b += k[2] + (((uint32_t)k[3])<<16);
+      c += k[4] + (((uint32_t)k[5])<<16);
+      mix(a,b,c);
+      length -= 12;
+      k += 6;
+    }
+
+    /*----------------------------- handle the last (probably partial) block */
+    k8 = (const uint8_t *)k;
+    switch(length)
+    {
+    case 12: c+=k[4]+(((uint32_t)k[5])<<16);
+             b+=k[2]+(((uint32_t)k[3])<<16);
+             a+=k[0]+(((uint32_t)k[1])<<16);
+             break;
+    case 11: c+=((uint32_t)k8[10])<<16;     /* fall through */
+    case 10: c+=k[4];
+             b+=k[2]+(((uint32_t)k[3])<<16);
+             a+=k[0]+(((uint32_t)k[1])<<16);
+             break;
+    case 9 : c+=k8[8];                      /* fall through */
+    case 8 : b+=k[2]+(((uint32_t)k[3])<<16);
+             a+=k[0]+(((uint32_t)k[1])<<16);
+             break;
+    case 7 : b+=((uint32_t)k8[6])<<16;      /* fall through */
+    case 6 : b+=k[2];
+             a+=k[0]+(((uint32_t)k[1])<<16);
+             break;
+    case 5 : b+=k8[4];                      /* fall through */
+    case 4 : a+=k[0]+(((uint32_t)k[1])<<16);
+             break;
+    case 3 : a+=((uint32_t)k8[2])<<16;      /* fall through */
+    case 2 : a+=k[0];
+             break;
+    case 1 : a+=k8[0];
+             break;
+    case 0 : return c;                     /* zero length requires no mixing */
+    }
+
+  } else {                        /* need to read the key one byte at a time */
+    const uint8_t *k = (const uint8_t *)key;
+
+    /*--------------- all but the last block: affect some 32 bits of (a,b,c) */
+    while (length > 12)
+    {
+      a += k[0];
+      a += ((uint32_t)k[1])<<8;
+      a += ((uint32_t)k[2])<<16;
+      a += ((uint32_t)k[3])<<24;
+      b += k[4];
+      b += ((uint32_t)k[5])<<8;
+      b += ((uint32_t)k[6])<<16;
+      b += ((uint32_t)k[7])<<24;
+      c += k[8];
+      c += ((uint32_t)k[9])<<8;
+      c += ((uint32_t)k[10])<<16;
+      c += ((uint32_t)k[11])<<24;
+      mix(a,b,c);
+      length -= 12;
+      k += 12;
+    }
+
+    /*-------------------------------- last block: affect all 32 bits of (c) */
+    switch(length)                   /* all the case statements fall through */
+    {
+    case 12: c+=((uint32_t)k[11])<<24;
+	     /* FALLTHROUGH */
+    case 11: c+=((uint32_t)k[10])<<16;
+	     /* FALLTHROUGH */
+    case 10: c+=((uint32_t)k[9])<<8;
+	     /* FALLTHROUGH */
+    case 9 : c+=k[8];
+	     /* FALLTHROUGH */
+    case 8 : b+=((uint32_t)k[7])<<24;
+	     /* FALLTHROUGH */
+    case 7 : b+=((uint32_t)k[6])<<16;
+	     /* FALLTHROUGH */
+    case 6 : b+=((uint32_t)k[5])<<8;
+	     /* FALLTHROUGH */
+    case 5 : b+=k[4];
+	     /* FALLTHROUGH */
+    case 4 : a+=((uint32_t)k[3])<<24;
+	     /* FALLTHROUGH */
+    case 3 : a+=((uint32_t)k[2])<<16;
+	     /* FALLTHROUGH */
+    case 2 : a+=((uint32_t)k[1])<<8;
+	     /* FALLTHROUGH */
+    case 1 : a+=k[0];
+             break;
+    case 0 : return c;
+    }
+  }
+
+  final(a,b,c);
+  return c;
+}

help-from-md.awk

@@ -0,0 +1,40 @@
+#!/usr/bin/awk -f
+
+# The caller must pass args: -v hfile=help-NAME.h NAME.NUM.md
+
+BEGIN {
+    heading = "/* DO NOT EDIT THIS FILE!  It is auto-generated from the option list in " ARGV[1] "! */"
+    findcomment = hfile
+    sub("\\.", "\\.", findcomment)
+    findcomment = "\\[comment\\].*" findcomment
+    backtick_cnt = 0
+    prints = ""
+}
+
+/^```/ {
+    backtick_cnt++
+    next
+}
+
+foundcomment {
+    if (backtick_cnt > 1) exit
+    if (backtick_cnt == 1) {
+	gsub(/"/, "\\\"")
+	prints = prints "\n  rprintf(F,\"" $0 "\\n\");"
+    }
+    next
+}
+
+$0 ~ findcomment {
+    foundcomment = 1
+    backtick_cnt = 0
+}
+
+END {
+    if (foundcomment && backtick_cnt > 1)
+	print heading "\n" prints > hfile
+    else {
+	print "Failed to find " hfile " section in " ARGV[1]
+	exit 1
+    }
+}

hlink.c

@@ -1,233 +1,566 @@
 /*
-   Copyright (C) Andrew Tridgell 1996
-   Copyright (C) Paul Mackerras 1996
-   Copyright (C) 2002 by Martin Pool <mbp@samba.org>
-
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
+ * Routines to support hard-linking.
+ *
+ * Copyright (C) 1996 Andrew Tridgell
+ * Copyright (C) 1996 Paul Mackerras
+ * Copyright (C) 2002 Martin Pool <mbp@samba.org>
+ * Copyright (C) 2004-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
 #include "rsync.h"
+#include "inums.h"
+#include "ifuncs.h"
 
 extern int dry_run;
-extern int verbose;
-extern int make_backups;
-extern struct file_list *the_file_list;
+extern int list_only;
+extern int am_sender;
+extern int inc_recurse;
+extern int do_xfers;
+extern int alt_dest_type;
+extern int preserve_acls;
+extern int preserve_xattrs;
+extern int protocol_version;
+extern int remove_source_files;
+extern int stdout_format_has_i;
+extern int maybe_ATTRS_REPORT;
+extern int unsort_ndx;
+extern char *basis_dir[MAX_BASIS_DIRS+1];
+extern struct file_list *cur_flist;
 
 #ifdef SUPPORT_HARD_LINKS
 
-#define FPTR(i) (the_file_list->files[i])
-#define LINKED(p1,p2) (FPTR(p1)->F_DEV == FPTR(p2)->F_DEV \
-		    && FPTR(p1)->F_INODE == FPTR(p2)->F_INODE)
+/* Starting with protocol 30, we use a simple hashtable on the sending side
+ * for hashing the st_dev and st_ino info.  The receiving side gets told
+ * (via flags and a "group index") which items are hard-linked together, so
+ * we can avoid the pool of dev+inode data.  For incremental recursion mode,
+ * the receiver will use a ndx hash to remember old pathnames. */
 
-static int hlink_compare(int *int1, int *int2)
-{
-	struct file_struct *f1 = FPTR(*int1);
-	struct file_struct *f2 = FPTR(*int2);
+static void *data_when_new = "";
 
-	if (f1->F_DEV != f2->F_DEV)
-		return (int) (f1->F_DEV > f2->F_DEV ? 1 : -1);
+static struct hashtable *dev_tbl;
 
-	if (f1->F_INODE != f2->F_INODE)
-		return (int) (f1->F_INODE > f2->F_INODE ? 1 : -1);
+static struct hashtable *prior_hlinks;
 
-	return f_name_cmp(f1, f2);
-}
-
-static int *hlink_list;
-static int hlink_count;
-
-/* Analyze the data in the hlink_list[], remove items that aren't multiply
- * linked, and replace the dev+inode data with the hlindex+next linked list. */
-static void link_idev_data(void)
-{
-	int head, from, to, start;
-
-	alloc_pool_t hlink_pool;
-	alloc_pool_t idev_pool = the_file_list->hlink_pool;
-
-	hlink_pool = pool_create(128 * 1024, sizeof (struct hlink),
-	    out_of_memory, POOL_INTERN);
-
-	for (from = to = 0; from < hlink_count; from++) {
-		start = from;
-		head = hlink_list[start];
-		while (from < hlink_count-1
-		    && LINKED(hlink_list[from], hlink_list[from+1])) {
-			pool_free(idev_pool, 0, FPTR(hlink_list[from])->link_u.idev);
-			FPTR(hlink_list[from])->link_u.links = pool_talloc(hlink_pool,
-			    struct hlink, 1, "hlink_list");
-
-			FPTR(hlink_list[from])->F_HLINDEX = to;
-			FPTR(hlink_list[from])->F_NEXT = hlink_list[from+1];
-			from++;
-		}
-		if (from > start) {
-			pool_free(idev_pool, 0, FPTR(hlink_list[from])->link_u.idev);
-			FPTR(hlink_list[from])->link_u.links = pool_talloc(hlink_pool,
-			    struct hlink, 1, "hlink_list");
-
-			FPTR(head)->flags |= FLAG_HLINK_TOL;
-			FPTR(hlink_list[from])->F_HLINDEX = to;
-			FPTR(hlink_list[from])->F_NEXT = head;
-			FPTR(hlink_list[from])->flags |= FLAG_HLINK_EOL;
-			hlink_list[to++] = head;
-		} else {
-			pool_free(idev_pool, 0, FPTR(head)->link_u.idev);
-			FPTR(head)->link_u.idev = NULL;
-		}
-	}
-
-	if (!to) {
-		free(hlink_list);
-		hlink_list = NULL;
-		pool_destroy(hlink_pool);
-		hlink_pool = NULL;
-	} else {
-		hlink_count = to;
-		hlink_list = realloc_array(hlink_list, int, hlink_count);
-		if (!hlink_list)
-			out_of_memory("init_hard_links");
-	}
-	the_file_list->hlink_pool = hlink_pool;
-	pool_destroy(idev_pool);
-}
-#endif
+static struct file_list *hlink_flist;
 
 void init_hard_links(void)
 {
-#ifdef SUPPORT_HARD_LINKS
+	if (am_sender || protocol_version < 30)
+		dev_tbl = hashtable_create(16, HT_KEY64);
+	else if (inc_recurse)
+		prior_hlinks = hashtable_create(1024, HT_KEY32);
+}
+
+struct ht_int64_node *idev_find(int64 dev, int64 ino)
+{
+	static struct ht_int64_node *dev_node = NULL;
+
+	/* Note that some OSes have a dev == 0, so increment to avoid storing a 0. */
+	if (!dev_node || dev_node->key != dev+1) {
+		/* We keep a separate hash table of inodes for every device. */
+		dev_node = hashtable_find(dev_tbl, dev+1, data_when_new);
+		if (dev_node->data == data_when_new) {
+			dev_node->data = hashtable_create(512, HT_KEY64);
+			if (DEBUG_GTE(HLINK, 3)) {
+				rprintf(FINFO, "[%s] created hashtable for dev %s\n",
+					who_am_i(), big_num(dev));
+			}
+		}
+	}
+
+	return hashtable_find(dev_node->data, ino, (void*)-1L);
+}
+
+void idev_destroy(void)
+{
 	int i;
 
-	if (the_file_list->count < 2)
-		return;
-
-	if (hlink_list)
-		free(hlink_list);
-
-	if (!(hlink_list = new_array(int, the_file_list->count)))
-		out_of_memory("init_hard_links");
-
-	hlink_count = 0;
-	for (i = 0; i < the_file_list->count; i++) {
-		if (FPTR(i)->link_u.idev)
-			hlink_list[hlink_count++] = i;
+	for (i = 0; i < dev_tbl->size; i++) {
+		struct ht_int32_node *node = HT_NODE(dev_tbl, dev_tbl->nodes, i);
+		if (node->data)
+			hashtable_destroy(node->data);
 	}
 
-	qsort(hlink_list, hlink_count,
-	    sizeof hlink_list[0], (int (*)()) hlink_compare);
-
-	if (!hlink_count) {
-		free(hlink_list);
-		hlink_list = NULL;
-	} else
-		link_idev_data();
-#endif
+	hashtable_destroy(dev_tbl);
 }
 
-int hard_link_check(struct file_struct *file, int ndx, int skip)
+static int hlink_compare_gnum(int *int1, int *int2)
 {
-#ifdef SUPPORT_HARD_LINKS
-	if (!hlink_list || !file->link_u.links)
-		return 0;
-	if (skip && !(file->flags & FLAG_HLINK_EOL))
-		hlink_list[file->F_HLINDEX] = file->F_NEXT;
-	if (hlink_list[file->F_HLINDEX] != ndx) {
-		if (verbose > 2) {
-			rprintf(FINFO, "\"%s\" is a hard link\n",
-				safe_fname(f_name(file)));
-		}
-		return 1;
-	}
-#endif
-	return 0;
+	struct file_struct *f1 = hlink_flist->sorted[*int1];
+	struct file_struct *f2 = hlink_flist->sorted[*int2];
+	int32 gnum1 = F_HL_GNUM(f1);
+	int32 gnum2 = F_HL_GNUM(f2);
+
+	if (gnum1 != gnum2)
+		return gnum1 > gnum2 ? 1 : -1;
+
+	return *int1 > *int2 ? 1 : -1;
 }
 
-#ifdef SUPPORT_HARD_LINKS
-int hard_link_one(struct file_struct *file, int ndx, char *fname,
-		  int statret, STRUCT_STAT *st, char *toname, int terse,
-		  int itemizing, enum logcode code)
+static void match_gnums(int32 *ndx_list, int ndx_count)
 {
-	if (do_link(toname, fname)) {
-		if (verbose) {
-			rsyserr(FERROR, errno, "link %s => %s failed",
-				full_fname(fname), safe_fname(toname));
+	int32 from, prev;
+	struct file_struct *file, *file_next;
+	struct ht_int32_node *node = NULL;
+	int32 gnum, gnum_next;
+
+	qsort(ndx_list, ndx_count, sizeof ndx_list[0], (int (*)()) hlink_compare_gnum);
+
+	for (from = 0; from < ndx_count; from++) {
+		file = hlink_flist->sorted[ndx_list[from]];
+		gnum = F_HL_GNUM(file);
+		if (inc_recurse) {
+			node = hashtable_find(prior_hlinks, gnum, data_when_new);
+			if (node->data == data_when_new) {
+				node->data = new_array0(char, 5);
+				assert(gnum >= hlink_flist->ndx_start);
+				file->flags |= FLAG_HLINK_FIRST;
+				prev = -1;
+			} else if (CVAL(node->data, 0) == 0) {
+				struct file_list *flist;
+				prev = IVAL(node->data, 1);
+				flist = flist_for_ndx(prev, NULL);
+				if (flist)
+					flist->files[prev - flist->ndx_start]->flags &= ~FLAG_HLINK_LAST;
+				else {
+					/* We skipped all prior files in this
+					 * group, so mark this as a "first". */
+					file->flags |= FLAG_HLINK_FIRST;
+					prev = -1;
+				}
+			} else
+				prev = -1;
+		} else {
+			file->flags |= FLAG_HLINK_FIRST;
+			prev = -1;
 		}
-		return -1;
-	}
-
-	if (itemizing) {
-		itemize(file, ndx, statret, st,
-			ITEM_LOCAL_CHANGE | ITEM_XNAME_FOLLOWS, 0,
-			terse ? "" : toname);
-	}
-	if (code && verbose && !terse) {
-		rprintf(code, "%s => %s\n",
-			safe_fname(fname), safe_fname(toname));
-	}
-	return 0;
-}
-#endif
-
-
-void hard_link_cluster(struct file_struct *file, int master, int itemizing,
-		       enum logcode code)
-{
-#ifdef SUPPORT_HARD_LINKS
-	char hlink1[MAXPATHLEN];
-	char *hlink2;
-	STRUCT_STAT st1, st2;
-	int statret, ndx = master;
-
-	if (link_stat(f_name_to(file, hlink1), &st1, 0) < 0)
-		return;
-	if (!(file->flags & FLAG_HLINK_TOL)) {
-		while (!(file->flags & FLAG_HLINK_EOL)) {
-			ndx = file->F_NEXT;
-			file = FPTR(ndx);
+		for ( ; from < ndx_count-1; file = file_next, gnum = gnum_next, from++) { /*SHARED ITERATOR*/
+			file_next = hlink_flist->sorted[ndx_list[from+1]];
+			gnum_next = F_HL_GNUM(file_next);
+			if (gnum != gnum_next)
+				break;
+			F_HL_PREV(file) = prev;
+			/* The linked list uses over-the-wire ndx values. */
+			if (unsort_ndx)
+				prev = F_NDX(file);
+			else
+				prev = ndx_list[from] + hlink_flist->ndx_start;
 		}
-	}
-	do {
-		ndx = file->F_NEXT;
-		file = FPTR(ndx);
-		if (ndx == master)
+		if (prev < 0 && !inc_recurse) {
+			/* Disable hard-link bit and set DONE so that
+			 * HLINK_BUMP()-dependent values are unaffected. */
+			file->flags &= ~(FLAG_HLINKED | FLAG_HLINK_FIRST);
+			file->flags |= FLAG_HLINK_DONE;
 			continue;
-		hlink2 = f_name(file);
-		if ((statret = link_stat(hlink2, &st2, 0)) == 0) {
-			if (st2.st_dev == st1.st_dev
-			    && st2.st_ino == st1.st_ino) {
-				if (itemizing) {
-					itemize(file, ndx, statret, &st2,
-						ITEM_LOCAL_CHANGE | ITEM_XNAME_FOLLOWS,
-						0, "");
-				}
-				continue;
+		}
+
+		file->flags |= FLAG_HLINK_LAST;
+		F_HL_PREV(file) = prev;
+		if (inc_recurse && CVAL(node->data, 0) == 0) {
+			if (unsort_ndx)
+				prev = F_NDX(file);
+			else
+				prev = ndx_list[from] + hlink_flist->ndx_start;
+			SIVAL(node->data, 1, prev);
+		}
+	}
+}
+
+/* Analyze the hard-links in the file-list by creating a list of all the
+ * items that have hlink data, sorting them, and matching up identical
+ * values into clusters.  These will be a single linked list from last
+ * to first when we're done. */
+void match_hard_links(struct file_list *flist)
+{
+	if (!list_only && flist->used) {
+		int i, ndx_count = 0;
+		int32 *ndx_list;
+
+		ndx_list = new_array(int32, flist->used);
+
+		for (i = 0; i < flist->used; i++) {
+			if (F_IS_HLINKED(flist->sorted[i]))
+				ndx_list[ndx_count++] = i;
+		}
+
+		hlink_flist = flist;
+
+		if (ndx_count)
+			match_gnums(ndx_list, ndx_count);
+
+		free(ndx_list);
+	}
+	if (protocol_version < 30)
+		idev_destroy();
+}
+
+static int maybe_hard_link(struct file_struct *file, int ndx,
+			   char *fname, int statret, stat_x *sxp,
+			   const char *oldname, STRUCT_STAT *old_stp,
+			   const char *realname, int itemizing, enum logcode code)
+{
+	if (statret == 0) {
+		if (sxp->st.st_dev == old_stp->st_dev
+		 && sxp->st.st_ino == old_stp->st_ino) {
+			if (itemizing) {
+				itemize(fname, file, ndx, statret, sxp,
+					ITEM_LOCAL_CHANGE | ITEM_XNAME_FOLLOWS,
+					0, "");
 			}
-			if (make_backups) {
-				if (!make_backup(hlink2))
-					continue;
-			} else if (robust_unlink(hlink2)) {
-				if (verbose > 0) {
-					rsyserr(FINFO, errno,
-						"unlink %s failed",
-						full_fname(hlink2));
+			if (INFO_GTE(NAME, 2) && maybe_ATTRS_REPORT)
+				rprintf(FCLIENT, "%s is uptodate\n", fname);
+			file->flags |= FLAG_HLINK_DONE;
+			return 0;
+		}
+	}
+
+	if (atomic_create(file, fname, NULL, oldname, MAKEDEV(0, 0), sxp, statret == 0 ? DEL_FOR_FILE : 0)) {
+		if (itemizing) {
+			itemize(fname, file, ndx, statret, sxp,
+				ITEM_LOCAL_CHANGE | ITEM_XNAME_FOLLOWS, 0,
+				realname);
+		}
+		if (code != FNONE && INFO_GTE(NAME, 1))
+			rprintf(code, "%s => %s\n", fname, realname);
+		return 0;
+	}
+
+	return -1;
+}
+
+/* Figure out if a prior entry is still there or if we just have a
+ * cached name for it. */
+static char *check_prior(struct file_struct *file, int gnum,
+			 int *prev_ndx_p, struct file_list **flist_p)
+{
+	struct file_struct *fp;
+	struct ht_int32_node *node;
+	int prev_ndx = F_HL_PREV(file);
+
+	while (1) {
+		struct file_list *flist;
+		if (prev_ndx < 0
+		 || (flist = flist_for_ndx(prev_ndx, NULL)) == NULL)
+			break;
+		fp = flist->files[prev_ndx - flist->ndx_start];
+		if (!(fp->flags & FLAG_SKIP_HLINK)) {
+			*prev_ndx_p = prev_ndx;
+			*flist_p = flist;
+			return NULL;
+		}
+		F_HL_PREV(file) = prev_ndx = F_HL_PREV(fp);
+	}
+
+	if (inc_recurse
+	 && (node = hashtable_find(prior_hlinks, gnum, NULL)) != NULL) {
+		assert(node->data != NULL);
+		if (CVAL(node->data, 0) != 0) {
+			*prev_ndx_p = -1;
+			*flist_p = NULL;
+			return node->data;
+		}
+		/* The prior file must have been skipped. */
+		F_HL_PREV(file) = -1;
+	}
+
+	*prev_ndx_p = -1;
+	*flist_p = NULL;
+	return NULL;
+}
+
+/* Only called if FLAG_HLINKED is set and FLAG_HLINK_FIRST is not.  Returns:
+ * 0 = process the file, 1 = skip the file, -1 = error occurred. */
+int hard_link_check(struct file_struct *file, int ndx, char *fname,
+		    int statret, stat_x *sxp, int itemizing,
+		    enum logcode code)
+{
+	STRUCT_STAT prev_st;
+	char namebuf[MAXPATHLEN], altbuf[MAXPATHLEN];
+	char *realname, *prev_name;
+	struct file_list *flist;
+	int gnum = inc_recurse ? F_HL_GNUM(file) : -1;
+	int prev_ndx;
+
+	prev_name = realname = check_prior(file, gnum, &prev_ndx, &flist);
+
+	if (!prev_name) {
+		struct file_struct *prev_file;
+
+		if (!flist) {
+			/* The previous file was skipped, so this one is
+			 * treated as if it were the first in its group. */
+			if (DEBUG_GTE(HLINK, 2)) {
+				rprintf(FINFO, "hlink for %d (%s,%d): virtual first\n",
+					ndx, f_name(file, NULL), gnum);
+			}
+			return 0;
+		}
+
+		prev_file = flist->files[prev_ndx - flist->ndx_start];
+
+		/* Is the previous link not complete yet? */
+		if (!(prev_file->flags & FLAG_HLINK_DONE)) {
+			/* Is the previous link being transferred? */
+			if (prev_file->flags & FLAG_FILE_SENT) {
+				/* Add ourselves to the list of files that will
+				 * be updated when the transfer completes, and
+				 * mark ourself as waiting for the transfer. */
+				F_HL_PREV(file) = F_HL_PREV(prev_file);
+				F_HL_PREV(prev_file) = ndx;
+				file->flags |= FLAG_FILE_SENT;
+				cur_flist->in_progress++;
+				if (DEBUG_GTE(HLINK, 2)) {
+					rprintf(FINFO, "hlink for %d (%s,%d): waiting for %d\n",
+						ndx, f_name(file, NULL), gnum, F_HL_PREV(file));
 				}
-				continue;
+				return 1;
+			}
+			if (DEBUG_GTE(HLINK, 2)) {
+				rprintf(FINFO, "hlink for %d (%s,%d): looking for a leader\n",
+					ndx, f_name(file, NULL), gnum);
+			}
+			return 0;
+		}
+
+		/* There is a finished file to link with! */
+		if (!(prev_file->flags & FLAG_HLINK_FIRST)) {
+			/* The previous previous is FIRST when prev is not. */
+			prev_name = realname = check_prior(prev_file, gnum, &prev_ndx, &flist);
+			/* Update our previous pointer to point to the FIRST. */
+			F_HL_PREV(file) = prev_ndx;
+		}
+
+		if (!prev_name) {
+			int alt_dest;
+
+			assert(flist != NULL);
+			prev_file = flist->files[prev_ndx - flist->ndx_start];
+			/* F_HL_PREV() is alt_dest value when DONE && FIRST. */
+			alt_dest = F_HL_PREV(prev_file);
+			if (DEBUG_GTE(HLINK, 2)) {
+				rprintf(FINFO, "hlink for %d (%s,%d): found flist match (alt %d)\n",
+					ndx, f_name(file, NULL), gnum, alt_dest);
+			}
+
+			if (alt_dest >= 0 && dry_run) {
+				pathjoin(namebuf, MAXPATHLEN, basis_dir[alt_dest],
+					 f_name(prev_file, NULL));
+				prev_name = namebuf;
+				realname = f_name(prev_file, altbuf);
+			} else {
+				prev_name = f_name(prev_file, namebuf);
+				realname = prev_name;
 			}
 		}
-		hard_link_one(file, ndx, hlink2, statret,
-			      &st2, hlink1, 0, itemizing, code);
-	} while (!(file->flags & FLAG_HLINK_EOL));
+	}
+
+	if (DEBUG_GTE(HLINK, 2)) {
+		rprintf(FINFO, "hlink for %d (%s,%d): leader is %d (%s)\n",
+			ndx, f_name(file, NULL), gnum, prev_ndx, prev_name);
+	}
+
+	if (link_stat(prev_name, &prev_st, 0) < 0) {
+		if (!dry_run || errno != ENOENT) {
+			rsyserr(FERROR_XFER, errno, "stat %s failed", full_fname(prev_name));
+			return -1;
+		}
+		/* A new hard-link will get a new dev & inode, so approximate
+		 * those values in dry-run mode by zeroing them. */
+		memset(&prev_st, 0, sizeof prev_st);
+	}
+
+	if (statret < 0 && basis_dir[0] != NULL) {
+		/* If we match an alt-dest item, we don't output this as a change. */
+		char cmpbuf[MAXPATHLEN];
+		stat_x alt_sx;
+		int j = 0;
+		init_stat_x(&alt_sx);
+		do {
+			pathjoin(cmpbuf, MAXPATHLEN, basis_dir[j], fname);
+			if (link_stat(cmpbuf, &alt_sx.st, 0) < 0)
+				continue;
+			if (alt_dest_type == LINK_DEST) {
+				if (prev_st.st_dev != alt_sx.st.st_dev
+				 || prev_st.st_ino != alt_sx.st.st_ino)
+					continue;
+				statret = 1;
+				if (stdout_format_has_i == 0
+				 || (!INFO_GTE(NAME, 2) && stdout_format_has_i < 2)) {
+					itemizing = 0;
+					code = FNONE;
+					if (INFO_GTE(NAME, 2) && maybe_ATTRS_REPORT)
+						rprintf(FCLIENT, "%s is uptodate\n", fname);
+				}
+				break;
+			}
+			if (!unchanged_file(cmpbuf, file, &alt_sx.st))
+				continue;
+			statret = 1;
+			if (unchanged_attrs(cmpbuf, file, &alt_sx))
+				break;
+		} while (basis_dir[++j] != NULL);
+		if (statret == 1) {
+			sxp->st = alt_sx.st;
+#ifdef SUPPORT_ACLS
+			if (preserve_acls && !S_ISLNK(file->mode)) {
+				free_acl(sxp);
+				if (!ACL_READY(alt_sx))
+					get_acl(cmpbuf, sxp);
+				else {
+					sxp->acc_acl = alt_sx.acc_acl;
+					sxp->def_acl = alt_sx.def_acl;
+					alt_sx.acc_acl = alt_sx.def_acl = NULL;
+				}
+			}
 #endif
+#ifdef SUPPORT_XATTRS
+			if (preserve_xattrs) {
+				free_xattr(sxp);
+				if (!XATTR_READY(alt_sx))
+					get_xattr(cmpbuf, sxp);
+				else {
+					sxp->xattr = alt_sx.xattr;
+					alt_sx.xattr = NULL;
+				}
+			}
+#endif
+		} else
+			free_stat_x(&alt_sx);
+	}
+
+	if (maybe_hard_link(file, ndx, fname, statret, sxp, prev_name, &prev_st,
+			    realname, itemizing, code) < 0)
+		return -1;
+
+	if (remove_source_files == 1 && do_xfers)
+		send_msg_int(MSG_SUCCESS, ndx);
+
+	return 1;
 }
+
+int hard_link_one(struct file_struct *file, const char *fname,
+		  const char *oldname, int terse)
+{
+	if (do_link(oldname, fname) < 0) {
+		enum logcode code;
+		if (terse) {
+			if (!INFO_GTE(NAME, 1))
+				return 0;
+			code = FINFO;
+		} else
+			code = FERROR_XFER;
+		rsyserr(code, errno, "link %s => %s failed",
+			full_fname(fname), oldname);
+		return 0;
+	}
+
+	file->flags |= FLAG_HLINK_DONE;
+
+	return 1;
+}
+
+void finish_hard_link(struct file_struct *file, const char *fname, int fin_ndx,
+		      STRUCT_STAT *stp, int itemizing, enum logcode code,
+		      int alt_dest)
+{
+	stat_x prev_sx;
+	STRUCT_STAT st;
+	char prev_name[MAXPATHLEN], alt_name[MAXPATHLEN];
+	const char *our_name;
+	struct file_list *flist;
+	int prev_statret, ndx, prev_ndx = F_HL_PREV(file);
+
+	if (stp == NULL && prev_ndx >= 0) {
+		if (link_stat(fname, &st, 0) < 0 && !dry_run) {
+			rsyserr(FERROR_XFER, errno, "stat %s failed",
+				full_fname(fname));
+			return;
+		}
+		stp = &st;
+	}
+
+	/* FIRST combined with DONE means we were the first to get done. */
+	file->flags |= FLAG_HLINK_FIRST | FLAG_HLINK_DONE;
+	F_HL_PREV(file) = alt_dest;
+	if (alt_dest >= 0 && dry_run) {
+		pathjoin(alt_name, MAXPATHLEN, basis_dir[alt_dest],
+			 f_name(file, NULL));
+		our_name = alt_name;
+	} else
+		our_name = fname;
+
+	init_stat_x(&prev_sx);
+
+	while ((ndx = prev_ndx) >= 0) {
+		int val;
+		flist = flist_for_ndx(ndx, "finish_hard_link");
+		file = flist->files[ndx - flist->ndx_start];
+		file->flags = (file->flags & ~FLAG_HLINK_FIRST) | FLAG_HLINK_DONE;
+		prev_ndx = F_HL_PREV(file);
+		F_HL_PREV(file) = fin_ndx;
+		prev_statret = link_stat(f_name(file, prev_name), &prev_sx.st, 0);
+		val = maybe_hard_link(file, ndx, prev_name, prev_statret, &prev_sx,
+				      our_name, stp, fname, itemizing, code);
+		flist->in_progress--;
+		free_stat_x(&prev_sx);
+		if (val < 0)
+			continue;
+		if (remove_source_files == 1 && do_xfers)
+			send_msg_int(MSG_SUCCESS, ndx);
+	}
+
+	if (inc_recurse) {
+		int gnum = F_HL_GNUM(file);
+		struct ht_int32_node *node = hashtable_find(prior_hlinks, gnum, NULL);
+		if (node == NULL) {
+			rprintf(FERROR, "Unable to find a hlink node for %d (%s)\n", gnum, f_name(file, prev_name));
+			exit_cleanup(RERR_MESSAGEIO);
+		}
+		if (node->data == NULL) {
+			rprintf(FERROR, "Hlink node data for %d is NULL (%s)\n", gnum, f_name(file, prev_name));
+			exit_cleanup(RERR_MESSAGEIO);
+		}
+		if (CVAL(node->data, 0) != 0) {
+			rprintf(FERROR, "Hlink node data for %d already has path=%s (%s)\n",
+				gnum, (char*)node->data, f_name(file, prev_name));
+			exit_cleanup(RERR_MESSAGEIO);
+		}
+		free(node->data);
+		node->data = strdup(our_name);
+	}
+}
+
+int skip_hard_link(struct file_struct *file, struct file_list **flist_p)
+{
+	struct file_list *flist;
+	int prev_ndx;
+
+	file->flags |= FLAG_SKIP_HLINK;
+	if (!(file->flags & FLAG_HLINK_LAST))
+		return -1;
+
+	check_prior(file, F_HL_GNUM(file), &prev_ndx, &flist);
+	if (prev_ndx >= 0) {
+		file = flist->files[prev_ndx - flist->ndx_start];
+		if (file->flags & (FLAG_HLINK_DONE|FLAG_FILE_SENT))
+			return -1;
+		file->flags |= FLAG_HLINK_LAST;
+		*flist_p = flist;
+	}
+
+	return prev_ndx;
+}
+#endif

ifuncs.h

@@ -0,0 +1,111 @@
+/* Inline functions for rsync.
+ *
+ * Copyright (C) 2007-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
+
+static inline void
+alloc_xbuf(xbuf *xb, size_t sz)
+{
+	xb->buf = new_array(char, sz);
+	xb->size = sz;
+	xb->len = xb->pos = 0;
+}
+
+static inline void
+realloc_xbuf(xbuf *xb, size_t sz)
+{
+	char *bf = realloc_array(xb->buf, char, sz);
+	xb->buf = bf;
+	xb->size = sz;
+}
+
+static inline void
+free_xbuf(xbuf *xb)
+{
+	if (xb->buf)
+		free(xb->buf);
+	memset(xb, 0, sizeof (xbuf));
+}
+
+static inline int
+to_wire_mode(mode_t mode)
+{
+#ifdef SUPPORT_LINKS
+#if _S_IFLNK != 0120000
+	if (S_ISLNK(mode))
+		return (mode & ~(_S_IFMT)) | 0120000;
+#endif
+#endif
+	return mode;
+}
+
+static inline mode_t
+from_wire_mode(int mode)
+{
+#if _S_IFLNK != 0120000
+	if ((mode & (_S_IFMT)) == 0120000)
+		return (mode & ~(_S_IFMT)) | _S_IFLNK;
+#endif
+	return mode;
+}
+
+static inline char *
+d_name(struct dirent *di)
+{
+#ifdef HAVE_BROKEN_READDIR
+	return (di->d_name - 2);
+#else
+	return di->d_name;
+#endif
+}
+
+static inline void
+init_stat_x(stat_x *sx_p)
+{
+#ifdef SUPPORT_ACLS
+	sx_p->acc_acl = sx_p->def_acl = NULL;
+#endif
+#ifdef SUPPORT_XATTRS
+	sx_p->xattr = NULL;
+#endif
+}
+
+static inline void
+free_stat_x(stat_x *sx_p)
+{
+#ifdef SUPPORT_ACLS
+    {
+	extern int preserve_acls;
+	if (preserve_acls)
+		free_acl(sx_p);
+    }
+#endif
+#ifdef SUPPORT_XATTRS
+    {
+	extern int preserve_xattrs;
+	if (preserve_xattrs)
+		free_xattr(sx_p);
+    }
+#endif
+}
+
+static inline char *my_strdup(const char *str, const char *file, int line)
+{
+    int len = strlen(str)+1;
+    char *buf = my_alloc(do_malloc, len, 1, file, line);
+    memcpy(buf, str, len);
+    return buf;
+}

inums.h

@@ -0,0 +1,57 @@
+/* Inline functions for rsync.
+ *
+ * Copyright (C) 2008-2019 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
+
+static inline char *
+big_num(int64 num)
+{
+	return do_big_num(num, 0, NULL);
+}
+
+static inline char *
+comma_num(int64 num)
+{
+	extern int human_readable;
+	return do_big_num(num, human_readable != 0, NULL);
+}
+
+static inline char *
+human_num(int64 num)
+{
+	extern int human_readable;
+	return do_big_num(num, human_readable, NULL);
+}
+
+static inline char *
+big_dnum(double dnum, int decimal_digits)
+{
+	return do_big_dnum(dnum, 0, decimal_digits);
+}
+
+static inline char *
+comma_dnum(double dnum, int decimal_digits)
+{
+	extern int human_readable;
+	return do_big_dnum(dnum, human_readable != 0, decimal_digits);
+}
+
+static inline char *
+human_dnum(double dnum, int decimal_digits)
+{
+	extern int human_readable;
+	return do_big_dnum(dnum, human_readable, decimal_digits);
+}

io.h

@@ -0,0 +1,52 @@
+/*
+ * Copyright (C) 2007-2019 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
+
+extern int protocol_version;
+
+static inline int32
+read_varint30(int f)
+{
+	if (protocol_version < 30)
+		return read_int(f);
+	return read_varint(f);
+}
+
+static inline int64
+read_varlong30(int f, uchar min_bytes)
+{
+	if (protocol_version < 30)
+		return read_longint(f);
+	return read_varlong(f, min_bytes);
+}
+
+static inline void
+write_varint30(int f, int32 x)
+{
+	if (protocol_version < 30)
+		write_int(f, x);
+	else
+		write_varint(f, x);
+}
+
+static inline void
+write_varlong30(int f, int64 x, uchar min_bytes)
+{
+	if (protocol_version < 30)
+		write_longint(f, x);
+	else
+		write_varlong(f, x, min_bytes);
+}

itypes.h

@@ -0,0 +1,65 @@
+/* Inline functions for rsync.
+ *
+ * Copyright (C) 2007-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
+
+static inline int
+isDigit(const char *ptr)
+{
+	return isdigit(*(unsigned char *)ptr);
+}
+
+static inline int
+isHexDigit(const char *ptr)
+{
+	return isxdigit(*(unsigned char *)ptr);
+}
+
+static inline int
+isPrint(const char *ptr)
+{
+	return isprint(*(unsigned char *)ptr);
+}
+
+static inline int
+isSpace(const char *ptr)
+{
+	return isspace(*(unsigned char *)ptr);
+}
+
+static inline int
+isLower(const char *ptr)
+{
+	return islower(*(unsigned char *)ptr);
+}
+
+static inline int
+isUpper(const char *ptr)
+{
+	return isupper(*(unsigned char *)ptr);
+}
+
+static inline int
+toLower(const char *ptr)
+{
+	return tolower(*(unsigned char *)ptr);
+}
+
+static inline int
+toUpper(const char *ptr)
+{
+	return toupper(*(unsigned char *)ptr);
+}

latest-year.h

@@ -0,0 +1 @@
+#define LATEST_YEAR "2020"

lib/.cvsignore

@@ -1 +0,0 @@
-dummy

lib/addrinfo.h

@@ -0,0 +1,180 @@
+/*
+PostgreSQL Database Management System
+(formerly known as Postgres, then as Postgres95)
+
+Portions Copyright (c) 1996-2005, The PostgreSQL Global Development Group
+
+Portions Copyright (c) 1994, The Regents of the University of California
+
+Permission to use, copy, modify, and distribute this software and its
+documentation for any purpose, without fee, and without a written agreement
+is hereby granted, provided that the above copyright notice and this paragraph
+and the following two paragraphs appear in all copies.
+
+IN NO EVENT SHALL THE UNIVERSITY OF CALIFORNIA BE LIABLE TO ANY PARTY FOR
+DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES, INCLUDING
+LOST PROFITS, ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION,
+EVEN IF THE UNIVERSITY OF CALIFORNIA HAS BEEN ADVISED OF THE POSSIBILITY OF
+SUCH DAMAGE.
+
+THE UNIVERSITY OF CALIFORNIA SPECIFICALLY DISCLAIMS ANY WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+AND FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
+ON AN "AS IS" BASIS, AND THE UNIVERSITY OF CALIFORNIA HAS NO OBLIGATIONS
+TO PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+
+*/
+
+/*-------------------------------------------------------------------------
+ *
+ * getaddrinfo.h
+ *	  Support getaddrinfo() on platforms that don't have it.
+ *
+ * Note: we use our own routines on platforms that don't HAVE_STRUCT_ADDRINFO,
+ * whether or not the library routine getaddrinfo() can be found.  This
+ * policy is needed because on some platforms a manually installed libbind.a
+ * may provide getaddrinfo(), yet the system headers may not provide the
+ * struct definitions needed to call it.  To avoid conflict with the libbind
+ * definition in such cases, we rename our routines to pg_xxx() via macros.
+ *
+ * This code will also work on platforms where struct addrinfo is defined
+ * in the system headers but no getaddrinfo() can be located.
+ *
+ * Copyright (c) 2003-2007, PostgreSQL Global Development Group
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef ADDRINFO_H
+#define ADDRINFO_H
+
+
+/* Various macros that ought to be in <netdb.h>, but might not be */
+
+#ifndef EAI_FAIL
+#define EAI_BADFLAGS	(-1)
+#define EAI_NONAME		(-2)
+#define EAI_AGAIN		(-3)
+#define EAI_FAIL		(-4)
+#define EAI_FAMILY		(-6)
+#define EAI_SOCKTYPE	(-7)
+#define EAI_SERVICE		(-8)
+#define EAI_MEMORY		(-10)
+#define EAI_SYSTEM		(-11)
+#endif   /* !EAI_FAIL */
+
+#ifndef AI_PASSIVE
+#define AI_PASSIVE		0x0001
+#endif
+
+#ifndef AI_NUMERICHOST
+/*
+ * some platforms don't support AI_NUMERICHOST; define as zero if using
+ * the system version of getaddrinfo...
+ */
+#if defined(HAVE_STRUCT_ADDRINFO) && defined(HAVE_GETADDRINFO)
+#define AI_NUMERICHOST	0
+#else
+#define AI_NUMERICHOST	0x0004
+#endif
+#endif
+
+#ifndef AI_CANONNAME
+#if defined(HAVE_STRUCT_ADDRINFO) && defined(HAVE_GETADDRINFO)
+#define AI_CANONNAME 0
+#else
+#define AI_CANONNAME 0x0008
+#endif
+#endif
+
+#ifndef AI_NUMERICSERV
+#if defined(HAVE_STRUCT_ADDRINFO) && defined(HAVE_GETADDRINFO)
+#define AI_NUMERICSERV 0
+#else
+#define AI_NUMERICSERV 0x0010
+#endif
+#endif
+
+#ifndef NI_NUMERICHOST
+#define NI_NUMERICHOST	1
+#endif
+
+#ifndef NI_NUMERICSERV
+#define NI_NUMERICSERV	2
+#endif
+
+#ifndef NI_NOFQDN
+#define NI_NOFQDN	4
+#endif
+
+#ifndef NI_NAMEREQD
+#define NI_NAMEREQD 	8
+#endif
+
+#ifndef NI_DGRAM
+#define NI_DGRAM	16
+#endif
+
+
+#ifndef NI_MAXHOST
+#define NI_MAXHOST	1025
+#endif
+
+#ifndef NI_MAXSERV
+#define NI_MAXSERV	32
+#endif
+
+#ifndef HAVE_STRUCT_ADDRINFO
+struct addrinfo
+{
+	int			ai_flags;
+	int			ai_family;
+	int			ai_socktype;
+	int			ai_protocol;
+	size_t		ai_addrlen;
+	struct sockaddr *ai_addr;
+	char	   *ai_canonname;
+	struct addrinfo *ai_next;
+};
+#endif   /* !HAVE_STRUCT_ADDRINFO */
+
+#ifndef HAVE_STRUCT_SOCKADDR_STORAGE
+struct sockaddr_storage {
+	unsigned short ss_family;
+	unsigned long ss_align;
+	char ss_padding[128 - sizeof (unsigned long)];
+};
+#endif	/* !HAVE_STRUCT_SOCKADDR_STORAGE */
+
+#ifndef HAVE_GETADDRINFO
+
+/* Rename private copies per comments above */
+#ifdef getaddrinfo
+#undef getaddrinfo
+#endif
+#define getaddrinfo pg_getaddrinfo
+
+#ifdef freeaddrinfo
+#undef freeaddrinfo
+#endif
+#define freeaddrinfo pg_freeaddrinfo
+
+#ifdef gai_strerror
+#undef gai_strerror
+#endif
+#define gai_strerror pg_gai_strerror
+
+#ifdef getnameinfo
+#undef getnameinfo
+#endif
+#define getnameinfo pg_getnameinfo
+
+extern int getaddrinfo(const char *node, const char *service,
+			const struct addrinfo * hints, struct addrinfo ** res);
+extern void freeaddrinfo(struct addrinfo * res);
+extern const char *gai_strerror(int errcode);
+extern int getnameinfo(const struct sockaddr * sa, socklen_t salen,
+			char *node, size_t nodelen,
+			char *service, size_t servicelen, int flags);
+#endif   /* !HAVE_GETADDRINFO */
+
+#endif   /* ADDRINFO_H */

lib/compat.c

@@ -1,42 +1,37 @@
-/* 
-   Copyright (C) Andrew Tridgell 1998
-   Copyright (C) 2002 by Martin Pool
-   
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
-
-/**
- * @file compat.c
+/*
+ * Reimplementations of standard functions for platforms that don't have them.
  *
- * Reimplementations of standard functions for platforms that don't
- * have them.
- **/
-
-
+ * Copyright (C) 1998 Andrew Tridgell
+ * Copyright (C) 2002 Martin Pool
+ * Copyright (C) 2004-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
 #include "rsync.h"
+#include "itypes.h"
 
+static char number_separator;
 
 #ifndef HAVE_STRDUP
  char *strdup(char *s)
 {
-  int l = strlen(s) + 1;
-  char *ret = (char *)malloc(l);
-  if (ret)
-    strcpy(ret,s);
-  return ret;
+	int len = strlen(s) + 1;
+	char *ret = (char *)malloc(len);
+	if (ret)
+		memcpy(ret, s, len);
+	return ret;
 }
 #endif
 
@@ -84,9 +79,9 @@
 
 #ifndef HAVE_STRPBRK
 /**
- * Find the first ocurrence in @p s of any character in @p accept.
+ * Find the first occurrence in @p s of any character in @p accept.
  *
- * Derived from glibc 
+ * Derived from glibc
  **/
  char *strpbrk(const char *s, const char *accept)
 {
@@ -105,7 +100,7 @@
 
 #ifndef HAVE_STRLCPY
 /**
- * Like strncpy but does not 0 fill the buffer and always null 
+ * Like strncpy but does not 0 fill the buffer and always null
  * terminates.
  *
  * @param bufsize is the size of the destination buffer.
@@ -128,7 +123,7 @@
 
 #ifndef HAVE_STRLCAT
 /**
- * Like strncat() but does not 0 fill the buffer and always null 
+ * Like strncat() but does not 0 fill the buffer and always null
  * terminates.
  *
  * @param bufsize length of the buffer, which should be one more than
@@ -150,49 +145,6 @@
 }
 #endif
 
-#ifdef REPLACE_INET_NTOA
- char *rep_inet_ntoa(struct in_addr ip)
-{
-	unsigned char *p = (unsigned char *)&ip.s_addr;
-	static char buf[18];
-#ifdef WORDS_BIGENDIAN
-	snprintf(buf, 18, "%d.%d.%d.%d", 
-		 (int)p[0], (int)p[1], (int)p[2], (int)p[3]);
-#else
-	snprintf(buf, 18, "%d.%d.%d.%d", 
-		 (int)p[3], (int)p[2], (int)p[1], (int)p[0]);
-#endif
-	return buf;
-}
-#endif
-
-#ifdef REPLACE_INET_ATON
- int inet_aton(const char *cp, struct in_addr *inp)
-{
-	unsigned int a1, a2, a3, a4;
-	unsigned long ret;
-
-	if (strcmp(cp, "255.255.255.255") == 0) {
-		inp->s_addr = (unsigned) -1;
-		return 0;
-	}
-
-	if (sscanf(cp, "%u.%u.%u.%u", &a1, &a2, &a3, &a4) != 4 ||
-	    a1 > 255 || a2 > 255 || a3 > 255 || a4 > 255) {
-		return 0;
-	}
-
-	ret = (a1 << 24) | (a2 << 16) | (a3 << 8) | a4;
-
-	inp->s_addr = htonl(ret);
-	
-	if (inp->s_addr == (unsigned) -1) {
-		return 0;
-	}
-	return 1;
-}
-#endif
-
 /* some systems don't take the 2nd argument */
 int sys_gettimeofday(struct timeval *tv)
 {
@@ -202,3 +154,122 @@ int sys_gettimeofday(struct timeval *tv)
 	return gettimeofday(tv);
 #endif
 }
+
+#define HUMANIFY(mult) \
+	do { \
+		if (num >= mult || num <= -mult) { \
+			double dnum = (double)num / mult; \
+			char units; \
+			if (num < 0) \
+				dnum = -dnum; \
+			if (dnum < mult) \
+				units = 'K'; \
+			else if ((dnum /= mult) < mult) \
+				units = 'M'; \
+			else if ((dnum /= mult) < mult) \
+				units = 'G'; \
+			else { \
+				dnum /= mult; \
+				units = 'T'; \
+			} \
+			if (num < 0) \
+				dnum = -dnum; \
+			snprintf(bufs[n], sizeof bufs[0], "%.2f%c", dnum, units); \
+			return bufs[n]; \
+		} \
+	} while (0)
+
+/* Return the int64 number as a string.  If the human_flag arg is non-zero,
+ * we may output the number in K, M, G, or T units.  If we don't add a unit
+ * suffix, we will append the fract string, if it is non-NULL.  We can
+ * return up to 4 buffers at a time. */
+char *do_big_num(int64 num, int human_flag, const char *fract)
+{
+	static char bufs[4][128]; /* more than enough room */
+	static unsigned int n;
+	char *s;
+	int len, negated;
+
+	if (human_flag && !number_separator) {
+		char buf[32];
+		snprintf(buf, sizeof buf, "%f", 3.14);
+		if (strchr(buf, '.') != NULL)
+			number_separator = ',';
+		else
+			number_separator = '.';
+	}
+
+	n = (n + 1) % (sizeof bufs / sizeof bufs[0]);
+
+	if (human_flag > 1) {
+		if (human_flag == 2)
+			HUMANIFY(1000);
+		else
+			HUMANIFY(1024);
+	}
+
+	s = bufs[n] + sizeof bufs[0] - 1;
+	if (fract) {
+		len = strlen(fract);
+		s -= len;
+		strlcpy(s, fract, len + 1);
+	} else
+		*s = '\0';
+
+	len = 0;
+
+	if (!num)
+		*--s = '0';
+	if (num < 0) {
+		/* A maximum-size negated number can't fit as a positive,
+		 * so do one digit in negated form to start us off. */
+		*--s = (char)(-(num % 10)) + '0';
+		num = -(num / 10);
+		len++;
+		negated = 1;
+	} else
+		negated = 0;
+
+	while (num) {
+		if (human_flag) {
+			if (len == 3) {
+				*--s = number_separator;
+				len = 1;
+			} else
+				len++;
+		}
+		*--s = (char)(num % 10) + '0';
+		num /= 10;
+	}
+
+	if (negated)
+		*--s = '-';
+
+	return s;
+}
+
+/* Return the double number as a string.  If the human_flag option is > 1,
+ * we may output the number in K, M, G, or T units.  The buffer we use for
+ * our result is either a single static buffer defined here, or a buffer
+ * we get from do_big_num(). */
+char *do_big_dnum(double dnum, int human_flag, int decimal_digits)
+{
+	static char tmp_buf[128];
+#if SIZEOF_INT64 >= 8
+	char *fract;
+
+	snprintf(tmp_buf, sizeof tmp_buf, "%.*f", decimal_digits, dnum);
+
+	if (!human_flag || (dnum < 1000.0 && dnum > -1000.0))
+		return tmp_buf;
+
+	for (fract = tmp_buf+1; isDigit(fract); fract++) {}
+
+	return do_big_num((int64)dnum, human_flag, fract);
+#else
+	/* A big number might lose digits converting to a too-short int64,
+	 * so let's just return the raw double conversion. */
+	snprintf(tmp_buf, sizeof tmp_buf, "%.*f", decimal_digits, dnum);
+	return tmp_buf;
+#endif
+}

lib/getaddrinfo.c

@@ -0,0 +1,504 @@
+/*
+PostgreSQL Database Management System
+(formerly known as Postgres, then as Postgres95)
+
+Portions Copyright (c) 1996-2005, The PostgreSQL Global Development Group
+
+Portions Copyright (c) 1994, The Regents of the University of California
+
+Permission to use, copy, modify, and distribute this software and its
+documentation for any purpose, without fee, and without a written agreement
+is hereby granted, provided that the above copyright notice and this paragraph
+and the following two paragraphs appear in all copies.
+
+IN NO EVENT SHALL THE UNIVERSITY OF CALIFORNIA BE LIABLE TO ANY PARTY FOR
+DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES, INCLUDING
+LOST PROFITS, ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION,
+EVEN IF THE UNIVERSITY OF CALIFORNIA HAS BEEN ADVISED OF THE POSSIBILITY OF
+SUCH DAMAGE.
+
+THE UNIVERSITY OF CALIFORNIA SPECIFICALLY DISCLAIMS ANY WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+AND FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
+ON AN "AS IS" BASIS, AND THE UNIVERSITY OF CALIFORNIA HAS NO OBLIGATIONS
+TO PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+
+*/
+
+/*-------------------------------------------------------------------------
+ *
+ * getaddrinfo.c
+ *	  Support getaddrinfo() on platforms that don't have it.
+ *
+ * We also supply getnameinfo() here, assuming that the platform will have
+ * it if and only if it has getaddrinfo().	If this proves false on some
+ * platform, we'll need to split this file and provide a separate configure
+ * test for getnameinfo().
+ *
+ * Copyright (c) 2003-2007, PostgreSQL Global Development Group
+ *
+ * Copyright (C) 2007 Jeremy Allison.
+ * Modified to return multiple IPv4 addresses for Samba.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+#include "rsync.h"
+
+#ifndef SMB_MALLOC
+#define SMB_MALLOC(s) malloc(s)
+#endif
+
+#ifndef SMB_STRDUP
+#define SMB_STRDUP(s) strdup(s)
+#endif
+
+#ifndef HOST_NAME_MAX
+#define HOST_NAME_MAX 255
+#endif
+
+static int check_hostent_err(struct hostent *hp)
+{
+#ifndef INET6
+	extern int h_errno;
+#endif
+	if (!hp) {
+		switch (h_errno) {
+			case HOST_NOT_FOUND:
+			case NO_DATA:
+				return EAI_NONAME;
+			case TRY_AGAIN:
+				return EAI_AGAIN;
+			case NO_RECOVERY:
+			default:
+				return EAI_FAIL;
+		}
+	}
+	if (!hp->h_name || hp->h_addrtype != AF_INET) {
+		return EAI_FAIL;
+	}
+	return 0;
+}
+
+static char *canon_name_from_hostent(struct hostent *hp,
+				int *perr)
+{
+	char *ret = NULL;
+
+	*perr = check_hostent_err(hp);
+	if (*perr) {
+		return NULL;
+	}
+	ret = SMB_STRDUP(hp->h_name);
+	if (!ret) {
+		*perr = EAI_MEMORY;
+	}
+	return ret;
+}
+
+static char *get_my_canon_name(int *perr)
+{
+	char name[HOST_NAME_MAX+1];
+
+	if (gethostname(name, HOST_NAME_MAX) == -1) {
+		*perr = EAI_FAIL;
+		return NULL;
+	}
+	/* Ensure null termination. */
+	name[HOST_NAME_MAX] = '\0';
+	return canon_name_from_hostent(gethostbyname(name), perr);
+}
+
+static char *get_canon_name_from_addr(struct in_addr ip,
+				int *perr)
+{
+	return canon_name_from_hostent(
+			gethostbyaddr((void *)&ip, sizeof ip, AF_INET),
+			perr);
+}
+
+static struct addrinfo *alloc_entry(const struct addrinfo *hints,
+				struct in_addr ip,
+				unsigned short port)
+{
+	struct sockaddr_in *psin = NULL;
+	struct addrinfo *ai = SMB_MALLOC(sizeof(*ai));
+
+	if (!ai) {
+		return NULL;
+	}
+	memset(ai, '\0', sizeof(*ai));
+
+	psin = SMB_MALLOC(sizeof(*psin));
+	if (!psin) {
+		free(ai);
+		return NULL;
+	}
+
+	memset(psin, '\0', sizeof(*psin));
+
+	psin->sin_family = AF_INET;
+	psin->sin_port = htons(port);
+	psin->sin_addr = ip;
+
+	ai->ai_flags = 0;
+	ai->ai_family = AF_INET;
+	ai->ai_socktype = hints->ai_socktype;
+	ai->ai_protocol = hints->ai_protocol;
+	ai->ai_addrlen = sizeof(*psin);
+	ai->ai_addr = (struct sockaddr *) psin;
+	ai->ai_canonname = NULL;
+	ai->ai_next = NULL;
+
+	return ai;
+}
+
+/*
+ * get address info for a single ipv4 address.
+ *
+ *	Bugs:	- servname can only be a number, not text.
+ */
+
+static int getaddr_info_single_addr(const char *service,
+				uint32 addr,
+				const struct addrinfo *hints,
+				struct addrinfo **res)
+{
+
+	struct addrinfo *ai = NULL;
+	struct in_addr ip;
+	unsigned short port = 0;
+
+	if (service) {
+		port = (unsigned short)atoi(service);
+	}
+	ip.s_addr = htonl(addr);
+
+	ai = alloc_entry(hints, ip, port);
+	if (!ai) {
+		return EAI_MEMORY;
+	}
+
+	/* If we're asked for the canonical name,
+	 * make sure it returns correctly. */
+	if (!(hints->ai_flags & AI_NUMERICSERV) &&
+			hints->ai_flags & AI_CANONNAME) {
+		int err;
+		if (addr == INADDR_LOOPBACK || addr == INADDR_ANY) {
+			ai->ai_canonname = get_my_canon_name(&err);
+		} else {
+			ai->ai_canonname =
+			get_canon_name_from_addr(ip,&err);
+		}
+		if (ai->ai_canonname == NULL) {
+			freeaddrinfo(ai);
+			return err;
+		}
+	}
+
+	*res = ai;
+	return 0;
+}
+
+/*
+ * get address info for multiple ipv4 addresses.
+ *
+ *	Bugs:	- servname can only be a number, not text.
+ */
+
+static int getaddr_info_name(const char *node,
+				const char *service,
+				const struct addrinfo *hints,
+				struct addrinfo **res)
+{
+	struct addrinfo *listp = NULL, *prevp = NULL;
+	char **pptr = NULL;
+	int err;
+	struct hostent *hp = NULL;
+	unsigned short port = 0;
+
+	if (service) {
+		port = (unsigned short)atoi(service);
+	}
+
+	hp = gethostbyname(node);
+	err = check_hostent_err(hp);
+	if (err) {
+		return err;
+	}
+
+	for(pptr = hp->h_addr_list; *pptr; pptr++) {
+		struct in_addr ip = *(struct in_addr *)*pptr;
+		struct addrinfo *ai = alloc_entry(hints, ip, port);
+
+		if (!ai) {
+			freeaddrinfo(listp);
+			return EAI_MEMORY;
+		}
+
+		if (!listp) {
+			listp = ai;
+			prevp = ai;
+			ai->ai_canonname = SMB_STRDUP(hp->h_name);
+			if (!ai->ai_canonname) {
+				freeaddrinfo(listp);
+				return EAI_MEMORY;
+			}
+		} else {
+			prevp->ai_next = ai;
+			prevp = ai;
+		}
+	}
+	*res = listp;
+	return 0;
+}
+
+/*
+ * get address info for ipv4 sockets.
+ *
+ *	Bugs:	- servname can only be a number, not text.
+ */
+
+int getaddrinfo(const char *node,
+		const char *service,
+		const struct addrinfo * hintp,
+		struct addrinfo ** res)
+{
+	struct addrinfo hints;
+
+	/* Setup the hints struct. */
+	if (hintp == NULL) {
+		memset(&hints, 0, sizeof(hints));
+		hints.ai_family = AF_INET;
+		hints.ai_socktype = SOCK_STREAM;
+	} else {
+		memcpy(&hints, hintp, sizeof(hints));
+	}
+
+	if (hints.ai_family != AF_INET && hints.ai_family != AF_UNSPEC) {
+		return EAI_FAMILY;
+	}
+
+	if (hints.ai_socktype == 0) {
+		hints.ai_socktype = SOCK_STREAM;
+	}
+
+	if (!node && !service) {
+		return EAI_NONAME;
+	}
+
+	if (node) {
+		if (node[0] == '\0') {
+			return getaddr_info_single_addr(service,
+					INADDR_ANY,
+					&hints,
+					res);
+		} else if (hints.ai_flags & AI_NUMERICHOST) {
+			struct in_addr ip;
+			if (inet_pton(AF_INET, node, &ip) <= 0)
+				return EAI_FAIL;
+			return getaddr_info_single_addr(service,
+					ntohl(ip.s_addr),
+					&hints,
+					res);
+		} else {
+			return getaddr_info_name(node,
+						service,
+						&hints,
+						res);
+		}
+	} else if (hints.ai_flags & AI_PASSIVE) {
+		return getaddr_info_single_addr(service,
+					INADDR_ANY,
+					&hints,
+					res);
+	}
+	return getaddr_info_single_addr(service,
+					INADDR_LOOPBACK,
+					&hints,
+					res);
+}
+
+
+void freeaddrinfo(struct addrinfo *res)
+{
+	struct addrinfo *next = NULL;
+
+	for (;res; res = next) {
+		next = res->ai_next;
+		if (res->ai_canonname) {
+			free(res->ai_canonname);
+		}
+		if (res->ai_addr) {
+			free(res->ai_addr);
+		}
+		free(res);
+	}
+}
+
+
+const char *gai_strerror(int errcode)
+{
+#ifdef HAVE_HSTRERROR
+	int			hcode;
+
+	switch (errcode)
+	{
+		case EAI_NONAME:
+			hcode = HOST_NOT_FOUND;
+			break;
+		case EAI_AGAIN:
+			hcode = TRY_AGAIN;
+			break;
+		case EAI_FAIL:
+		default:
+			hcode = NO_RECOVERY;
+			break;
+	}
+
+	return hstrerror(hcode);
+#else							/* !HAVE_HSTRERROR */
+
+	switch (errcode)
+	{
+		case EAI_NONAME:
+			return "Unknown host";
+		case EAI_AGAIN:
+			return "Host name lookup failure";
+#ifdef EAI_BADFLAGS
+		case EAI_BADFLAGS:
+			return "Invalid argument";
+#endif
+#ifdef EAI_FAMILY
+		case EAI_FAMILY:
+			return "Address family not supported";
+#endif
+#ifdef EAI_MEMORY
+		case EAI_MEMORY:
+			return "Not enough memory";
+#endif
+#ifdef EAI_NODATA
+		case EAI_NODATA:
+			return "No host data of that type was found";
+#endif
+#ifdef EAI_SERVICE
+		case EAI_SERVICE:
+			return "Class type not found";
+#endif
+#ifdef EAI_SOCKTYPE
+		case EAI_SOCKTYPE:
+			return "Socket type not supported";
+#endif
+		default:
+			return "Unknown server error";
+	}
+#endif   /* HAVE_HSTRERROR */
+}
+
+static int gethostnameinfo(const struct sockaddr *sa,
+			char *node,
+			size_t nodelen,
+			int flags)
+{
+	int ret = -1;
+	char *p = NULL;
+
+	if (!(flags & NI_NUMERICHOST)) {
+		struct hostent *hp = gethostbyaddr(
+				(void *)&((struct sockaddr_in *)sa)->sin_addr,
+				sizeof (struct in_addr),
+				sa->sa_family);
+		ret = check_hostent_err(hp);
+		if (ret == 0) {
+			/* Name looked up successfully. */
+			ret = snprintf(node, nodelen, "%s", hp->h_name);
+			if (ret < 0 || (size_t)ret >= nodelen) {
+				return EAI_MEMORY;
+			}
+			if (flags & NI_NOFQDN) {
+				p = strchr(node,'.');
+				if (p) {
+					*p = '\0';
+				}
+			}
+			return 0;
+		}
+
+		if (flags & NI_NAMEREQD) {
+			/* If we require a name and didn't get one,
+			 * automatically fail. */
+			return ret;
+		}
+		/* Otherwise just fall into the numeric host code... */
+	}
+	p = inet_ntoa(((struct sockaddr_in *)sa)->sin_addr);
+	ret = snprintf(node, nodelen, "%s", p);
+	if (ret < 0 || (size_t)ret >= nodelen) {
+		return EAI_MEMORY;
+	}
+	return 0;
+}
+
+static int getservicenameinfo(const struct sockaddr *sa,
+			char *service,
+			size_t servicelen,
+			int flags)
+{
+	int ret = -1;
+	int port = ntohs(((struct sockaddr_in *)sa)->sin_port);
+
+	if (!(flags & NI_NUMERICSERV)) {
+		struct servent *se = getservbyport(
+				port,
+				(flags & NI_DGRAM) ? "udp" : "tcp");
+		if (se && se->s_name) {
+			/* Service name looked up successfully. */
+			ret = snprintf(service, servicelen, "%s", se->s_name);
+			if (ret < 0 || (size_t)ret >= servicelen) {
+				return EAI_MEMORY;
+			}
+			return 0;
+		}
+		/* Otherwise just fall into the numeric service code... */
+	}
+	ret = snprintf(service, servicelen, "%d", port);
+	if (ret < 0 || (size_t)ret >= servicelen) {
+		return EAI_MEMORY;
+	}
+	return 0;
+}
+
+/*
+ * Convert an ipv4 address to a hostname.
+ *
+ * Bugs:	- No IPv6 support.
+ */
+int getnameinfo(const struct sockaddr *sa, socklen_t salen,
+			char *node, size_t nodelen,
+			char *service, size_t servicelen, int flags)
+{
+
+	/* Invalid arguments. */
+	if (sa == NULL || (node == NULL && service == NULL)) {
+		return EAI_FAIL;
+	}
+
+	if (sa->sa_family != AF_INET) {
+		return EAI_FAIL;
+	}
+
+	if (salen < (socklen_t)sizeof (struct sockaddr_in)) {
+		return EAI_FAIL;
+	}
+
+	if (node) {
+		int ret = gethostnameinfo(sa, node, nodelen, flags);
+		if (ret)
+			return ret;
+	}
+
+	if (service) {
+		return getservicenameinfo(sa, service, servicelen, flags);
+	}
+	return 0;
+}

lib/getpass.c

@@ -0,0 +1,72 @@
+/*
+ * An implementation of getpass for systems that lack one.
+ *
+ * Copyright (C) 2013 Roman Donchenko
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <termios.h>
+
+#include "rsync.h"
+
+char *getpass(const char *prompt)
+{
+	static char password[256];
+
+	BOOL tty_changed = False, read_success;
+	struct termios tty_old, tty_new;
+	FILE *in = stdin, *out = stderr;
+	FILE *tty = fopen("/dev/tty", "w+");
+
+	if (tty)
+		in = out = tty;
+
+	if (tcgetattr(fileno(in), &tty_old) == 0) {
+		tty_new = tty_old;
+		tty_new.c_lflag &= ~(ECHO | ISIG);
+
+		if (tcsetattr(fileno(in), TCSAFLUSH, &tty_new) == 0)
+			tty_changed = True;
+	}
+
+	if (!tty_changed)
+		fputs("(WARNING: will be visible) ", out);
+	fputs(prompt, out);
+	fflush(out);
+
+	read_success = fgets(password, sizeof password, in) != NULL;
+
+	/* Print the newline that hasn't been echoed. */
+	fputc('\n', out);
+
+	if (tty_changed)
+		tcsetattr(fileno(in), TCSAFLUSH, &tty_old);
+
+	if (tty)
+		fclose(tty);
+
+	if (read_success) {
+		/* Remove the trailing newline. */
+		size_t password_len = strlen(password);
+		if (password_len && password[password_len - 1] == '\n')
+			password[password_len - 1] = '\0';
+
+		return password;
+	}
+
+	return NULL;
+}

lib/inet_ntop.c

@@ -75,13 +75,14 @@ inet_ntop4(const unsigned char *src, char *dst, size_t size)
 {
 	static const char *fmt = "%u.%u.%u.%u";
 	char tmp[sizeof "255.255.255.255"];
+	size_t len;
 
-	if ((size_t)sprintf(tmp, fmt, src[0], src[1], src[2], src[3]) >= size)
-	{
+	len = snprintf(tmp, sizeof tmp, fmt, src[0], src[1], src[2], src[3]);
+	if (len >= size) {
 		errno = ENOSPC;
 		return (NULL);
 	}
-	strcpy(dst, tmp);
+	memcpy(dst, tmp, len + 1);
 
 	return (dst);
 }
@@ -106,7 +107,7 @@ inet_ntop6(const unsigned char *src, char *dst, size_t size)
 	char tmp[sizeof "ffff:ffff:ffff:ffff:ffff:ffff:255.255.255.255"], *tp;
 	struct { int base, len; } best, cur;
 	unsigned int words[NS_IN6ADDRSZ / NS_INT16SZ];
-	int i;
+	int i, inc;
 
 	/*
 	 * Preprocess:
@@ -157,13 +158,14 @@ inet_ntop6(const unsigned char *src, char *dst, size_t size)
 		/* Is this address an encapsulated IPv4? */
 		if (i == 6 && best.base == 0 &&
 		    (best.len == 6 || (best.len == 5 && words[5] == 0xffff))) {
-			if (!inet_ntop4(src+12, tp,
-					sizeof tmp - (tp - tmp)))
+			if (!inet_ntop4(src+12, tp, sizeof tmp - (tp - tmp)))
 				return (NULL);
 			tp += strlen(tp);
 			break;
 		}
-		tp += sprintf(tp, "%x", words[i]);
+		inc = snprintf(tp, 5, "%x", words[i]);
+		assert(inc < 5);
+		tp += inc;
 	}
 	/* Was it a trailing run of 0x00's? */
 	if (best.base != -1 && (best.base + best.len) ==
@@ -178,7 +180,7 @@ inet_ntop6(const unsigned char *src, char *dst, size_t size)
 		errno = ENOSPC;
 		return (NULL);
 	}
-	strcpy(dst, tmp);
+	memcpy(dst, tmp, tp - tmp);
 	return (dst);
 }
 #endif /* AF_INET6 */

lib/md-defines.h

@@ -0,0 +1,17 @@
+/* Keep this simple so both C and ASM can use it */
+
+#define MD4_DIGEST_LEN 16
+#define MD5_DIGEST_LEN 16
+#define MAX_DIGEST_LEN MD5_DIGEST_LEN
+
+#define CSUM_CHUNK 64
+
+#define CSUM_NONE 0
+#define CSUM_MD4_ARCHAIC 1
+#define CSUM_MD4_BUSTED 2
+#define CSUM_MD4_OLD 3
+#define CSUM_MD4 4
+#define CSUM_MD5 5
+#define CSUM_XXH64 6
+#define CSUM_XXH3_64 7
+#define CSUM_XXH3_128 8

lib/md5-asm-x86_64.S

@@ -0,0 +1,701 @@
+/*
+ * x86-64 optimized assembler MD5 implementation
+ *
+ * Author: Marc Bevand, 2004
+ *
+ * This code was placed in the public domain by the author. The original
+ * publication can be found at:
+ *
+ * https://www.zorinaq.com/papers/md5-amd64.html
+ */
+/*
+ * No modifications were made aside from changing the function and file names.
+ * The MD5_CTX structure as expected here (from OpenSSL) is binary compatible
+ * with the md_context used by rsync, for the fields accessed.
+ *
+ * Benchmarks (in MB/s)            C     ASM
+ * - Intel Atom D2700            302     334
+ * - Intel i7-7700hq             351     376
+ * - AMD ThreadRipper 2950x      728     784
+ *
+ * The original code was also incorporated into OpenSSL. It has since been
+ * modified there. Those changes have not been made here due to licensing
+ * incompatibilities. Benchmarks of those changes on the above CPUs did not
+ * show any significant difference in performance, though.
+ */
+
+#include "config.h"
+#include "md-defines.h"
+
+#if !defined USE_OPENSSL && CSUM_CHUNK == 64
+
+#ifdef __APPLE__
+#define md5_process_asm _md5_process_asm
+#endif
+
+.text
+.align 16
+
+.globl md5_process_asm
+md5_process_asm:
+	push	%rbp
+	push	%rbx
+	push	%r12
+	push	%r13			# not really useful (r13 is unused)
+	push	%r14
+	push	%r15
+
+	# rdi = arg #1 (ctx, MD5_CTX pointer)
+	# rsi = arg #2 (ptr, data pointer)
+	# rdx = arg #3 (nbr, number of 16-word blocks to process)
+	mov	%rdi,		%rbp	# rbp = ctx
+	shl	$6,		%rdx	# rdx = nbr in bytes
+	lea	(%rsi,%rdx),	%rdi	# rdi = end
+	mov	0*4(%rbp),	%eax	# eax = ctx->A
+	mov	1*4(%rbp),	%ebx	# ebx = ctx->B
+	mov	2*4(%rbp),	%ecx	# ecx = ctx->C
+	mov	3*4(%rbp),	%edx	# edx = ctx->D
+	# end is 'rdi'
+	# ptr is 'rsi'
+	# A is 'eax'
+	# B is 'ebx'
+	# C is 'ecx'
+	# D is 'edx'
+
+	cmp	%rdi,		%rsi		# cmp end with ptr
+	je	1f				# jmp if ptr == end
+
+	# BEGIN of loop over 16-word blocks
+2:	# save old values of A, B, C, D
+	mov	%eax,		%r8d
+	mov	%ebx,		%r9d
+	mov	%ecx,		%r14d
+	mov	%edx,		%r15d
+ mov	0*4(%rsi),	%r10d		/* (NEXT STEP) X[0] */
+ mov	%edx,		%r11d		/* (NEXT STEP) z' = %edx */
+	xor	%ecx,		%r11d		/* y ^ ... */
+	lea	-680876936(%eax,%r10d),%eax		/* Const + dst + ... */
+	and	%ebx,		%r11d		/* x & ... */
+	xor	%edx,		%r11d		/* z ^ ... */
+	mov	1*4(%rsi),%r10d		/* (NEXT STEP) X[1] */
+	add	%r11d,		%eax		/* dst += ... */
+	rol	$7,		%eax		/* dst <<< s */
+	mov	%ecx,		%r11d		/* (NEXT STEP) z' = %ecx */
+	add	%ebx,		%eax		/* dst += x */
+	xor	%ebx,		%r11d		/* y ^ ... */
+	lea	-389564586(%edx,%r10d),%edx		/* Const + dst + ... */
+	and	%eax,		%r11d		/* x & ... */
+	xor	%ecx,		%r11d		/* z ^ ... */
+	mov	2*4(%rsi),%r10d		/* (NEXT STEP) X[2] */
+	add	%r11d,		%edx		/* dst += ... */
+	rol	$12,		%edx		/* dst <<< s */
+	mov	%ebx,		%r11d		/* (NEXT STEP) z' = %ebx */
+	add	%eax,		%edx		/* dst += x */
+	xor	%eax,		%r11d		/* y ^ ... */
+	lea	606105819(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	and	%edx,		%r11d		/* x & ... */
+	xor	%ebx,		%r11d		/* z ^ ... */
+	mov	3*4(%rsi),%r10d		/* (NEXT STEP) X[3] */
+	add	%r11d,		%ecx		/* dst += ... */
+	rol	$17,		%ecx		/* dst <<< s */
+	mov	%eax,		%r11d		/* (NEXT STEP) z' = %eax */
+	add	%edx,		%ecx		/* dst += x */
+	xor	%edx,		%r11d		/* y ^ ... */
+	lea	-1044525330(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	and	%ecx,		%r11d		/* x & ... */
+	xor	%eax,		%r11d		/* z ^ ... */
+	mov	4*4(%rsi),%r10d		/* (NEXT STEP) X[4] */
+	add	%r11d,		%ebx		/* dst += ... */
+	rol	$22,		%ebx		/* dst <<< s */
+	mov	%edx,		%r11d		/* (NEXT STEP) z' = %edx */
+	add	%ecx,		%ebx		/* dst += x */
+	xor	%ecx,		%r11d		/* y ^ ... */
+	lea	-176418897(%eax,%r10d),%eax		/* Const + dst + ... */
+	and	%ebx,		%r11d		/* x & ... */
+	xor	%edx,		%r11d		/* z ^ ... */
+	mov	5*4(%rsi),%r10d		/* (NEXT STEP) X[5] */
+	add	%r11d,		%eax		/* dst += ... */
+	rol	$7,		%eax		/* dst <<< s */
+	mov	%ecx,		%r11d		/* (NEXT STEP) z' = %ecx */
+	add	%ebx,		%eax		/* dst += x */
+	xor	%ebx,		%r11d		/* y ^ ... */
+	lea	1200080426(%edx,%r10d),%edx		/* Const + dst + ... */
+	and	%eax,		%r11d		/* x & ... */
+	xor	%ecx,		%r11d		/* z ^ ... */
+	mov	6*4(%rsi),%r10d		/* (NEXT STEP) X[6] */
+	add	%r11d,		%edx		/* dst += ... */
+	rol	$12,		%edx		/* dst <<< s */
+	mov	%ebx,		%r11d		/* (NEXT STEP) z' = %ebx */
+	add	%eax,		%edx		/* dst += x */
+	xor	%eax,		%r11d		/* y ^ ... */
+	lea	-1473231341(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	and	%edx,		%r11d		/* x & ... */
+	xor	%ebx,		%r11d		/* z ^ ... */
+	mov	7*4(%rsi),%r10d		/* (NEXT STEP) X[7] */
+	add	%r11d,		%ecx		/* dst += ... */
+	rol	$17,		%ecx		/* dst <<< s */
+	mov	%eax,		%r11d		/* (NEXT STEP) z' = %eax */
+	add	%edx,		%ecx		/* dst += x */
+	xor	%edx,		%r11d		/* y ^ ... */
+	lea	-45705983(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	and	%ecx,		%r11d		/* x & ... */
+	xor	%eax,		%r11d		/* z ^ ... */
+	mov	8*4(%rsi),%r10d		/* (NEXT STEP) X[8] */
+	add	%r11d,		%ebx		/* dst += ... */
+	rol	$22,		%ebx		/* dst <<< s */
+	mov	%edx,		%r11d		/* (NEXT STEP) z' = %edx */
+	add	%ecx,		%ebx		/* dst += x */
+	xor	%ecx,		%r11d		/* y ^ ... */
+	lea	1770035416(%eax,%r10d),%eax		/* Const + dst + ... */
+	and	%ebx,		%r11d		/* x & ... */
+	xor	%edx,		%r11d		/* z ^ ... */
+	mov	9*4(%rsi),%r10d		/* (NEXT STEP) X[9] */
+	add	%r11d,		%eax		/* dst += ... */
+	rol	$7,		%eax		/* dst <<< s */
+	mov	%ecx,		%r11d		/* (NEXT STEP) z' = %ecx */
+	add	%ebx,		%eax		/* dst += x */
+	xor	%ebx,		%r11d		/* y ^ ... */
+	lea	-1958414417(%edx,%r10d),%edx		/* Const + dst + ... */
+	and	%eax,		%r11d		/* x & ... */
+	xor	%ecx,		%r11d		/* z ^ ... */
+	mov	10*4(%rsi),%r10d		/* (NEXT STEP) X[10] */
+	add	%r11d,		%edx		/* dst += ... */
+	rol	$12,		%edx		/* dst <<< s */
+	mov	%ebx,		%r11d		/* (NEXT STEP) z' = %ebx */
+	add	%eax,		%edx		/* dst += x */
+	xor	%eax,		%r11d		/* y ^ ... */
+	lea	-42063(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	and	%edx,		%r11d		/* x & ... */
+	xor	%ebx,		%r11d		/* z ^ ... */
+	mov	11*4(%rsi),%r10d		/* (NEXT STEP) X[11] */
+	add	%r11d,		%ecx		/* dst += ... */
+	rol	$17,		%ecx		/* dst <<< s */
+	mov	%eax,		%r11d		/* (NEXT STEP) z' = %eax */
+	add	%edx,		%ecx		/* dst += x */
+	xor	%edx,		%r11d		/* y ^ ... */
+	lea	-1990404162(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	and	%ecx,		%r11d		/* x & ... */
+	xor	%eax,		%r11d		/* z ^ ... */
+	mov	12*4(%rsi),%r10d		/* (NEXT STEP) X[12] */
+	add	%r11d,		%ebx		/* dst += ... */
+	rol	$22,		%ebx		/* dst <<< s */
+	mov	%edx,		%r11d		/* (NEXT STEP) z' = %edx */
+	add	%ecx,		%ebx		/* dst += x */
+	xor	%ecx,		%r11d		/* y ^ ... */
+	lea	1804603682(%eax,%r10d),%eax		/* Const + dst + ... */
+	and	%ebx,		%r11d		/* x & ... */
+	xor	%edx,		%r11d		/* z ^ ... */
+	mov	13*4(%rsi),%r10d		/* (NEXT STEP) X[13] */
+	add	%r11d,		%eax		/* dst += ... */
+	rol	$7,		%eax		/* dst <<< s */
+	mov	%ecx,		%r11d		/* (NEXT STEP) z' = %ecx */
+	add	%ebx,		%eax		/* dst += x */
+	xor	%ebx,		%r11d		/* y ^ ... */
+	lea	-40341101(%edx,%r10d),%edx		/* Const + dst + ... */
+	and	%eax,		%r11d		/* x & ... */
+	xor	%ecx,		%r11d		/* z ^ ... */
+	mov	14*4(%rsi),%r10d		/* (NEXT STEP) X[14] */
+	add	%r11d,		%edx		/* dst += ... */
+	rol	$12,		%edx		/* dst <<< s */
+	mov	%ebx,		%r11d		/* (NEXT STEP) z' = %ebx */
+	add	%eax,		%edx		/* dst += x */
+	xor	%eax,		%r11d		/* y ^ ... */
+	lea	-1502002290(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	and	%edx,		%r11d		/* x & ... */
+	xor	%ebx,		%r11d		/* z ^ ... */
+	mov	15*4(%rsi),%r10d		/* (NEXT STEP) X[15] */
+	add	%r11d,		%ecx		/* dst += ... */
+	rol	$17,		%ecx		/* dst <<< s */
+	mov	%eax,		%r11d		/* (NEXT STEP) z' = %eax */
+	add	%edx,		%ecx		/* dst += x */
+	xor	%edx,		%r11d		/* y ^ ... */
+	lea	1236535329(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	and	%ecx,		%r11d		/* x & ... */
+	xor	%eax,		%r11d		/* z ^ ... */
+	mov	0*4(%rsi),%r10d		/* (NEXT STEP) X[0] */
+	add	%r11d,		%ebx		/* dst += ... */
+	rol	$22,		%ebx		/* dst <<< s */
+	mov	%edx,		%r11d		/* (NEXT STEP) z' = %edx */
+	add	%ecx,		%ebx		/* dst += x */
+ mov	1*4(%rsi),	%r10d		/* (NEXT STEP) X[1] */
+ mov	%edx,		%r11d		/* (NEXT STEP) z' = %edx */
+ mov	%edx,		%r12d		/* (NEXT STEP) z' = %edx */
+	not	%r11d				/* not z */
+	lea	-165796510(%eax,%r10d),%eax		/* Const + dst + ... */
+	and	%ebx,		%r12d		/* x & z */
+	and	%ecx,		%r11d		/* y & (not z) */
+	mov	6*4(%rsi),%r10d		/* (NEXT STEP) X[6] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%ecx,		%r11d		/* (NEXT STEP) z' = %ecx */
+	add	%r12d,		%eax		/* dst += ... */
+	mov	%ecx,		%r12d		/* (NEXT STEP) z' = %ecx */
+	rol	$5,		%eax		/* dst <<< s */
+	add	%ebx,		%eax		/* dst += x */
+	not	%r11d				/* not z */
+	lea	-1069501632(%edx,%r10d),%edx		/* Const + dst + ... */
+	and	%eax,		%r12d		/* x & z */
+	and	%ebx,		%r11d		/* y & (not z) */
+	mov	11*4(%rsi),%r10d		/* (NEXT STEP) X[11] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%ebx,		%r11d		/* (NEXT STEP) z' = %ebx */
+	add	%r12d,		%edx		/* dst += ... */
+	mov	%ebx,		%r12d		/* (NEXT STEP) z' = %ebx */
+	rol	$9,		%edx		/* dst <<< s */
+	add	%eax,		%edx		/* dst += x */
+	not	%r11d				/* not z */
+	lea	643717713(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	and	%edx,		%r12d		/* x & z */
+	and	%eax,		%r11d		/* y & (not z) */
+	mov	0*4(%rsi),%r10d		/* (NEXT STEP) X[0] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%eax,		%r11d		/* (NEXT STEP) z' = %eax */
+	add	%r12d,		%ecx		/* dst += ... */
+	mov	%eax,		%r12d		/* (NEXT STEP) z' = %eax */
+	rol	$14,		%ecx		/* dst <<< s */
+	add	%edx,		%ecx		/* dst += x */
+	not	%r11d				/* not z */
+	lea	-373897302(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	and	%ecx,		%r12d		/* x & z */
+	and	%edx,		%r11d		/* y & (not z) */
+	mov	5*4(%rsi),%r10d		/* (NEXT STEP) X[5] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%edx,		%r11d		/* (NEXT STEP) z' = %edx */
+	add	%r12d,		%ebx		/* dst += ... */
+	mov	%edx,		%r12d		/* (NEXT STEP) z' = %edx */
+	rol	$20,		%ebx		/* dst <<< s */
+	add	%ecx,		%ebx		/* dst += x */
+	not	%r11d				/* not z */
+	lea	-701558691(%eax,%r10d),%eax		/* Const + dst + ... */
+	and	%ebx,		%r12d		/* x & z */
+	and	%ecx,		%r11d		/* y & (not z) */
+	mov	10*4(%rsi),%r10d		/* (NEXT STEP) X[10] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%ecx,		%r11d		/* (NEXT STEP) z' = %ecx */
+	add	%r12d,		%eax		/* dst += ... */
+	mov	%ecx,		%r12d		/* (NEXT STEP) z' = %ecx */
+	rol	$5,		%eax		/* dst <<< s */
+	add	%ebx,		%eax		/* dst += x */
+	not	%r11d				/* not z */
+	lea	38016083(%edx,%r10d),%edx		/* Const + dst + ... */
+	and	%eax,		%r12d		/* x & z */
+	and	%ebx,		%r11d		/* y & (not z) */
+	mov	15*4(%rsi),%r10d		/* (NEXT STEP) X[15] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%ebx,		%r11d		/* (NEXT STEP) z' = %ebx */
+	add	%r12d,		%edx		/* dst += ... */
+	mov	%ebx,		%r12d		/* (NEXT STEP) z' = %ebx */
+	rol	$9,		%edx		/* dst <<< s */
+	add	%eax,		%edx		/* dst += x */
+	not	%r11d				/* not z */
+	lea	-660478335(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	and	%edx,		%r12d		/* x & z */
+	and	%eax,		%r11d		/* y & (not z) */
+	mov	4*4(%rsi),%r10d		/* (NEXT STEP) X[4] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%eax,		%r11d		/* (NEXT STEP) z' = %eax */
+	add	%r12d,		%ecx		/* dst += ... */
+	mov	%eax,		%r12d		/* (NEXT STEP) z' = %eax */
+	rol	$14,		%ecx		/* dst <<< s */
+	add	%edx,		%ecx		/* dst += x */
+	not	%r11d				/* not z */
+	lea	-405537848(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	and	%ecx,		%r12d		/* x & z */
+	and	%edx,		%r11d		/* y & (not z) */
+	mov	9*4(%rsi),%r10d		/* (NEXT STEP) X[9] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%edx,		%r11d		/* (NEXT STEP) z' = %edx */
+	add	%r12d,		%ebx		/* dst += ... */
+	mov	%edx,		%r12d		/* (NEXT STEP) z' = %edx */
+	rol	$20,		%ebx		/* dst <<< s */
+	add	%ecx,		%ebx		/* dst += x */
+	not	%r11d				/* not z */
+	lea	568446438(%eax,%r10d),%eax		/* Const + dst + ... */
+	and	%ebx,		%r12d		/* x & z */
+	and	%ecx,		%r11d		/* y & (not z) */
+	mov	14*4(%rsi),%r10d		/* (NEXT STEP) X[14] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%ecx,		%r11d		/* (NEXT STEP) z' = %ecx */
+	add	%r12d,		%eax		/* dst += ... */
+	mov	%ecx,		%r12d		/* (NEXT STEP) z' = %ecx */
+	rol	$5,		%eax		/* dst <<< s */
+	add	%ebx,		%eax		/* dst += x */
+	not	%r11d				/* not z */
+	lea	-1019803690(%edx,%r10d),%edx		/* Const + dst + ... */
+	and	%eax,		%r12d		/* x & z */
+	and	%ebx,		%r11d		/* y & (not z) */
+	mov	3*4(%rsi),%r10d		/* (NEXT STEP) X[3] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%ebx,		%r11d		/* (NEXT STEP) z' = %ebx */
+	add	%r12d,		%edx		/* dst += ... */
+	mov	%ebx,		%r12d		/* (NEXT STEP) z' = %ebx */
+	rol	$9,		%edx		/* dst <<< s */
+	add	%eax,		%edx		/* dst += x */
+	not	%r11d				/* not z */
+	lea	-187363961(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	and	%edx,		%r12d		/* x & z */
+	and	%eax,		%r11d		/* y & (not z) */
+	mov	8*4(%rsi),%r10d		/* (NEXT STEP) X[8] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%eax,		%r11d		/* (NEXT STEP) z' = %eax */
+	add	%r12d,		%ecx		/* dst += ... */
+	mov	%eax,		%r12d		/* (NEXT STEP) z' = %eax */
+	rol	$14,		%ecx		/* dst <<< s */
+	add	%edx,		%ecx		/* dst += x */
+	not	%r11d				/* not z */
+	lea	1163531501(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	and	%ecx,		%r12d		/* x & z */
+	and	%edx,		%r11d		/* y & (not z) */
+	mov	13*4(%rsi),%r10d		/* (NEXT STEP) X[13] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%edx,		%r11d		/* (NEXT STEP) z' = %edx */
+	add	%r12d,		%ebx		/* dst += ... */
+	mov	%edx,		%r12d		/* (NEXT STEP) z' = %edx */
+	rol	$20,		%ebx		/* dst <<< s */
+	add	%ecx,		%ebx		/* dst += x */
+	not	%r11d				/* not z */
+	lea	-1444681467(%eax,%r10d),%eax		/* Const + dst + ... */
+	and	%ebx,		%r12d		/* x & z */
+	and	%ecx,		%r11d		/* y & (not z) */
+	mov	2*4(%rsi),%r10d		/* (NEXT STEP) X[2] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%ecx,		%r11d		/* (NEXT STEP) z' = %ecx */
+	add	%r12d,		%eax		/* dst += ... */
+	mov	%ecx,		%r12d		/* (NEXT STEP) z' = %ecx */
+	rol	$5,		%eax		/* dst <<< s */
+	add	%ebx,		%eax		/* dst += x */
+	not	%r11d				/* not z */
+	lea	-51403784(%edx,%r10d),%edx		/* Const + dst + ... */
+	and	%eax,		%r12d		/* x & z */
+	and	%ebx,		%r11d		/* y & (not z) */
+	mov	7*4(%rsi),%r10d		/* (NEXT STEP) X[7] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%ebx,		%r11d		/* (NEXT STEP) z' = %ebx */
+	add	%r12d,		%edx		/* dst += ... */
+	mov	%ebx,		%r12d		/* (NEXT STEP) z' = %ebx */
+	rol	$9,		%edx		/* dst <<< s */
+	add	%eax,		%edx		/* dst += x */
+	not	%r11d				/* not z */
+	lea	1735328473(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	and	%edx,		%r12d		/* x & z */
+	and	%eax,		%r11d		/* y & (not z) */
+	mov	12*4(%rsi),%r10d		/* (NEXT STEP) X[12] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%eax,		%r11d		/* (NEXT STEP) z' = %eax */
+	add	%r12d,		%ecx		/* dst += ... */
+	mov	%eax,		%r12d		/* (NEXT STEP) z' = %eax */
+	rol	$14,		%ecx		/* dst <<< s */
+	add	%edx,		%ecx		/* dst += x */
+	not	%r11d				/* not z */
+	lea	-1926607734(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	and	%ecx,		%r12d		/* x & z */
+	and	%edx,		%r11d		/* y & (not z) */
+	mov	0*4(%rsi),%r10d		/* (NEXT STEP) X[0] */
+	or	%r11d,		%r12d		/* (y & (not z)) | (x & z) */
+	mov	%edx,		%r11d		/* (NEXT STEP) z' = %edx */
+	add	%r12d,		%ebx		/* dst += ... */
+	mov	%edx,		%r12d		/* (NEXT STEP) z' = %edx */
+	rol	$20,		%ebx		/* dst <<< s */
+	add	%ecx,		%ebx		/* dst += x */
+ mov	5*4(%rsi),	%r10d		/* (NEXT STEP) X[5] */
+ mov	%ecx,		%r11d		/* (NEXT STEP) y' = %ecx */
+	lea	-378558(%eax,%r10d),%eax		/* Const + dst + ... */
+	mov	8*4(%rsi),%r10d		/* (NEXT STEP) X[8] */
+	xor	%edx,		%r11d		/* z ^ ... */
+	xor	%ebx,		%r11d		/* x ^ ... */
+	add	%r11d,		%eax		/* dst += ... */
+	rol	$4,		%eax		/* dst <<< s */
+	mov	%ebx,		%r11d		/* (NEXT STEP) y' = %ebx */
+	add	%ebx,		%eax		/* dst += x */
+	lea	-2022574463(%edx,%r10d),%edx		/* Const + dst + ... */
+	mov	11*4(%rsi),%r10d		/* (NEXT STEP) X[11] */
+	xor	%ecx,		%r11d		/* z ^ ... */
+	xor	%eax,		%r11d		/* x ^ ... */
+	add	%r11d,		%edx		/* dst += ... */
+	rol	$11,		%edx		/* dst <<< s */
+	mov	%eax,		%r11d		/* (NEXT STEP) y' = %eax */
+	add	%eax,		%edx		/* dst += x */
+	lea	1839030562(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	mov	14*4(%rsi),%r10d		/* (NEXT STEP) X[14] */
+	xor	%ebx,		%r11d		/* z ^ ... */
+	xor	%edx,		%r11d		/* x ^ ... */
+	add	%r11d,		%ecx		/* dst += ... */
+	rol	$16,		%ecx		/* dst <<< s */
+	mov	%edx,		%r11d		/* (NEXT STEP) y' = %edx */
+	add	%edx,		%ecx		/* dst += x */
+	lea	-35309556(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	mov	1*4(%rsi),%r10d		/* (NEXT STEP) X[1] */
+	xor	%eax,		%r11d		/* z ^ ... */
+	xor	%ecx,		%r11d		/* x ^ ... */
+	add	%r11d,		%ebx		/* dst += ... */
+	rol	$23,		%ebx		/* dst <<< s */
+	mov	%ecx,		%r11d		/* (NEXT STEP) y' = %ecx */
+	add	%ecx,		%ebx		/* dst += x */
+	lea	-1530992060(%eax,%r10d),%eax		/* Const + dst + ... */
+	mov	4*4(%rsi),%r10d		/* (NEXT STEP) X[4] */
+	xor	%edx,		%r11d		/* z ^ ... */
+	xor	%ebx,		%r11d		/* x ^ ... */
+	add	%r11d,		%eax		/* dst += ... */
+	rol	$4,		%eax		/* dst <<< s */
+	mov	%ebx,		%r11d		/* (NEXT STEP) y' = %ebx */
+	add	%ebx,		%eax		/* dst += x */
+	lea	1272893353(%edx,%r10d),%edx		/* Const + dst + ... */
+	mov	7*4(%rsi),%r10d		/* (NEXT STEP) X[7] */
+	xor	%ecx,		%r11d		/* z ^ ... */
+	xor	%eax,		%r11d		/* x ^ ... */
+	add	%r11d,		%edx		/* dst += ... */
+	rol	$11,		%edx		/* dst <<< s */
+	mov	%eax,		%r11d		/* (NEXT STEP) y' = %eax */
+	add	%eax,		%edx		/* dst += x */
+	lea	-155497632(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	mov	10*4(%rsi),%r10d		/* (NEXT STEP) X[10] */
+	xor	%ebx,		%r11d		/* z ^ ... */
+	xor	%edx,		%r11d		/* x ^ ... */
+	add	%r11d,		%ecx		/* dst += ... */
+	rol	$16,		%ecx		/* dst <<< s */
+	mov	%edx,		%r11d		/* (NEXT STEP) y' = %edx */
+	add	%edx,		%ecx		/* dst += x */
+	lea	-1094730640(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	mov	13*4(%rsi),%r10d		/* (NEXT STEP) X[13] */
+	xor	%eax,		%r11d		/* z ^ ... */
+	xor	%ecx,		%r11d		/* x ^ ... */
+	add	%r11d,		%ebx		/* dst += ... */
+	rol	$23,		%ebx		/* dst <<< s */
+	mov	%ecx,		%r11d		/* (NEXT STEP) y' = %ecx */
+	add	%ecx,		%ebx		/* dst += x */
+	lea	681279174(%eax,%r10d),%eax		/* Const + dst + ... */
+	mov	0*4(%rsi),%r10d		/* (NEXT STEP) X[0] */
+	xor	%edx,		%r11d		/* z ^ ... */
+	xor	%ebx,		%r11d		/* x ^ ... */
+	add	%r11d,		%eax		/* dst += ... */
+	rol	$4,		%eax		/* dst <<< s */
+	mov	%ebx,		%r11d		/* (NEXT STEP) y' = %ebx */
+	add	%ebx,		%eax		/* dst += x */
+	lea	-358537222(%edx,%r10d),%edx		/* Const + dst + ... */
+	mov	3*4(%rsi),%r10d		/* (NEXT STEP) X[3] */
+	xor	%ecx,		%r11d		/* z ^ ... */
+	xor	%eax,		%r11d		/* x ^ ... */
+	add	%r11d,		%edx		/* dst += ... */
+	rol	$11,		%edx		/* dst <<< s */
+	mov	%eax,		%r11d		/* (NEXT STEP) y' = %eax */
+	add	%eax,		%edx		/* dst += x */
+	lea	-722521979(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	mov	6*4(%rsi),%r10d		/* (NEXT STEP) X[6] */
+	xor	%ebx,		%r11d		/* z ^ ... */
+	xor	%edx,		%r11d		/* x ^ ... */
+	add	%r11d,		%ecx		/* dst += ... */
+	rol	$16,		%ecx		/* dst <<< s */
+	mov	%edx,		%r11d		/* (NEXT STEP) y' = %edx */
+	add	%edx,		%ecx		/* dst += x */
+	lea	76029189(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	mov	9*4(%rsi),%r10d		/* (NEXT STEP) X[9] */
+	xor	%eax,		%r11d		/* z ^ ... */
+	xor	%ecx,		%r11d		/* x ^ ... */
+	add	%r11d,		%ebx		/* dst += ... */
+	rol	$23,		%ebx		/* dst <<< s */
+	mov	%ecx,		%r11d		/* (NEXT STEP) y' = %ecx */
+	add	%ecx,		%ebx		/* dst += x */
+	lea	-640364487(%eax,%r10d),%eax		/* Const + dst + ... */
+	mov	12*4(%rsi),%r10d		/* (NEXT STEP) X[12] */
+	xor	%edx,		%r11d		/* z ^ ... */
+	xor	%ebx,		%r11d		/* x ^ ... */
+	add	%r11d,		%eax		/* dst += ... */
+	rol	$4,		%eax		/* dst <<< s */
+	mov	%ebx,		%r11d		/* (NEXT STEP) y' = %ebx */
+	add	%ebx,		%eax		/* dst += x */
+	lea	-421815835(%edx,%r10d),%edx		/* Const + dst + ... */
+	mov	15*4(%rsi),%r10d		/* (NEXT STEP) X[15] */
+	xor	%ecx,		%r11d		/* z ^ ... */
+	xor	%eax,		%r11d		/* x ^ ... */
+	add	%r11d,		%edx		/* dst += ... */
+	rol	$11,		%edx		/* dst <<< s */
+	mov	%eax,		%r11d		/* (NEXT STEP) y' = %eax */
+	add	%eax,		%edx		/* dst += x */
+	lea	530742520(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	mov	2*4(%rsi),%r10d		/* (NEXT STEP) X[2] */
+	xor	%ebx,		%r11d		/* z ^ ... */
+	xor	%edx,		%r11d		/* x ^ ... */
+	add	%r11d,		%ecx		/* dst += ... */
+	rol	$16,		%ecx		/* dst <<< s */
+	mov	%edx,		%r11d		/* (NEXT STEP) y' = %edx */
+	add	%edx,		%ecx		/* dst += x */
+	lea	-995338651(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	mov	0*4(%rsi),%r10d		/* (NEXT STEP) X[0] */
+	xor	%eax,		%r11d		/* z ^ ... */
+	xor	%ecx,		%r11d		/* x ^ ... */
+	add	%r11d,		%ebx		/* dst += ... */
+	rol	$23,		%ebx		/* dst <<< s */
+	mov	%ecx,		%r11d		/* (NEXT STEP) y' = %ecx */
+	add	%ecx,		%ebx		/* dst += x */
+ mov	0*4(%rsi),	%r10d		/* (NEXT STEP) X[0] */
+ mov	$0xffffffff,	%r11d
+ xor	%edx,		%r11d		/* (NEXT STEP) not z' = not %edx*/
+	lea	-198630844(%eax,%r10d),%eax		/* Const + dst + ... */
+	or	%ebx,		%r11d		/* x | ... */
+	xor	%ecx,		%r11d		/* y ^ ... */
+	add	%r11d,		%eax		/* dst += ... */
+	mov	7*4(%rsi),%r10d		/* (NEXT STEP) X[7] */
+	mov	$0xffffffff,	%r11d
+	rol	$6,		%eax		/* dst <<< s */
+	xor	%ecx,		%r11d		/* (NEXT STEP) not z' = not %ecx */
+	add	%ebx,		%eax		/* dst += x */
+	lea	1126891415(%edx,%r10d),%edx		/* Const + dst + ... */
+	or	%eax,		%r11d		/* x | ... */
+	xor	%ebx,		%r11d		/* y ^ ... */
+	add	%r11d,		%edx		/* dst += ... */
+	mov	14*4(%rsi),%r10d		/* (NEXT STEP) X[14] */
+	mov	$0xffffffff,	%r11d
+	rol	$10,		%edx		/* dst <<< s */
+	xor	%ebx,		%r11d		/* (NEXT STEP) not z' = not %ebx */
+	add	%eax,		%edx		/* dst += x */
+	lea	-1416354905(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	or	%edx,		%r11d		/* x | ... */
+	xor	%eax,		%r11d		/* y ^ ... */
+	add	%r11d,		%ecx		/* dst += ... */
+	mov	5*4(%rsi),%r10d		/* (NEXT STEP) X[5] */
+	mov	$0xffffffff,	%r11d
+	rol	$15,		%ecx		/* dst <<< s */
+	xor	%eax,		%r11d		/* (NEXT STEP) not z' = not %eax */
+	add	%edx,		%ecx		/* dst += x */
+	lea	-57434055(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	or	%ecx,		%r11d		/* x | ... */
+	xor	%edx,		%r11d		/* y ^ ... */
+	add	%r11d,		%ebx		/* dst += ... */
+	mov	12*4(%rsi),%r10d		/* (NEXT STEP) X[12] */
+	mov	$0xffffffff,	%r11d
+	rol	$21,		%ebx		/* dst <<< s */
+	xor	%edx,		%r11d		/* (NEXT STEP) not z' = not %edx */
+	add	%ecx,		%ebx		/* dst += x */
+	lea	1700485571(%eax,%r10d),%eax		/* Const + dst + ... */
+	or	%ebx,		%r11d		/* x | ... */
+	xor	%ecx,		%r11d		/* y ^ ... */
+	add	%r11d,		%eax		/* dst += ... */
+	mov	3*4(%rsi),%r10d		/* (NEXT STEP) X[3] */
+	mov	$0xffffffff,	%r11d
+	rol	$6,		%eax		/* dst <<< s */
+	xor	%ecx,		%r11d		/* (NEXT STEP) not z' = not %ecx */
+	add	%ebx,		%eax		/* dst += x */
+	lea	-1894986606(%edx,%r10d),%edx		/* Const + dst + ... */
+	or	%eax,		%r11d		/* x | ... */
+	xor	%ebx,		%r11d		/* y ^ ... */
+	add	%r11d,		%edx		/* dst += ... */
+	mov	10*4(%rsi),%r10d		/* (NEXT STEP) X[10] */
+	mov	$0xffffffff,	%r11d
+	rol	$10,		%edx		/* dst <<< s */
+	xor	%ebx,		%r11d		/* (NEXT STEP) not z' = not %ebx */
+	add	%eax,		%edx		/* dst += x */
+	lea	-1051523(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	or	%edx,		%r11d		/* x | ... */
+	xor	%eax,		%r11d		/* y ^ ... */
+	add	%r11d,		%ecx		/* dst += ... */
+	mov	1*4(%rsi),%r10d		/* (NEXT STEP) X[1] */
+	mov	$0xffffffff,	%r11d
+	rol	$15,		%ecx		/* dst <<< s */
+	xor	%eax,		%r11d		/* (NEXT STEP) not z' = not %eax */
+	add	%edx,		%ecx		/* dst += x */
+	lea	-2054922799(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	or	%ecx,		%r11d		/* x | ... */
+	xor	%edx,		%r11d		/* y ^ ... */
+	add	%r11d,		%ebx		/* dst += ... */
+	mov	8*4(%rsi),%r10d		/* (NEXT STEP) X[8] */
+	mov	$0xffffffff,	%r11d
+	rol	$21,		%ebx		/* dst <<< s */
+	xor	%edx,		%r11d		/* (NEXT STEP) not z' = not %edx */
+	add	%ecx,		%ebx		/* dst += x */
+	lea	1873313359(%eax,%r10d),%eax		/* Const + dst + ... */
+	or	%ebx,		%r11d		/* x | ... */
+	xor	%ecx,		%r11d		/* y ^ ... */
+	add	%r11d,		%eax		/* dst += ... */
+	mov	15*4(%rsi),%r10d		/* (NEXT STEP) X[15] */
+	mov	$0xffffffff,	%r11d
+	rol	$6,		%eax		/* dst <<< s */
+	xor	%ecx,		%r11d		/* (NEXT STEP) not z' = not %ecx */
+	add	%ebx,		%eax		/* dst += x */
+	lea	-30611744(%edx,%r10d),%edx		/* Const + dst + ... */
+	or	%eax,		%r11d		/* x | ... */
+	xor	%ebx,		%r11d		/* y ^ ... */
+	add	%r11d,		%edx		/* dst += ... */
+	mov	6*4(%rsi),%r10d		/* (NEXT STEP) X[6] */
+	mov	$0xffffffff,	%r11d
+	rol	$10,		%edx		/* dst <<< s */
+	xor	%ebx,		%r11d		/* (NEXT STEP) not z' = not %ebx */
+	add	%eax,		%edx		/* dst += x */
+	lea	-1560198380(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	or	%edx,		%r11d		/* x | ... */
+	xor	%eax,		%r11d		/* y ^ ... */
+	add	%r11d,		%ecx		/* dst += ... */
+	mov	13*4(%rsi),%r10d		/* (NEXT STEP) X[13] */
+	mov	$0xffffffff,	%r11d
+	rol	$15,		%ecx		/* dst <<< s */
+	xor	%eax,		%r11d		/* (NEXT STEP) not z' = not %eax */
+	add	%edx,		%ecx		/* dst += x */
+	lea	1309151649(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	or	%ecx,		%r11d		/* x | ... */
+	xor	%edx,		%r11d		/* y ^ ... */
+	add	%r11d,		%ebx		/* dst += ... */
+	mov	4*4(%rsi),%r10d		/* (NEXT STEP) X[4] */
+	mov	$0xffffffff,	%r11d
+	rol	$21,		%ebx		/* dst <<< s */
+	xor	%edx,		%r11d		/* (NEXT STEP) not z' = not %edx */
+	add	%ecx,		%ebx		/* dst += x */
+	lea	-145523070(%eax,%r10d),%eax		/* Const + dst + ... */
+	or	%ebx,		%r11d		/* x | ... */
+	xor	%ecx,		%r11d		/* y ^ ... */
+	add	%r11d,		%eax		/* dst += ... */
+	mov	11*4(%rsi),%r10d		/* (NEXT STEP) X[11] */
+	mov	$0xffffffff,	%r11d
+	rol	$6,		%eax		/* dst <<< s */
+	xor	%ecx,		%r11d		/* (NEXT STEP) not z' = not %ecx */
+	add	%ebx,		%eax		/* dst += x */
+	lea	-1120210379(%edx,%r10d),%edx		/* Const + dst + ... */
+	or	%eax,		%r11d		/* x | ... */
+	xor	%ebx,		%r11d		/* y ^ ... */
+	add	%r11d,		%edx		/* dst += ... */
+	mov	2*4(%rsi),%r10d		/* (NEXT STEP) X[2] */
+	mov	$0xffffffff,	%r11d
+	rol	$10,		%edx		/* dst <<< s */
+	xor	%ebx,		%r11d		/* (NEXT STEP) not z' = not %ebx */
+	add	%eax,		%edx		/* dst += x */
+	lea	718787259(%ecx,%r10d),%ecx		/* Const + dst + ... */
+	or	%edx,		%r11d		/* x | ... */
+	xor	%eax,		%r11d		/* y ^ ... */
+	add	%r11d,		%ecx		/* dst += ... */
+	mov	9*4(%rsi),%r10d		/* (NEXT STEP) X[9] */
+	mov	$0xffffffff,	%r11d
+	rol	$15,		%ecx		/* dst <<< s */
+	xor	%eax,		%r11d		/* (NEXT STEP) not z' = not %eax */
+	add	%edx,		%ecx		/* dst += x */
+	lea	-343485551(%ebx,%r10d),%ebx		/* Const + dst + ... */
+	or	%ecx,		%r11d		/* x | ... */
+	xor	%edx,		%r11d		/* y ^ ... */
+	add	%r11d,		%ebx		/* dst += ... */
+	mov	0*4(%rsi),%r10d		/* (NEXT STEP) X[0] */
+	mov	$0xffffffff,	%r11d
+	rol	$21,		%ebx		/* dst <<< s */
+	xor	%edx,		%r11d		/* (NEXT STEP) not z' = not %edx */
+	add	%ecx,		%ebx		/* dst += x */
+	# add old values of A, B, C, D
+	add	%r8d,	%eax
+	add	%r9d,	%ebx
+	add	%r14d,	%ecx
+	add	%r15d,	%edx
+
+	# loop control
+	add	$64,		%rsi		# ptr += 64
+	cmp	%rdi,		%rsi		# cmp end with ptr
+	jb	2b				# jmp if ptr < end
+	# END of loop over 16-word blocks
+1:
+	mov	%eax,		0*4(%rbp)	# ctx->A = A
+	mov	%ebx,		1*4(%rbp)	# ctx->B = B
+	mov	%ecx,		2*4(%rbp)	# ctx->C = C
+	mov	%edx,		3*4(%rbp)	# ctx->D = D
+
+	pop	%r15
+	pop	%r14
+	pop	%r13				# not really useful (r13 is unused)
+	pop	%r12
+	pop	%rbx
+	pop	%rbp
+	ret
+
+#endif /* !USE_OPENSSL ... */

lib/md5.c

@@ -0,0 +1,320 @@
+/*
+ * RFC 1321 compliant MD5 implementation
+ *
+ * Copyright (C) 2001-2003 Christophe Devine
+ * Copyright (C) 2007-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
+
+#include "rsync.h"
+
+#ifndef USE_OPENSSL
+void md5_begin(md_context *ctx)
+{
+	ctx->A = 0x67452301;
+	ctx->B = 0xEFCDAB89;
+	ctx->C = 0x98BADCFE;
+	ctx->D = 0x10325476;
+
+	ctx->totalN = ctx->totalN2 = 0;
+}
+
+static void md5_process(md_context *ctx, const uchar data[CSUM_CHUNK])
+{
+	uint32 X[16], A, B, C, D;
+
+	A = ctx->A;
+	B = ctx->B;
+	C = ctx->C;
+	D = ctx->D;
+
+	X[0] = IVALu(data, 0);
+	X[1] = IVALu(data, 4);
+	X[2] = IVALu(data, 8);
+	X[3] = IVALu(data, 12);
+	X[4] = IVALu(data, 16);
+	X[5] = IVALu(data, 20);
+	X[6] = IVALu(data, 24);
+	X[7] = IVALu(data, 28);
+	X[8] = IVALu(data, 32);
+	X[9] = IVALu(data, 36);
+	X[10] = IVALu(data, 40);
+	X[11] = IVALu(data, 44);
+	X[12] = IVALu(data, 48);
+	X[13] = IVALu(data, 52);
+	X[14] = IVALu(data, 56);
+	X[15] = IVALu(data, 60);
+
+#define S(x,n) ((x << n) | ((x & 0xFFFFFFFF) >> (32 - n)))
+
+#define P(a,b,c,d,k,s,t) a += F(b,c,d) + X[k] + t, a = S(a,s) + b
+
+#define F(x,y,z) (z ^ (x & (y ^ z)))
+
+	P(A, B, C, D,  0,  7, 0xD76AA478);
+	P(D, A, B, C,  1, 12, 0xE8C7B756);
+	P(C, D, A, B,  2, 17, 0x242070DB);
+	P(B, C, D, A,  3, 22, 0xC1BDCEEE);
+	P(A, B, C, D,  4,  7, 0xF57C0FAF);
+	P(D, A, B, C,  5, 12, 0x4787C62A);
+	P(C, D, A, B,  6, 17, 0xA8304613);
+	P(B, C, D, A,  7, 22, 0xFD469501);
+	P(A, B, C, D,  8,  7, 0x698098D8);
+	P(D, A, B, C,  9, 12, 0x8B44F7AF);
+	P(C, D, A, B, 10, 17, 0xFFFF5BB1);
+	P(B, C, D, A, 11, 22, 0x895CD7BE);
+	P(A, B, C, D, 12,  7, 0x6B901122);
+	P(D, A, B, C, 13, 12, 0xFD987193);
+	P(C, D, A, B, 14, 17, 0xA679438E);
+	P(B, C, D, A, 15, 22, 0x49B40821);
+
+#undef F
+#define F(x,y,z) (y ^ (z & (x ^ y)))
+
+	P(A, B, C, D,  1,  5, 0xF61E2562);
+	P(D, A, B, C,  6,  9, 0xC040B340);
+	P(C, D, A, B, 11, 14, 0x265E5A51);
+	P(B, C, D, A,  0, 20, 0xE9B6C7AA);
+	P(A, B, C, D,  5,  5, 0xD62F105D);
+	P(D, A, B, C, 10,  9, 0x02441453);
+	P(C, D, A, B, 15, 14, 0xD8A1E681);
+	P(B, C, D, A,  4, 20, 0xE7D3FBC8);
+	P(A, B, C, D,  9,  5, 0x21E1CDE6);
+	P(D, A, B, C, 14,  9, 0xC33707D6);
+	P(C, D, A, B,  3, 14, 0xF4D50D87);
+	P(B, C, D, A,  8, 20, 0x455A14ED);
+	P(A, B, C, D, 13,  5, 0xA9E3E905);
+	P(D, A, B, C,  2,  9, 0xFCEFA3F8);
+	P(C, D, A, B,  7, 14, 0x676F02D9);
+	P(B, C, D, A, 12, 20, 0x8D2A4C8A);
+
+#undef F
+#define F(x,y,z) (x ^ y ^ z)
+
+	P(A, B, C, D,  5,  4, 0xFFFA3942);
+	P(D, A, B, C,  8, 11, 0x8771F681);
+	P(C, D, A, B, 11, 16, 0x6D9D6122);
+	P(B, C, D, A, 14, 23, 0xFDE5380C);
+	P(A, B, C, D,  1,  4, 0xA4BEEA44);
+	P(D, A, B, C,  4, 11, 0x4BDECFA9);
+	P(C, D, A, B,  7, 16, 0xF6BB4B60);
+	P(B, C, D, A, 10, 23, 0xBEBFBC70);
+	P(A, B, C, D, 13,  4, 0x289B7EC6);
+	P(D, A, B, C,  0, 11, 0xEAA127FA);
+	P(C, D, A, B,  3, 16, 0xD4EF3085);
+	P(B, C, D, A,  6, 23, 0x04881D05);
+	P(A, B, C, D,  9,  4, 0xD9D4D039);
+	P(D, A, B, C, 12, 11, 0xE6DB99E5);
+	P(C, D, A, B, 15, 16, 0x1FA27CF8);
+	P(B, C, D, A,  2, 23, 0xC4AC5665);
+
+#undef F
+#define F(x,y,z) (y ^ (x | ~z))
+
+	P(A, B, C, D,  0,  6, 0xF4292244);
+	P(D, A, B, C,  7, 10, 0x432AFF97);
+	P(C, D, A, B, 14, 15, 0xAB9423A7);
+	P(B, C, D, A,  5, 21, 0xFC93A039);
+	P(A, B, C, D, 12,  6, 0x655B59C3);
+	P(D, A, B, C,  3, 10, 0x8F0CCC92);
+	P(C, D, A, B, 10, 15, 0xFFEFF47D);
+	P(B, C, D, A,  1, 21, 0x85845DD1);
+	P(A, B, C, D,  8,  6, 0x6FA87E4F);
+	P(D, A, B, C, 15, 10, 0xFE2CE6E0);
+	P(C, D, A, B,  6, 15, 0xA3014314);
+	P(B, C, D, A, 13, 21, 0x4E0811A1);
+	P(A, B, C, D,  4,  6, 0xF7537E82);
+	P(D, A, B, C, 11, 10, 0xBD3AF235);
+	P(C, D, A, B,  2, 15, 0x2AD7D2BB);
+	P(B, C, D, A,  9, 21, 0xEB86D391);
+
+#undef F
+
+	ctx->A += A;
+	ctx->B += B;
+	ctx->C += C;
+	ctx->D += D;
+}
+
+#if defined HAVE_ASM && CSUM_CHUNK == 64
+extern void md5_process_asm(md_context *ctx, const void *data, size_t num);
+#endif
+
+void md5_update(md_context *ctx, const uchar *input, uint32 length)
+{
+	uint32 left, fill;
+
+	if (!length)
+		return;
+
+	left = ctx->totalN & 0x3F;
+	fill = CSUM_CHUNK - left;
+
+	ctx->totalN += length;
+	ctx->totalN &= 0xFFFFFFFF;
+
+	if (ctx->totalN < length)
+		ctx->totalN2++;
+
+	if (left && length >= fill) {
+		memcpy(ctx->buffer + left, input, fill);
+		md5_process(ctx, ctx->buffer);
+		length -= fill;
+		input  += fill;
+		left = 0;
+	}
+
+#if defined HAVE_ASM && CSUM_CHUNK == 64
+	if (length >= CSUM_CHUNK) {
+		uint32 chunks = length / CSUM_CHUNK;
+		md5_process_asm(ctx, input, chunks);
+		length -= chunks * CSUM_CHUNK;
+		input += chunks * CSUM_CHUNK;
+	}
+#else
+	while (length >= CSUM_CHUNK) {
+		md5_process(ctx, input);
+		length -= CSUM_CHUNK;
+		input  += CSUM_CHUNK;
+	}
+#endif
+
+	if (length)
+		memcpy(ctx->buffer + left, input, length);
+}
+
+static uchar md5_padding[CSUM_CHUNK] = { 0x80 };
+
+void md5_result(md_context *ctx, uchar digest[MD5_DIGEST_LEN])
+{
+	uint32 last, padn;
+	uint32 high, low;
+	uchar msglen[8];
+
+	high = (ctx->totalN >> 29)
+	     | (ctx->totalN2 <<  3);
+	low  = (ctx->totalN <<  3);
+
+	SIVALu(msglen, 0, low);
+	SIVALu(msglen, 4, high);
+
+	last = ctx->totalN & 0x3F;
+	padn = last < 56 ? 56 - last : 120 - last;
+
+	md5_update(ctx, md5_padding, padn);
+	md5_update(ctx, msglen, 8);
+
+	SIVALu(digest, 0, ctx->A);
+	SIVALu(digest, 4, ctx->B);
+	SIVALu(digest, 8, ctx->C);
+	SIVALu(digest, 12, ctx->D);
+}
+#endif
+
+#ifdef TEST_MD5
+
+void get_md5(uchar *out, const uchar *input, int n)
+{
+	md_context ctx;
+	md5_begin(&ctx);
+	md5_update(&ctx, input, n);
+	md5_result(&ctx, out);
+}
+
+#include <stdlib.h>
+#include <stdio.h>
+
+/*
+ * those are the standard RFC 1321 test vectors
+ */
+
+static struct {
+    char *str, *md5;
+} tests[] = {
+ { "",
+   "d41d8cd98f00b204e9800998ecf8427e" },
+ { "a",
+   "0cc175b9c0f1b6a831c399e269772661" },
+ { "abc",
+   "900150983cd24fb0d6963f7d28e17f72" },
+ { "message digest",
+   "f96b697d7cb7938d525a2f31aaf161d0" },
+ { "abcdefghijklmnopqrstuvwxyz",
+   "c3fcd3d76192e4007dfb496cca67e13b" },
+ { "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789",
+   "d174ab98d277d9f5a5611c2c9f419d9f" },
+ { "12345678901234567890123456789012345678901234567890123456789012345678901234567890",
+   "57edf4a22be3c955ac49da2e2107b67a" },
+ { NULL, NULL }
+};
+
+int main(int argc, char *argv[])
+{
+	FILE *f;
+	int i, j;
+	char output[33];
+	md_context ctx;
+	uchar buf[1000];
+	uchar md5sum[MD5_DIGEST_LEN];
+
+	if (argc < 2) {
+		printf("\nMD5 Validation Tests:\n\n");
+
+		for (i = 0; tests[i].str; i++) {
+			char *str = tests[i].str;
+			char *chk = tests[i].md5;
+
+			printf("  Test %d ", i + 1);
+
+			get_md5(md5sum, str, strlen(str));
+
+			for (j = 0; j < MD5_DIGEST_LEN; j++)
+				sprintf(output + j * 2, "%02x", md5sum[j]);
+
+			if (memcmp(output, chk, 32)) {
+				printf("failed!\n");
+				return 1;
+			}
+
+			printf("passed.\n");
+		}
+
+		printf("\n");
+		return 0;
+	}
+
+	while (--argc) {
+		if (!(f = fopen(*++argv, "rb"))) {
+			perror("fopen");
+			return 1;
+		}
+
+		md5_begin(&ctx);
+
+		while ((i = fread(buf, 1, sizeof buf, f)) > 0)
+			md5_update(&ctx, buf, i);
+
+		md5_result(&ctx, md5sum);
+
+		for (j = 0; j < MD5_DIGEST_LEN; j++)
+			printf("%02x", md5sum[j]);
+
+		printf("  %s\n", *argv);
+	}
+
+	return 0;
+}
+
+#endif

lib/mdfour.c

@@ -1,32 +1,32 @@
 /* 
-   Unix SMB/Netbios implementation.
-   Version 1.9.
-   a implementation of MD4 designed for use in the SMB authentication protocol
-   Copyright (C) Andrew Tridgell 1997-1998.
-   
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
+ * Unix SMB/Netbios implementation.
+ * Version 1.9.
+ * An implementation of MD4 designed for use in the SMB authentication protocol.
+ *
+ * Copyright (C) 1997-1998 Andrew Tridgell
+ * Copyright (C) 2005-2020 Wayne Davison
+ * 
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
 #include "rsync.h"
 
 /* NOTE: This code makes no attempt to be fast! 
+ *
+ * It assumes that a int is at least 32 bits long. */
 
-   It assumes that a int is at least 32 bits long
-*/
-
-static struct mdfour *m;
+static md_context *m;
 
 #define MASK32 (0xffffffff)
 
@@ -48,32 +48,31 @@ static void mdfour64(uint32 *M)
 	A = m->A; B = m->B; C = m->C; D = m->D; 
 	AA = A; BB = B; CC = C; DD = D;
 
-        ROUND1(A,B,C,D,  0,  3);  ROUND1(D,A,B,C,  1,  7);  
+	ROUND1(A,B,C,D,  0,  3);  ROUND1(D,A,B,C,  1,  7);  
 	ROUND1(C,D,A,B,  2, 11);  ROUND1(B,C,D,A,  3, 19);
-        ROUND1(A,B,C,D,  4,  3);  ROUND1(D,A,B,C,  5,  7);  
+	ROUND1(A,B,C,D,  4,  3);  ROUND1(D,A,B,C,  5,  7);  
 	ROUND1(C,D,A,B,  6, 11);  ROUND1(B,C,D,A,  7, 19);
-        ROUND1(A,B,C,D,  8,  3);  ROUND1(D,A,B,C,  9,  7);  
+	ROUND1(A,B,C,D,  8,  3);  ROUND1(D,A,B,C,  9,  7);  
 	ROUND1(C,D,A,B, 10, 11);  ROUND1(B,C,D,A, 11, 19);
-        ROUND1(A,B,C,D, 12,  3);  ROUND1(D,A,B,C, 13,  7);  
+	ROUND1(A,B,C,D, 12,  3);  ROUND1(D,A,B,C, 13,  7);  
 	ROUND1(C,D,A,B, 14, 11);  ROUND1(B,C,D,A, 15, 19);	
 
-
-        ROUND2(A,B,C,D,  0,  3);  ROUND2(D,A,B,C,  4,  5);  
+	ROUND2(A,B,C,D,  0,  3);  ROUND2(D,A,B,C,  4,  5);  
 	ROUND2(C,D,A,B,  8,  9);  ROUND2(B,C,D,A, 12, 13);
-        ROUND2(A,B,C,D,  1,  3);  ROUND2(D,A,B,C,  5,  5);  
+	ROUND2(A,B,C,D,  1,  3);  ROUND2(D,A,B,C,  5,  5);  
 	ROUND2(C,D,A,B,  9,  9);  ROUND2(B,C,D,A, 13, 13);
-        ROUND2(A,B,C,D,  2,  3);  ROUND2(D,A,B,C,  6,  5);  
+	ROUND2(A,B,C,D,  2,  3);  ROUND2(D,A,B,C,  6,  5);  
 	ROUND2(C,D,A,B, 10,  9);  ROUND2(B,C,D,A, 14, 13);
-        ROUND2(A,B,C,D,  3,  3);  ROUND2(D,A,B,C,  7,  5);  
+	ROUND2(A,B,C,D,  3,  3);  ROUND2(D,A,B,C,  7,  5);  
 	ROUND2(C,D,A,B, 11,  9);  ROUND2(B,C,D,A, 15, 13);
 
 	ROUND3(A,B,C,D,  0,  3);  ROUND3(D,A,B,C,  8,  9);  
 	ROUND3(C,D,A,B,  4, 11);  ROUND3(B,C,D,A, 12, 15);
-        ROUND3(A,B,C,D,  2,  3);  ROUND3(D,A,B,C, 10,  9);  
+	ROUND3(A,B,C,D,  2,  3);  ROUND3(D,A,B,C, 10,  9);  
 	ROUND3(C,D,A,B,  6, 11);  ROUND3(B,C,D,A, 14, 15);
-        ROUND3(A,B,C,D,  1,  3);  ROUND3(D,A,B,C,  9,  9);  
+	ROUND3(A,B,C,D,  1,  3);  ROUND3(D,A,B,C,  9,  9);  
 	ROUND3(C,D,A,B,  5, 11);  ROUND3(B,C,D,A, 13, 15);
-        ROUND3(A,B,C,D,  3,  3);  ROUND3(D,A,B,C, 11,  9);  
+	ROUND3(A,B,C,D,  3,  3);  ROUND3(D,A,B,C, 11,  9);  
 	ROUND3(C,D,A,B,  7, 11);  ROUND3(B,C,D,A, 15, 15);
 
 	A += AA; B += BB; 
@@ -85,16 +84,17 @@ static void mdfour64(uint32 *M)
 	m->A = A; m->B = B; m->C = C; m->D = D;
 }
 
-static void copy64(uint32 *M, unsigned char *in)
+static void copy64(uint32 *M, const uchar *in)
 {
 	int i;
 
-	for (i=0;i<16;i++)
-		M[i] = (in[i*4+3]<<24) | (in[i*4+2]<<16) |
-			(in[i*4+1]<<8) | (in[i*4+0]<<0);
+	for (i = 0; i < MD4_DIGEST_LEN; i++) {
+		M[i] = (in[i*4+3] << 24) | (in[i*4+2] << 16)
+		     | (in[i*4+1] << 8)  | (in[i*4+0] << 0);
+	}
 }
 
-static void copy4(unsigned char *out,uint32 x)
+static void copy4(uchar *out,uint32 x)
 {
 	out[0] = x&0xFF;
 	out[1] = (x>>8)&0xFF;
@@ -102,7 +102,7 @@ static void copy4(unsigned char *out,uint32 x)
 	out[3] = (x>>24)&0xFF;
 }
 
-void mdfour_begin(struct mdfour *md)
+void mdfour_begin(md_context *md)
 {
 	md->A = 0x67452301;
 	md->B = 0xefcdab89;
@@ -112,27 +112,26 @@ void mdfour_begin(struct mdfour *md)
 	md->totalN2 = 0;
 }
 
-
-static void mdfour_tail(unsigned char *in, uint32 n)
+static void mdfour_tail(const uchar *in, uint32 length)
 {
-	unsigned char buf[128];
+	uchar buf[128];
 	uint32 M[16];
 	extern int protocol_version;
 
 	/*
 	 * Count total number of bits, modulo 2^64
 	 */
-	m->totalN += n << 3;
-	if (m->totalN < (n << 3)) {
+	m->totalN += length << 3;
+	if (m->totalN < (length << 3))
 		m->totalN2++;
-	}
-	m->totalN2 += n >> 29;
+	m->totalN2 += length >> 29;
 
 	memset(buf, 0, 128);
-	if (n) memcpy(buf, in, n);
-	buf[n] = 0x80;
+	if (length)
+		memcpy(buf, in, length);
+	buf[length] = 0x80;
 
-	if (n <= 55) {
+	if (length <= 55) {
 		copy4(buf+56, m->totalN);
 		/*
 		 * Prior to protocol version 27 only the number of bits
@@ -140,9 +139,8 @@ static void mdfour_tail(unsigned char *in, uint32 n)
 		 * of bits modulo 2^64, which was fixed starting with
 		 * protocol version 27.
 		 */
-		if (protocol_version >= 27) {
+		if (protocol_version >= 27)
 			copy4(buf+60, m->totalN2);
-		}
 		copy64(M, buf);
 		mdfour64(M);
 	} else {
@@ -153,9 +151,8 @@ static void mdfour_tail(unsigned char *in, uint32 n)
 		 * of bits modulo 2^64, which was fixed starting with
 		 * protocol version 27.
 		 */
-		if (protocol_version >= 27) {
+		if (protocol_version >= 27)
 			copy4(buf+124, m->totalN2); 
-		}
 		copy64(M, buf);
 		mdfour64(M);
 		copy64(M, buf+64);
@@ -163,56 +160,56 @@ static void mdfour_tail(unsigned char *in, uint32 n)
 	}
 }
 
-void mdfour_update(struct mdfour *md, unsigned char *in, uint32 n)
+void mdfour_update(md_context *md, const uchar *in, uint32 length)
 {
 	uint32 M[16];
 
 	m = md;
 
-	if (n == 0) mdfour_tail(in, n);
+	if (length == 0)
+		mdfour_tail(in, length);
 
-	while (n >= 64) {
+	while (length >= 64) {
 		copy64(M, in);
 		mdfour64(M);
 		in += 64;
-		n -= 64;
+		length -= 64;
 		m->totalN += 64 << 3;
-		if (m->totalN < 64 << 3) {
+		if (m->totalN < 64 << 3)
 			m->totalN2++;
-		}
 	}
 
-	if (n) mdfour_tail(in, n);
+	if (length)
+		mdfour_tail(in, length);
 }
 
-
-void mdfour_result(struct mdfour *md, unsigned char *out)
+void mdfour_result(md_context *md, uchar digest[MD4_DIGEST_LEN])
 {
 	m = md;
 
-	copy4(out, m->A);
-	copy4(out+4, m->B);
-	copy4(out+8, m->C);
-	copy4(out+12, m->D);
-}
-
-
-void mdfour(unsigned char *out, unsigned char *in, int n)
-{
-	struct mdfour md;
-	mdfour_begin(&md);
-	mdfour_update(&md, in, n);
-	mdfour_result(&md, out);
+	copy4(digest, m->A);
+	copy4(digest+4, m->B);
+	copy4(digest+8, m->C);
+	copy4(digest+12, m->D);
 }
 
 #ifdef TEST_MDFOUR
+
+void mdfour(uchar digest[MD4_DIGEST_LEN], uchar *in, int length)
+{
+	md_context md;
+	mdfour_begin(&md);
+	mdfour_update(&md, in, length);
+	mdfour_result(&md, digest);
+}
+
 int protocol_version = 28;
 
 static void file_checksum1(char *fname)
 {
 	int fd, i, was_multiple_of_64 = 1;
-	struct mdfour md;
-	unsigned char buf[64*1024], sum[16];
+	md_context md;
+	uchar buf[64*1024], sum[MD4_DIGEST_LEN];
 	
 	fd = open(fname,O_RDONLY);
 	if (fd == -1) {
@@ -223,7 +220,7 @@ static void file_checksum1(char *fname)
 	mdfour_begin(&md);
 
 	while (1) {
-		int n = read(fd, buf, sizeof(buf));
+		int n = read(fd, buf, sizeof buf);
 		if (n <= 0)
 			break;
 		was_multiple_of_64 = !(n % 64);
@@ -236,54 +233,15 @@ static void file_checksum1(char *fname)
 
 	mdfour_result(&md, sum);
 
-	for (i=0;i<16;i++)
+	for (i = 0; i < MD4_DIGEST_LEN; i++)
 		printf("%02X", sum[i]);
 	printf("\n");
 }
 
-#if 0
-#include "../md4.h"
-
-static void file_checksum2(char *fname)
-{
-	int fd, i;
-	MDstruct md;
-	unsigned char buf[64], sum[16];
-
-	fd = open(fname,O_RDONLY);
-	if (fd == -1) {
-		perror("fname");
-		exit(1);
-	}
-	
-	MDbegin(&md);
-
-	while (1) {
-		int n = read(fd, buf, sizeof(buf));
-		if (n <= 0) break;
-		MDupdate(&md, buf, n*8);
-	}
-
-	if (!md.done) {
-		MDupdate(&md, buf, 0);
-	}
-
-	close(fd);
-
-	memcpy(sum, md.buffer, 16);
-
-	for (i=0;i<16;i++)
-		printf("%02X", sum[i]);
-	printf("\n");
-}
-#endif
-
  int main(int argc, char *argv[])
 {
-	file_checksum1(argv[1]);
-#if 0
-	file_checksum2(argv[1]);
-#endif
+	while (--argc)
+		file_checksum1(*++argv);
 	return 0;
 }
 #endif

lib/mdfour.h

@@ -1,35 +0,0 @@
-/* 
-   Unix SMB/Netbios implementation.
-   Version 1.9.
-   a implementation of MD4 designed for use in the SMB authentication protocol
-   Copyright (C) Andrew Tridgell 1997-1998.
-   
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
-
-struct mdfour {
-	uint32 A, B, C, D;
-	uint32 totalN;          /* bit count, lower 32 bits */
-	uint32 totalN2;         /* bit count, upper 32 bits */
-};
-
-void mdfour_begin(struct mdfour *md);
-void mdfour_update(struct mdfour *md, unsigned char *in, uint32 n);
-void mdfour_result(struct mdfour *md, unsigned char *out);
-void mdfour(unsigned char *out, unsigned char *in, int n);
-
-
-
-

lib/mdigest.h

@@ -0,0 +1,29 @@
+/* The include file for both the MD4 and MD5 routines. */
+
+#ifdef USE_OPENSSL
+#include "openssl/md4.h"
+#include "openssl/md5.h"
+#endif
+#include "md-defines.h"
+
+typedef struct {
+	uint32 A, B, C, D;
+	uint32 totalN;          /* bit count, lower 32 bits */
+	uint32 totalN2;         /* bit count, upper 32 bits */
+	uchar buffer[CSUM_CHUNK];
+} md_context;
+
+void mdfour_begin(md_context *md);
+void mdfour_update(md_context *md, const uchar *in, uint32 length);
+void mdfour_result(md_context *md, uchar digest[MD4_DIGEST_LEN]);
+
+#ifndef USE_OPENSSL
+#define MD5_CTX md_context
+#define MD5_Init md5_begin
+#define MD5_Update md5_update
+#define MD5_Final(digest, cptr) md5_result(cptr, digest)
+
+void md5_begin(md_context *ctx);
+void md5_update(md_context *ctx, const uchar *input, uint32 length);
+void md5_result(md_context *ctx, uchar digest[MD5_DIGEST_LEN]);
+#endif

lib/permstring.c

@@ -1,41 +1,39 @@
-/* 
-   Copyright (C) Andrew Tridgell 1996
-   Copyright (C) Paul Mackerras 1996
-   Copyright (C) 2001 by Martin Pool <mbp@samba.org>
-   
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
+/*
+ * A single utility routine.
+ *
+ * Copyright (C) 1996 Andrew Tridgell
+ * Copyright (C) 1996 Paul Mackerras
+ * Copyright (C) 2001 Martin Pool <mbp@samba.org>
+ * Copyright (C) 2003-2019 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
 #include "rsync.h"
 
-/**
- * Produce a string representation of Unix mode bits like that used by
- * ls(1).
- *
- * @param buf buffer of at least 11 characters
- **/
-void permstring(char *perms,
-		int mode)
+/* Produce a string representation of Unix mode bits like that used by ls(1).
+ * The "buf" buffer must be at least 11 characters. */
+void permstring(char *perms, mode_t mode)
 {
 	static const char *perm_map = "rwxrwxrwx";
 	int i;
 
-	strcpy(perms, "----------");
-	
-	for (i=0;i<9;i++) {
-		if (mode & (1<<i)) perms[9-i] = perm_map[8-i];
+	strlcpy(perms, "----------", 11);
+
+	for (i = 0; i < 9; i++) {
+		if (mode & (1 << i))
+			perms[9-i] = perm_map[8-i];
 	}
 
 	/* Handle setuid/sticky bits.  You might think the indices are
@@ -46,18 +44,22 @@ void permstring(char *perms,
 
 	if (mode & S_ISGID)
 		perms[6] = (mode & S_IXGRP) ? 's' : 'S';
-	
+
 #ifdef S_ISVTX
 	if (mode & S_ISVTX)
 		perms[9] = (mode & S_IXOTH) ? 't' : 'T';
 #endif
-		
-	if (S_ISLNK(mode)) perms[0] = 'l';
-	if (S_ISDIR(mode)) perms[0] = 'd';
-	if (S_ISBLK(mode)) perms[0] = 'b';
-	if (S_ISCHR(mode)) perms[0] = 'c';
-	if (S_ISSOCK(mode)) perms[0] = 's';
-	if (S_ISFIFO(mode)) perms[0] = 'p';
-}
 
-	
+	if (S_ISDIR(mode))
+		perms[0] = 'd';
+	else if (S_ISLNK(mode))
+		perms[0] = 'l';
+	else if (S_ISBLK(mode))
+		perms[0] = 'b';
+	else if (S_ISCHR(mode))
+		perms[0] = 'c';
+	else if (S_ISSOCK(mode))
+		perms[0] = 's';
+	else if (S_ISFIFO(mode))
+		perms[0] = 'p';
+}

lib/permstring.h

@@ -1,3 +1,3 @@
 #define PERMSTRING_SIZE 11
 
-void permstring(char *perms, int mode);
+void permstring(char *perms, mode_t mode);

lib/pool_alloc.3

@@ -28,7 +28,7 @@ See \fB\\$1\fP in \fB\\$2\fP for details.
 ..
 .TH POOL_ALLOC 3
 .SH NAME
-pool_alloc, pool_free, pool_talloc, pool_tfree, pool_create, pool_destroy
+pool_alloc, pool_free, pool_free_old, pool_talloc, pool_tfree, pool_create, pool_destroy, pool_boundary
 \- Allocate and free memory in managed allocation pools.
 .SH SYNOPSIS
 .B #include "pool_alloc.h"
@@ -39,42 +39,45 @@ pool_alloc, pool_free, pool_talloc, pool_tfree, pool_create, pool_destroy
 
 \fBvoid *pool_alloc(struct alloc_pool *\fIpool\fB, size_t \fIsize\fB, char *\fImsg\fB);
 
-\fBvoid pool_free(struct alloc_pool *\fIpool\fB, sise_t \fIsize\fB, void *\fIaddr\fB);
+\fBvoid pool_free(struct alloc_pool *\fIpool\fB, size_t \fIsize\fB, void *\fIaddr\fB);
+
+\fBvoid pool_free_old(struct alloc_pool *\fIpool\fB, void *\fIaddr\fB);
 
 \fBvoid *pool_talloc(struct alloc_pool *\fIpool\fB, \fItype\fB), int \fIcount\fB, char *\fImsg\fB);
 
 \fBvoid pool_tfree(struct alloc_pool *\fIpool\fB, \fItype\fB, int \fIcount\fB, void *\fIaddr\fB);
+
+\fBvoid pool_boundary(struct alloc_pool *\fIpool\fB, sise_t \fIsize\fB);
 .SH DESCRIPTION
 .P
 The pool allocation routines use
 .B malloc()
 for underlying memory management.
-What allocation pools do is cause
-memory within a given pool to be in large contigious blocks
-(called extents) that when freed will be reusable.  Unlike
-.B malloc()
+What allocation pools do is cause memory within a given pool
+to be allocated in large contiguous blocks
+(called extents) that will be reusable when freed.  Unlike
+.BR malloc() ,
 the allocations are not managed individually.
-Instead each extent tracks the total free memory within the
+Instead, each extent tracks the total free memory within the
 extent.  Each extent can either be used to allocate memory
 or to manage the freeing of memory within that extent.
 When an extent has less free memory than a given
-allocation request or when the first request to free
-memory within that extent is received the extent ceases to
-be used for allocation.
+allocation request, the current extent ceases to be used
+for allocation.  See also the
+.B pool_boundary()
+function.
 .P
 This form of memory management is suited to large numbers of small
 related allocations that are held for a while
 and then freed as a group.
 Because the
-underlying allocations are done in large contigious extents
-when an extent is freed it releases a large enough
-contigious block of memory to be useful to subsequent
+underlying allocations are done in large contiguous extents,
+when an extent is freed, it can release a large enough
+contiguous block of memory to allow the memory to be returned
+to the OS for use by whatever program needs it.
+You can allocate from one or more memory pools and/or
 .B malloc()
-and
-.B pool_alloc()
-calls even if allocations from other pools or from
-.B malloc()
-are made between allocations from a given pool.
+all at the same time without interfering with how pools work.
 .P
 .B pool_create()
 Creates an allocation pool for subsequent calls to the pool
@@ -90,22 +93,46 @@ Specifying
 .B 0
 for
 .I quantum
-Will produce a quantum that should meet maximal allignment
+will produce a quantum that should meet maximal alignment
 on most platforms.
-If the
-.B POOL_QALIGN
-.I flag
-is set allocations will be aligned to addresses that are a
+Unless
+.B POOL_NO_QALIGN
+is set in the
+.IR flags ,
+allocations will be aligned to addresses that are a
 multiple of
 .IR quantum .
-If the
+A
+.B NULL
+may be specified for the
+.I bomb
+function pointer if it is not needed.  (See the
+.B pool_alloc()
+function for how it is used.)
+If
 .B POOL_CLEAR
-.I flag
-is set all allocations from the pool will be zero filled.
+is set in the
+.IR flags ,
+all allocations from the pool will be initialized to zeros.
+If either
+.B POOL_PREPEND
+or
+.B POOL_INTERN
+is specified in the
+.IR flags ,
+each extent's data structure will be allocated at the start of the
+.IR size -length
+buffer (rather than as a separate, non-pool allocation), with the
+former extending the
+.I size
+to hold the structure, and the latter subtracting the structure's
+length from the indicated
+.IR size .
 .P
 .B pool_destroy()
-destroys an allocation pool and frees all memory allocated
-in that pool.
+destroys an allocation
+.I pool
+and frees all its associated memory.
 .P
 .B pool_alloc()
 allocates
@@ -115,57 +142,97 @@ bytes from the specified
 If
 .I size
 is
-.B 0
+.BR 0 ,
 .I quantum
-bytes will be freed.
-If the requested memory cannot be allocated
-.B pool_alloc()
-will call
+bytes will be allocated.
+If the pool has been created without
+.BR POOL_NO_QALIGN ,
+every chunk of memory that is returned will be suitably aligned.
+You can use this with the default
+.I quantum
+size to ensure that all memory can store a variable of any type.
+If the requested memory cannot be allocated, the
 .I bomb()
-function, if defined, with
+function will be called with
 .I msg
-as it's sole argument and
+as its sole argument (if the function was defined at the time
+the pool was created), and then a
 .B NULL
-will be returned.
+address is returned (assuming that the bomb function didn't exit).
 .P
 .B pool_free()
 frees
 .I size
-bytes pointed to by
+bytes pointed to by an
 .I addr
-previously allocated in the specified
+that was previously allocated in the specified
 .IR pool .
-The memory freed within an extent will not be reusable until
-all of the memory in that extent has been freed but 
-depending on the order in which the
-allocations are freed some extents may be released for reuse
-while others are still in use.
 If
 .I size
 is
-.B 0
+.BR 0 ,
 .I quantum
 bytes will be freed.
+The memory freed within an extent will not be reusable until
+all of the memory in that extent has been freed with one
+exception: the most recent pool allocation may be freed back
+into the pool prior to making any further allocations.
+If enough free calls are made to indicate that an extent has no
+remaining allocated objects (as computed by the total freed size for
+an extent), its memory will be completely freed back to the system.
 If
 .I addr
 is
-.B 0
-no memory will be freed but subsequent allocations will come
+.BR NULL ,
+no memory will be freed, but subsequent allocations will come
 from a new extent.
 .P
-.B pool_talloc()
-is a macro that take a
-.I type
+.B pool_free_old()
+takes a boundary
+.I addr
+value that was returned by
+.B pool_boundary()
+and frees up any extents in the
+.I pool
+that have data allocated from that point backward in time.
+NOTE: you must NOT mix calls to both
+.B pool_free
 and
+.B pool_free_old
+on the same pool!
+.P
+.B pool_boundary()
+asks for a boundary value that can be sent to 
+.B pool_free_old()
+at a later time to free up all memory allocated prior to a particular
+moment in time.
+If the extent that holds the boundary point has allocations from after the
+boundary point, it will not be freed until a future
+.B pool_free_old()
+call encompasses the entirety of the extent's data.
+If
+.I len
+is non-zero, the call will also check if the active extent has at least
+that much free memory available in it, and if not, it will mark the
+extent as inactive, forcing a new extent to be used for future allocations.
+(You can specify -1 for
+.I len
+if you want to force a new extent to start.)
+.P
+.B pool_talloc()
+is a macro that takes a
+.I type
+and a
 .I count
-instead of
-.I size
-and will cast the return value to the correct type.
+instead of a
+.IR size .
+It casts the return value to the correct pointer type.
 .P
 .B pool_tfree
-is a macro to free memory previously allocated in the
-specified
-.IR pool .
+is a macro that calls
+.B pool_free
+on memory that was allocated by
+.BR pool_talloc() .
 .SH RETURN VALUE
 .B pool_create()
 returns a pointer to
@@ -176,9 +243,6 @@ and
 .B pool_talloc()
 return pointers to the allocated memory,
 or NULL if the request fails.
-For each extent so long as no allocations are smaller than varaible
-allignment requirements this pointer will be suitably
-alligned for any kind of variable.
 The return type of
 .B pool_alloc()
 will normally require casting to the desired type but
@@ -186,7 +250,12 @@ will normally require casting to the desired type but
 will returns a pointer of the requested
 .IR type .
 .P
+.B pool_boundary()
+returns a pointer that should only be used in a call to
+.BR pool_free_old() .
+.P
 .BR pool_free() ,
+.BR pool_free_old() ,
 .B pool_tfree()
 and
 .B pool_destroy()

lib/pool_alloc.c

@@ -2,21 +2,20 @@
 
 #define POOL_DEF_EXTENT	(32 * 1024)
 
+#define POOL_QALIGN_P2		(1<<16)		/* power-of-2 qalign	*/
+
 struct alloc_pool
 {
 	size_t			size;		/* extent size		*/
 	size_t			quantum;	/* allocation quantum	*/
-	struct pool_extent	*live;		/* current extent for
-						 * allocations		*/
-	struct pool_extent	*free;		/* unfreed extent list	*/
-	void			(*bomb)();
-						/* function to call if
+	struct pool_extent	*extents;	/* top extent is "live" */
+	void			(*bomb)();	/* function to call if
 						 * malloc fails		*/
 	int			flags;
 
 	/* statistical data */
 	unsigned long		e_created;	/* extents created	*/
-	unsigned long		e_freed;	/* extents detroyed	*/
+	unsigned long		e_freed;	/* extents destroyed	*/
 	int64			n_allocated;	/* calls to alloc	*/
 	int64			n_freed;	/* calls to free	*/
 	int64			b_allocated;	/* cum. bytes allocated	*/
@@ -25,16 +24,18 @@ struct alloc_pool
 
 struct pool_extent
 {
+	struct pool_extent	*next;
 	void			*start;		/* starting address	*/
 	size_t			free;		/* free bytecount	*/
-	size_t			bound;		/* bytes bound by padding,
-						 * overhead and freed	*/
-	struct pool_extent	*next;
+	size_t			bound;		/* trapped free bytes	*/
 };
 
 struct align_test {
-    void *foo;
-    int64 bar;
+	uchar foo;
+	union {
+	    int64 i;
+	    void *p;
+	} bar;
 };
 
 #define MINALIGN	offsetof(struct align_test, bar)
@@ -44,24 +45,44 @@ struct align_test {
 #define PTR_ADD(b,o)	( (void*) ((char*)(b) + (o)) )
 
 alloc_pool_t
-pool_create(size_t size, size_t quantum,
-    void (*bomb)(char *), int flags)
+pool_create(size_t size, size_t quantum, void (*bomb)(const char *), int flags)
 {
-	struct alloc_pool	*pool;
+	struct alloc_pool *pool;
 
-	if (!(pool = (struct alloc_pool*) malloc(sizeof (struct alloc_pool))))
-		return pool;
-	memset(pool, 0, sizeof (struct alloc_pool));
-
-	pool->size = size	/* round extent size to min alignment reqs */
-	    ? (size + MINALIGN - 1) & ~(MINALIGN - 1)
-	    : POOL_DEF_EXTENT;
-	if (pool->flags & POOL_INTERN)
-	{
-		pool->size -= sizeof (struct pool_extent);
-		flags |= POOL_APPEND;
+	if ((MINALIGN & (MINALIGN - 1)) != 0) {
+		if (bomb)
+			(*bomb)("Compiler error: MINALIGN is not a power of 2\n");
+		return NULL;
 	}
-	pool->quantum = quantum ? quantum : MINALIGN;
+
+	if (!(pool = new0(struct alloc_pool)))
+		return NULL;
+
+	if (!size)
+		size = POOL_DEF_EXTENT;
+	if (!quantum)
+		quantum = MINALIGN;
+
+	if (flags & POOL_INTERN) {
+		if (size <= sizeof (struct pool_extent))
+			size = quantum;
+		else
+			size -= sizeof (struct pool_extent);
+		flags |= POOL_PREPEND;
+	}
+
+	if (quantum <= 1)
+		flags = (flags | POOL_NO_QALIGN) & ~POOL_QALIGN_P2;
+	else if (!(flags & POOL_NO_QALIGN)) {
+		if (size % quantum)
+			size += quantum - size % quantum;
+		/* If quantum is a power of 2, we'll avoid using modulus. */
+		if (!(quantum & (quantum - 1)))
+			flags |= POOL_QALIGN_P2;
+	}
+
+	pool->size = size;
+	pool->quantum = quantum;
 	pool->bomb = bomb;
 	pool->flags = flags;
 
@@ -72,30 +93,26 @@ void
 pool_destroy(alloc_pool_t p)
 {
 	struct alloc_pool *pool = (struct alloc_pool *) p;
-	struct pool_extent	*cur, *next;
+	struct pool_extent *cur, *next;
 
 	if (!pool)
 		return;
 
-	if (pool->live)
-	{
-		cur = pool->live;
-		free(cur->start);
-		if (!(pool->flags & POOL_APPEND))
-			free(cur);
-	}
-	for (cur = pool->free; cur; cur = next)
-	{
+	for (cur = pool->extents; cur; cur = next) {
 		next = cur->next;
-		free(cur->start);
-		if (!(pool->flags & POOL_APPEND))
+		if (pool->flags & POOL_PREPEND)
+			free(PTR_ADD(cur->start, -sizeof (struct pool_extent)));
+		else {
+			free(cur->start);
 			free(cur);
+		}
 	}
+
 	free(pool);
 }
 
 void *
-pool_alloc(alloc_pool_t p, size_t len, char *bomb)
+pool_alloc(alloc_pool_t p, size_t len, const char *bomb_msg)
 {
 	struct alloc_pool *pool = (struct alloc_pool *) p;
 	if (!pool)
@@ -103,57 +120,42 @@ pool_alloc(alloc_pool_t p, size_t len, char *bomb)
 
 	if (!len)
 		len = pool->quantum;
-	else if (pool->quantum > 1 && len % pool->quantum)
-		len += pool->quantum - len % pool->quantum;
+	else if (pool->flags & POOL_QALIGN_P2) {
+		if (len & (pool->quantum - 1))
+			len += pool->quantum - (len & (pool->quantum - 1));
+	} else if (!(pool->flags & POOL_NO_QALIGN)) {
+		if (len % pool->quantum)
+			len += pool->quantum - len % pool->quantum;
+	}
 
 	if (len > pool->size)
-		goto bomb;
+		goto bomb_out;
 
-	if (!pool->live || len > pool->live->free)
-	{
-		void	*start;
-		size_t	free;
-		size_t	bound;
-		size_t	sqew;
-		size_t	asize;
-
-		if (pool->live)
-		{
-			pool->live->next = pool->free;
-			pool->free = pool->live;
-		}
-
-		free = pool->size;
-		bound = 0;
+	if (!pool->extents || len > pool->extents->free) {
+		void *start;
+		size_t asize;
+		struct pool_extent *ext;
 
 		asize = pool->size;
-		if (pool->flags & POOL_APPEND)
+		if (pool->flags & POOL_PREPEND)
 			asize += sizeof (struct pool_extent);
 
-		if (!(start = (void *) malloc(asize)))
-			goto bomb;
+		if (!(start = new_array(char, asize)))
+			goto bomb_out;
 
 		if (pool->flags & POOL_CLEAR)
-			memset(start, 0, pool->size);
+			memset(start, 0, asize);
 
-		if (pool->flags & POOL_APPEND)
-		{
-			pool->live = PTR_ADD(start, free);
-		}
-		else if (!(pool->live = (struct pool_extent *) malloc(sizeof (struct pool_extent))))
-		{
-			goto bomb;
-		}
-		if (pool->flags & POOL_QALIGN && pool->quantum > 1
-		    && (sqew = (size_t)PTR_ADD(start, free) % pool->quantum))
-		{
-			bound  += sqew;
-			free -= sqew;
-		}
-		pool->live->start = start;
-		pool->live->free = free;
-		pool->live->bound = bound;
-		pool->live->next = NULL;
+		if (pool->flags & POOL_PREPEND) {
+			ext = start;
+			start = PTR_ADD(start, sizeof (struct pool_extent));
+		} else if (!(ext = new(struct pool_extent)))
+			goto bomb_out;
+		ext->start = start;
+		ext->free = pool->size;
+		ext->bound = 0;
+		ext->next = pool->extents;
+		pool->extents = ext;
 
 		pool->e_created++;
 	}
@@ -161,71 +163,51 @@ pool_alloc(alloc_pool_t p, size_t len, char *bomb)
 	pool->n_allocated++;
 	pool->b_allocated += len;
 
-	pool->live->free -= len;
+	pool->extents->free -= len;
 
-	return PTR_ADD(pool->live->start, pool->live->free);
+	return PTR_ADD(pool->extents->start, pool->extents->free);
 
-bomb:
+  bomb_out:
 	if (pool->bomb)
-		(*pool->bomb)(bomb);
+		(*pool->bomb)(bomb_msg);
 	return NULL;
 }
 
+/* This function allows you to declare memory in the pool that you are done
+ * using.  If you free all the memory in a pool's extent, that extent will
+ * be freed. */
 void
 pool_free(alloc_pool_t p, size_t len, void *addr)
 {
-	struct alloc_pool *pool = (struct alloc_pool *) p;
-	struct pool_extent	*cur;
-	struct pool_extent	*prev;
+	struct alloc_pool *pool = (struct alloc_pool *)p;
+	struct pool_extent *cur, *prev;
 
 	if (!pool)
 		return;
 
-	if (!len)
-		len = pool->quantum;
-	else if (pool->quantum > 1 && len % pool->quantum)
-		len += pool->quantum - len % pool->quantum;
-
-	if (!addr && pool->live)
-	{
-		pool->live->next = pool->free;
-		pool->free = pool->live;
-		pool->live = NULL;
+	if (!addr) {
+		/* A NULL addr starts a fresh extent for new allocations. */
+		if ((cur = pool->extents) != NULL && cur->free != pool->size) {
+			cur->bound += cur->free;
+			cur->free = 0;
+		}
 		return;
 	}
+
+	if (!len)
+		len = pool->quantum;
+	else if (pool->flags & POOL_QALIGN_P2) {
+		if (len & (pool->quantum - 1))
+			len += pool->quantum - (len & (pool->quantum - 1));
+	} else if (!(pool->flags & POOL_NO_QALIGN)) {
+		if (len % pool->quantum)
+			len += pool->quantum - len % pool->quantum;
+	}
+
 	pool->n_freed++;
 	pool->b_freed += len;
 
-	cur = pool->live;
-	if (cur
-	    && addr >= cur->start
-	    && addr < PTR_ADD(cur->start, pool->size))
-	{
-		if (addr == PTR_ADD(cur->start, cur->free))
-		{
-			if (pool->flags & POOL_CLEAR)
-				memset(addr, 0, len);
-			pool->b_freed += len;
-		} else {
-			cur->bound += len;
-		}
-		if (cur->free + cur->bound >= pool->size)
-		{
-			size_t sqew;
-
-			cur->free = pool->size;
-			cur->bound = 0;
-			if (pool->flags & POOL_QALIGN && pool->quantum > 1
-			    && (sqew = (size_t)PTR_ADD(cur->start, cur->free) % pool->quantum))
-			{
-				cur->bound += sqew;
-				cur->free -= sqew;
-			}
-		}
-		return;
-	}
-	for (prev = NULL, cur = pool->free; cur; prev = cur, cur = cur->next)
-	{
+	for (prev = NULL, cur = pool->extents; cur; prev = cur, cur = cur->next) {
 		if (addr >= cur->start
 		    && addr < PTR_ADD(cur->start, pool->size))
 			break;
@@ -233,45 +215,141 @@ pool_free(alloc_pool_t p, size_t len, void *addr)
 	if (!cur)
 		return;
 
-	if (prev)
-	{
-		prev->next = cur->next;
-		cur->next = pool->free;
-		pool->free = cur;
+	if (!prev) {
+		/* The "live" extent is kept ready for more allocations. */
+		if (cur->free + cur->bound + len >= pool->size) {
+			if (pool->flags & POOL_CLEAR) {
+				memset(PTR_ADD(cur->start, cur->free), 0,
+				       pool->size - cur->free);
+			}
+			cur->free = pool->size;
+			cur->bound = 0;
+		} else if (addr == PTR_ADD(cur->start, cur->free)) {
+			if (pool->flags & POOL_CLEAR)
+				memset(addr, 0, len);
+			cur->free += len;
+		} else
+			cur->bound += len;
+	} else {
+		cur->bound += len;
+
+		if (cur->free + cur->bound >= pool->size) {
+			prev->next = cur->next;
+			if (pool->flags & POOL_PREPEND)
+				free(PTR_ADD(cur->start, -sizeof (struct pool_extent)));
+			else {
+				free(cur->start);
+				free(cur);
+			}
+			pool->e_freed++;
+		} else if (prev != pool->extents) {
+			/* Move the extent to be the first non-live extent. */
+			prev->next = cur->next;
+			cur->next = pool->extents->next;
+			pool->extents->next = cur;
+		}
 	}
-	cur->bound += len;
+}
 
-	if (cur->free + cur->bound >= pool->size)
-	{
-		pool->free = cur->next;
+/* This allows you to declare that the given address marks the edge of some
+ * pool memory that is no longer needed.  Any extents that hold only data
+ * older than the boundary address are freed.  NOTE: You MUST NOT USE BOTH
+ * pool_free() and pool_free_old() on the same pool!! */
+void
+pool_free_old(alloc_pool_t p, void *addr)
+{
+	struct alloc_pool *pool = (struct alloc_pool *)p;
+	struct pool_extent *cur, *prev, *next;
 
-		free(cur->start);
-		if (!(pool->flags & POOL_APPEND))
+	if (!pool || !addr)
+		return;
+
+	for (prev = NULL, cur = pool->extents; cur; prev = cur, cur = cur->next) {
+		if (addr >= cur->start
+		    && addr < PTR_ADD(cur->start, pool->size))
+			break;
+	}
+	if (!cur)
+		return;
+
+	if (addr == PTR_ADD(cur->start, cur->free)) {
+		if (prev) {
+			prev->next = NULL;
+			next = cur;
+		} else {
+			/* The most recent live extent can just be reset. */
+			if (pool->flags & POOL_CLEAR)
+				memset(addr, 0, pool->size - cur->free);
+			cur->free = pool->size;
+			cur->bound = 0;
+			next = cur->next;
+			cur->next = NULL;
+		}
+	} else {
+		next = cur->next;
+		cur->next = NULL;
+	}
+
+	while ((cur = next) != NULL) {
+		next = cur->next;
+		if (pool->flags & POOL_PREPEND)
+			free(PTR_ADD(cur->start, -sizeof (struct pool_extent)));
+		else {
+			free(cur->start);
 			free(cur);
+		}
 		pool->e_freed++;
 	}
-	return;
+}
+
+/* If the current extent doesn't have "len" free space in it, mark it as full
+ * so that the next alloc will start a new extent.  If len is (size_t)-1, this
+ * bump will always occur.  The function returns a boundary address that can
+ * be used with pool_free_old(), or a NULL if no memory is allocated. */
+void *
+pool_boundary(alloc_pool_t p, size_t len)
+{
+	struct alloc_pool *pool = (struct alloc_pool *)p;
+	struct pool_extent *cur;
+
+	if (!pool || !pool->extents)
+		return NULL;
+
+	cur = pool->extents;
+
+	if (cur->free < len) {
+		cur->bound += cur->free;
+		cur->free = 0;
+	}
+
+	return PTR_ADD(cur->start, cur->free);
 }
 
 #define FDPRINT(label, value) \
-	snprintf(buf, BUFSIZ, label, value), \
-	write(fd, buf, strlen(buf));
+	do { \
+		int len = snprintf(buf, sizeof buf, label, value); \
+		if (write(fd, buf, len) != len) \
+			ret = -1; \
+	} while (0)
 
 #define FDEXTSTAT(ext) \
-	snprintf(buf, BUFSIZ, "  %12ld  %5ld\n", \
-		(long) ext->free, \
-		(long) ext->bound), \
-	write(fd, buf, strlen(buf))
+	do { \
+		int len = snprintf(buf, sizeof buf, "  %12ld  %5ld\n", \
+				   (long)ext->free, (long)ext->bound); \
+		if (write(fd, buf, len) != len) \
+			ret = -1; \
+	} while (0)
 
-void
+int
 pool_stats(alloc_pool_t p, int fd, int summarize)
 {
 	struct alloc_pool *pool = (struct alloc_pool *) p;
-	struct pool_extent	*cur;
+	struct pool_extent *cur;
 	char buf[BUFSIZ];
+	int ret = 0;
 
 	if (!pool)
-		return;
+		return ret;
 
 	FDPRINT("  Extent size:       %12ld\n",	(long)	pool->size);
 	FDPRINT("  Alloc quantum:     %12ld\n",	(long)	pool->quantum);
@@ -279,26 +357,20 @@ pool_stats(alloc_pool_t p, int fd, int summarize)
 	FDPRINT("  Extents freed:     %12ld\n",		pool->e_freed);
 	FDPRINT("  Alloc count:       %12.0f\n", (double) pool->n_allocated);
 	FDPRINT("  Free Count:        %12.0f\n", (double) pool->n_freed);
-	FDPRINT("  Alloc bytes:       %12.0f\n", (double) pool->b_allocated);
-	FDPRINT("  Free bytes:        %12.0f\n", (double) pool->b_freed);
+	FDPRINT("  Bytes allocated:   %12.0f\n", (double) pool->b_allocated);
+	FDPRINT("  Bytes freed:       %12.0f\n", (double) pool->b_freed);
 
 	if (summarize)
-		return;
+		return ret;
 
-	if (!pool->live && !pool->free)
-		return;
+	if (!pool->extents)
+		return ret;
 
-	write(fd, "\n", 1);
+	if (write(fd, "\n", 1) != 1)
+		ret = -1;
 
-	if (pool->live)
-	{
-		FDEXTSTAT(pool->live);
-	}
-	strcpy(buf, "   FREE    BOUND\n");
-	write(fd, buf, strlen(buf));
-
-	for (cur = pool->free; cur; cur = cur->next)
-	{
+	for (cur = pool->extents; cur; cur = cur->next)
 		FDEXTSTAT(cur);
-	}
+
+	return ret;
 }

lib/pool_alloc.h

@@ -1,20 +1,21 @@
 #include <stddef.h>
 
 #define POOL_CLEAR	(1<<0)		/* zero fill allocations	*/
-#define POOL_QALIGN	(1<<1)		/* align data to quanta		*/
+#define POOL_NO_QALIGN	(1<<1)		/* don't align data to quanta	*/
 #define POOL_INTERN	(1<<2)		/* Allocate extent structures	*/
-#define POOL_APPEND	(1<<3)		/*   or appended to extent data	*/
+#define POOL_PREPEND	(1<<3)		/*   or prepend to extent data	*/
 
 typedef void *alloc_pool_t;
 
-alloc_pool_t pool_create(size_t size, size_t quantum, void (*bomb)(char *), int flags);
+alloc_pool_t pool_create(size_t size, size_t quantum, void (*bomb)(const char *), int flags);
 void pool_destroy(alloc_pool_t pool);
-void *pool_alloc(alloc_pool_t pool, size_t size, char *bomb);
+void *pool_alloc(alloc_pool_t pool, size_t size, const char *bomb_msg);
 void pool_free(alloc_pool_t pool, size_t size, void *addr);
+void pool_free_old(alloc_pool_t pool, void *addr);
+void *pool_boundary(alloc_pool_t pool, size_t size);
 
-#define pool_talloc(pool, type, count, bomb) \
-	((type *)pool_alloc(pool, sizeof(type) * count, bomb))
+#define pool_talloc(pool, type, count, bomb_msg) \
+	((type *)pool_alloc(pool, sizeof(type) * count, bomb_msg))
 
 #define pool_tfree(pool, type, count, addr) \
 	(pool_free(pool, sizeof(type) * count, addr))
-

lib/sysacls.h

@@ -0,0 +1,305 @@
+/*
+ * Unix SMB/Netbios implementation.
+ * Version 2.2.x
+ * Portable SMB ACL interface
+ * Copyright (C) Jeremy Allison 2000
+ * Copyright (C) 2007-2019 Wayne Davison
+ *
+ * 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 3 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
+ * with this program; if not, visit the http://fsf.org website.
+ */
+
+#ifdef SUPPORT_ACLS
+
+#ifdef HAVE_SYS_ACL_H
+#include <sys/acl.h>
+#endif
+#ifdef HAVE_ACL_LIBACL_H
+#include <acl/libacl.h>
+#endif
+
+#define SMB_MALLOC(cnt) new_array(char, cnt)
+#define SMB_MALLOC_P(obj) new_array(obj, 1)
+#define SMB_MALLOC_ARRAY(obj, cnt) new_array(obj, cnt)
+#define SMB_REALLOC(mem, cnt) realloc_array(mem, char, cnt)
+#define slprintf snprintf
+
+#if defined HAVE_POSIX_ACLS /*-----------------------------------------------*/
+
+/* This is an identity mapping (just remove the SMB_). */
+
+#define SMB_ACL_TAG_T		acl_tag_t
+#define SMB_ACL_TYPE_T		acl_type_t
+
+/* Types of ACLs. */
+#define SMB_ACL_USER		ACL_USER
+#define SMB_ACL_USER_OBJ	ACL_USER_OBJ
+#define SMB_ACL_GROUP		ACL_GROUP
+#define SMB_ACL_GROUP_OBJ	ACL_GROUP_OBJ
+#define SMB_ACL_OTHER		ACL_OTHER
+#define SMB_ACL_MASK		ACL_MASK
+
+#define SMB_ACL_T		acl_t
+
+#define SMB_ACL_ENTRY_T		acl_entry_t
+
+#define SMB_ACL_FIRST_ENTRY	ACL_FIRST_ENTRY
+#define SMB_ACL_NEXT_ENTRY	ACL_NEXT_ENTRY
+
+#define SMB_ACL_TYPE_ACCESS	ACL_TYPE_ACCESS
+#define SMB_ACL_TYPE_DEFAULT	ACL_TYPE_DEFAULT
+
+#define SMB_ACL_VALID_NAME_BITS	(4 | 2 | 1)
+#define SMB_ACL_VALID_OBJ_BITS	(4 | 2 | 1)
+
+#define SMB_ACL_NEED_SORT
+
+#elif defined HAVE_TRU64_ACLS /*---------------------------------------------*/
+
+/* This is for DEC/Compaq Tru64 UNIX */
+
+#define SMB_ACL_TAG_T		acl_tag_t
+#define SMB_ACL_TYPE_T		acl_type_t
+
+/* Types of ACLs. */
+#define SMB_ACL_USER		ACL_USER
+#define SMB_ACL_USER_OBJ	ACL_USER_OBJ
+#define SMB_ACL_GROUP		ACL_GROUP
+#define SMB_ACL_GROUP_OBJ	ACL_GROUP_OBJ
+#define SMB_ACL_OTHER		ACL_OTHER
+#define SMB_ACL_MASK		ACL_MASK
+
+#define SMB_ACL_T		acl_t
+
+#define SMB_ACL_ENTRY_T		acl_entry_t
+
+#define SMB_ACL_FIRST_ENTRY	0
+#define SMB_ACL_NEXT_ENTRY	1
+
+#define SMB_ACL_TYPE_ACCESS	ACL_TYPE_ACCESS
+#define SMB_ACL_TYPE_DEFAULT	ACL_TYPE_DEFAULT
+
+#define SMB_ACL_VALID_NAME_BITS	(4 | 2 | 1)
+#define SMB_ACL_VALID_OBJ_BITS	(4 | 2 | 1)
+
+#define SMB_ACL_NEED_SORT
+
+#elif defined HAVE_UNIXWARE_ACLS || defined HAVE_SOLARIS_ACLS /*-------------*/
+
+/* Donated by Michael Davidson <md@sco.COM> for UnixWare / OpenUNIX.
+ * Modified by Toomas Soome <tsoome@ut.ee> for Solaris.  */
+
+/* SVR4.2 ES/MP ACLs */
+typedef int SMB_ACL_TAG_T;
+typedef int SMB_ACL_TYPE_T;
+
+/* Types of ACLs. */
+#define SMB_ACL_USER		USER
+#define SMB_ACL_USER_OBJ	USER_OBJ
+#define SMB_ACL_GROUP		GROUP
+#define SMB_ACL_GROUP_OBJ	GROUP_OBJ
+#define SMB_ACL_OTHER		OTHER_OBJ
+#define SMB_ACL_MASK		CLASS_OBJ
+
+typedef struct SMB_ACL_T {
+	int size;
+	int count;
+	int next;
+	struct acl acl[1];
+} *SMB_ACL_T;
+
+typedef struct acl *SMB_ACL_ENTRY_T;
+
+#define SMB_ACL_FIRST_ENTRY	0
+#define SMB_ACL_NEXT_ENTRY	1
+
+#define SMB_ACL_TYPE_ACCESS	0
+#define SMB_ACL_TYPE_DEFAULT	1
+
+#define SMB_ACL_VALID_NAME_BITS	(4 | 2 | 1)
+#define SMB_ACL_VALID_OBJ_BITS	(4 | 2 | 1)
+
+#define SMB_ACL_NEED_SORT
+
+#ifdef __CYGWIN__
+#define SMB_ACL_LOSES_SPECIAL_MODE_BITS
+#endif
+
+#elif defined HAVE_HPUX_ACLS /*----------------------------------------------*/
+
+/* Based on the Solaris & UnixWare code. */
+
+#undef GROUP
+#include <sys/aclv.h>
+
+/* SVR4.2 ES/MP ACLs */
+typedef int SMB_ACL_TAG_T;
+typedef int SMB_ACL_TYPE_T;
+
+/* Types of ACLs. */
+#define SMB_ACL_USER		USER
+#define SMB_ACL_USER_OBJ	USER_OBJ
+#define SMB_ACL_GROUP		GROUP
+#define SMB_ACL_GROUP_OBJ	GROUP_OBJ
+#define SMB_ACL_OTHER		OTHER_OBJ
+#define SMB_ACL_MASK		CLASS_OBJ
+
+typedef struct SMB_ACL_T {
+	int size;
+	int count;
+	int next;
+	struct acl acl[1];
+} *SMB_ACL_T;
+
+typedef struct acl *SMB_ACL_ENTRY_T;
+
+#define SMB_ACL_FIRST_ENTRY	0
+#define SMB_ACL_NEXT_ENTRY	1
+
+#define SMB_ACL_TYPE_ACCESS	0
+#define SMB_ACL_TYPE_DEFAULT	1
+
+#define SMB_ACL_VALID_NAME_BITS	(4 | 2 | 1)
+#define SMB_ACL_VALID_OBJ_BITS	(4 | 2 | 1)
+
+#define SMB_ACL_NEED_SORT
+
+#elif defined HAVE_IRIX_ACLS /*----------------------------------------------*/
+
+/* IRIX ACLs */
+
+#define SMB_ACL_TAG_T		acl_tag_t
+#define SMB_ACL_TYPE_T		acl_type_t
+
+/* Types of ACLs. */
+#define SMB_ACL_USER		ACL_USER
+#define SMB_ACL_USER_OBJ	ACL_USER_OBJ
+#define SMB_ACL_GROUP		ACL_GROUP
+#define SMB_ACL_GROUP_OBJ	ACL_GROUP_OBJ
+#define SMB_ACL_OTHER		ACL_OTHER_OBJ
+#define SMB_ACL_MASK		ACL_MASK
+
+typedef struct SMB_ACL_T {
+	int next;
+	BOOL freeaclp;
+	struct acl *aclp;
+} *SMB_ACL_T;
+
+#define SMB_ACL_ENTRY_T		acl_entry_t
+
+#define SMB_ACL_FIRST_ENTRY	0
+#define SMB_ACL_NEXT_ENTRY	1
+
+#define SMB_ACL_TYPE_ACCESS	ACL_TYPE_ACCESS
+#define SMB_ACL_TYPE_DEFAULT	ACL_TYPE_DEFAULT
+
+#define SMB_ACL_VALID_NAME_BITS	(4 | 2 | 1)
+#define SMB_ACL_VALID_OBJ_BITS	(4 | 2 | 1)
+
+#define SMB_ACL_NEED_SORT
+
+#elif defined HAVE_AIX_ACLS /*-----------------------------------------------*/
+
+/* Donated by Medha Date, mdate@austin.ibm.com, for IBM */
+
+#include "/usr/include/acl.h"
+
+struct acl_entry_link{
+	struct acl_entry_link *prevp;
+	struct new_acl_entry *entryp;
+	struct acl_entry_link *nextp;
+	int count;
+};
+
+struct new_acl_entry{
+	unsigned short ace_len;
+	unsigned short ace_type;
+	unsigned int ace_access;
+	struct ace_id ace_id[1];
+};
+
+#define SMB_ACL_ENTRY_T		struct new_acl_entry*
+#define SMB_ACL_T		struct acl_entry_link*
+ 
+#define SMB_ACL_TAG_T		unsigned short
+#define SMB_ACL_TYPE_T		int
+
+/* Types of ACLs. */
+#define SMB_ACL_USER		ACEID_USER
+#define SMB_ACL_USER_OBJ	3
+#define SMB_ACL_GROUP		ACEID_GROUP
+#define SMB_ACL_GROUP_OBJ	4
+#define SMB_ACL_OTHER		5
+#define SMB_ACL_MASK		6
+
+#define SMB_ACL_FIRST_ENTRY	1
+#define SMB_ACL_NEXT_ENTRY	2
+
+#define SMB_ACL_TYPE_ACCESS	0
+#define SMB_ACL_TYPE_DEFAULT	1
+
+#define SMB_ACL_VALID_NAME_BITS	(4 | 2 | 1)
+#define SMB_ACL_VALID_OBJ_BITS	(4 | 2 | 1)
+
+#define SMB_ACL_NEED_SORT
+
+#elif defined(HAVE_OSX_ACLS) /*----------------------------------------------*/
+
+/* Special handling for OS X ACLs */
+
+#define SMB_ACL_TAG_T		acl_tag_t
+#define SMB_ACL_TYPE_T		acl_type_t
+
+#define SMB_ACL_T		acl_t
+
+#define SMB_ACL_ENTRY_T		acl_entry_t
+
+#define SMB_ACL_USER		1
+#define SMB_ACL_GROUP		2
+
+#define SMB_ACL_FIRST_ENTRY	ACL_FIRST_ENTRY
+#define SMB_ACL_NEXT_ENTRY	ACL_NEXT_ENTRY
+
+#define SMB_ACL_TYPE_ACCESS	ACL_TYPE_EXTENDED
+#define SMB_ACL_TYPE_DEFAULT	ACL_TYPE_DEFAULT
+
+#define SMB_ACL_VALID_NAME_BITS	((1<<25)-1)
+#define SMB_ACL_VALID_OBJ_BITS	0
+
+/*#undef SMB_ACL_NEED_SORT*/
+
+#else /*---------------------------------------------------------------------*/
+
+/* Unknown platform. */
+
+#error Cannot handle ACLs on this platform!
+
+#endif
+
+int sys_acl_get_entry(SMB_ACL_T the_acl, int entry_id, SMB_ACL_ENTRY_T *entry_p);
+int sys_acl_get_tag_type(SMB_ACL_ENTRY_T entry_d, SMB_ACL_TAG_T *tag_type_p);
+int sys_acl_get_info(SMB_ACL_ENTRY_T entry, SMB_ACL_TAG_T *tag_type_p, uint32 *bits_p, id_t *u_g_id_p);
+SMB_ACL_T sys_acl_get_file(const char *path_p, SMB_ACL_TYPE_T type);
+SMB_ACL_T sys_acl_get_fd(int fd);
+SMB_ACL_T sys_acl_init(int count);
+int sys_acl_create_entry(SMB_ACL_T *pacl, SMB_ACL_ENTRY_T *pentry);
+int sys_acl_set_info(SMB_ACL_ENTRY_T entry, SMB_ACL_TAG_T tagtype, uint32 bits, id_t u_g_id);
+int sys_acl_set_access_bits(SMB_ACL_ENTRY_T entry, uint32 bits);
+int sys_acl_valid(SMB_ACL_T theacl);
+int sys_acl_set_file(const char *name, SMB_ACL_TYPE_T acltype, SMB_ACL_T theacl);
+int sys_acl_set_fd(int fd, SMB_ACL_T theacl);
+int sys_acl_delete_def_file(const char *name);
+int sys_acl_free_acl(SMB_ACL_T the_acl);
+int no_acl_syscall_error(int err);
+
+#endif /* SUPPORT_ACLS */

lib/sysxattrs.c

@@ -0,0 +1,300 @@
+/*
+ * Extended attribute support for rsync.
+ *
+ * Copyright (C) 2004 Red Hat, Inc.
+ * Copyright (C) 2003-2019 Wayne Davison
+ * Written by Jay Fenlason.
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
+
+#include "rsync.h"
+#include "sysxattrs.h"
+
+#ifdef SUPPORT_XATTRS
+
+#ifdef HAVE_OSX_XATTRS
+#define GETXATTR_FETCH_LIMIT (64*1024*1024)
+#endif
+
+#if defined HAVE_LINUX_XATTRS
+
+ssize_t sys_lgetxattr(const char *path, const char *name, void *value, size_t size)
+{
+	return lgetxattr(path, name, value, size);
+}
+
+ssize_t sys_fgetxattr(int filedes, const char *name, void *value, size_t size)
+{
+	return fgetxattr(filedes, name, value, size);
+}
+
+int sys_lsetxattr(const char *path, const char *name, const void *value, size_t size)
+{
+	return lsetxattr(path, name, value, size, 0);
+}
+
+int sys_lremovexattr(const char *path, const char *name)
+{
+	return lremovexattr(path, name);
+}
+
+ssize_t sys_llistxattr(const char *path, char *list, size_t size)
+{
+	return llistxattr(path, list, size);
+}
+
+#elif HAVE_OSX_XATTRS
+
+ssize_t sys_lgetxattr(const char *path, const char *name, void *value, size_t size)
+{
+	ssize_t len = getxattr(path, name, value, size, 0, XATTR_NOFOLLOW);
+
+	/* If we're retrieving data, handle resource forks > 64MB specially */
+	if (value != NULL && len == GETXATTR_FETCH_LIMIT && (size_t)len < size) {
+		/* getxattr will only return 64MB of data at a time, need to call again with a new offset */
+		u_int32_t offset = len;
+		size_t data_retrieved = len;
+		while (data_retrieved < size) {
+			len = getxattr(path, name, value + offset, size - data_retrieved, offset, XATTR_NOFOLLOW);
+			if (len <= 0)
+				break;
+			data_retrieved += len;
+			offset += (u_int32_t)len;
+		}
+		len = data_retrieved;
+	}
+
+	return len;
+}
+
+ssize_t sys_fgetxattr(int filedes, const char *name, void *value, size_t size)
+{
+	return fgetxattr(filedes, name, value, size, 0, 0);
+}
+
+int sys_lsetxattr(const char *path, const char *name, const void *value, size_t size)
+{
+	return setxattr(path, name, value, size, 0, XATTR_NOFOLLOW);
+}
+
+int sys_lremovexattr(const char *path, const char *name)
+{
+	return removexattr(path, name, XATTR_NOFOLLOW);
+}
+
+ssize_t sys_llistxattr(const char *path, char *list, size_t size)
+{
+	return listxattr(path, list, size, XATTR_NOFOLLOW);
+}
+
+#elif HAVE_FREEBSD_XATTRS
+
+ssize_t sys_lgetxattr(const char *path, const char *name, void *value, size_t size)
+{
+	return extattr_get_link(path, EXTATTR_NAMESPACE_USER, name, value, size);
+}
+
+ssize_t sys_fgetxattr(int filedes, const char *name, void *value, size_t size)
+{
+	return extattr_get_fd(filedes, EXTATTR_NAMESPACE_USER, name, value, size);
+}
+
+int sys_lsetxattr(const char *path, const char *name, const void *value, size_t size)
+{
+	return extattr_set_link(path, EXTATTR_NAMESPACE_USER, name, value, size);
+}
+
+int sys_lremovexattr(const char *path, const char *name)
+{
+	return extattr_delete_link(path, EXTATTR_NAMESPACE_USER, name);
+}
+
+ssize_t sys_llistxattr(const char *path, char *list, size_t size)
+{
+	unsigned char keylen;
+	ssize_t off, len = extattr_list_link(path, EXTATTR_NAMESPACE_USER, list, size);
+
+	if (len <= 0 || (size_t)len > size)
+		return len;
+
+	/* FreeBSD puts a single-byte length before each string, with no '\0'
+	 * terminator.  We need to change this into a series of null-terminted
+	 * strings.  Since the size is the same, we can simply transform the
+	 * output in place. */
+	for (off = 0; off < len; off += keylen + 1) {
+		keylen = ((unsigned char*)list)[off];
+		if (off + keylen >= len) {
+			/* Should be impossible, but kernel bugs happen! */
+			errno = EINVAL;
+			return -1;
+		}
+		memmove(list+off, list+off+1, keylen);
+		list[off+keylen] = '\0';
+	}
+
+	return len;
+}
+
+#elif HAVE_SOLARIS_XATTRS
+
+static ssize_t read_xattr(int attrfd, void *buf, size_t buflen)
+{
+	STRUCT_STAT sb;
+	ssize_t ret;
+
+	if (fstat(attrfd, &sb) < 0)
+		ret = -1;
+	else if (sb.st_size > SSIZE_MAX) {
+		errno = ERANGE;
+		ret = -1;
+	} else if (buflen == 0)
+		ret = sb.st_size;
+	else if (sb.st_size > buflen) {
+		errno = ERANGE;
+		ret = -1;
+	} else {
+		size_t bufpos;
+		for (bufpos = 0; bufpos < sb.st_size; ) {
+			ssize_t cnt = read(attrfd, buf + bufpos, sb.st_size - bufpos);
+			if (cnt <= 0) {
+				if (cnt < 0 && errno == EINTR)
+					continue;
+				bufpos = -1;
+				break;
+			}
+			bufpos += cnt;
+		}
+		ret = bufpos;
+	}
+
+	close(attrfd);
+
+	return ret;
+}
+
+ssize_t sys_lgetxattr(const char *path, const char *name, void *value, size_t size)
+{
+	int attrfd;
+
+	if ((attrfd = attropen(path, name, O_RDONLY)) < 0) {
+		errno = ENOATTR;
+		return -1;
+	}
+
+	return read_xattr(attrfd, value, size);
+}
+
+ssize_t sys_fgetxattr(int filedes, const char *name, void *value, size_t size)
+{
+	int attrfd;
+
+	if ((attrfd = openat(filedes, name, O_RDONLY|O_XATTR, 0)) < 0) {
+		errno = ENOATTR;
+		return -1;
+	}
+
+	return read_xattr(attrfd, value, size);
+}
+
+int sys_lsetxattr(const char *path, const char *name, const void *value, size_t size)
+{
+	int attrfd;
+	size_t bufpos;
+	mode_t mode = S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP;
+
+	if ((attrfd = attropen(path, name, O_CREAT|O_TRUNC|O_WRONLY, mode)) < 0)
+		return -1;
+
+	for (bufpos = 0; bufpos < size; ) {
+		ssize_t cnt = write(attrfd, value+bufpos, size);
+		if (cnt <= 0) {
+			if (cnt < 0 && errno == EINTR)
+				continue;
+			bufpos = -1;
+			break;
+		}
+		bufpos += cnt;
+	}
+
+	close(attrfd);
+
+	return bufpos > 0 ? 0 : -1;
+}
+
+int sys_lremovexattr(const char *path, const char *name)
+{
+	int attrdirfd;
+	int ret;
+
+	if ((attrdirfd = attropen(path, ".", O_RDONLY)) < 0)
+		return -1;
+
+	ret = unlinkat(attrdirfd, name, 0);
+
+	close(attrdirfd);
+
+	return ret;
+}
+
+ssize_t sys_llistxattr(const char *path, char *list, size_t size)
+{
+	int attrdirfd;
+	DIR *dirp;
+	struct dirent *dp;
+	ssize_t ret = 0;
+
+	if ((attrdirfd = attropen(path, ".", O_RDONLY)) < 0) {
+		errno = ENOTSUP;
+		return -1;
+	}
+
+	if ((dirp = fdopendir(attrdirfd)) == NULL) {
+		close(attrdirfd);
+		return -1;
+	}
+
+	while ((dp = readdir(dirp))) {
+		int len = strlen(dp->d_name);
+
+		if (dp->d_name[0] == '.' && (len == 1 || (len == 2 && dp->d_name[1] == '.')))
+			continue;
+		if (len == 11 && dp->d_name[0] == 'S' && strncmp(dp->d_name, "SUNWattr_r", 10) == 0
+		 && (dp->d_name[10] == 'o' || dp->d_name[10] == 'w'))
+			continue;
+
+		if ((ret += len+1) > size) {
+			if (size == 0)
+				continue;
+			ret = -1;
+			errno = ERANGE;
+			break;
+		}
+		memcpy(list, dp->d_name, len+1);
+		list += len+1;
+	}
+
+	closedir(dirp);
+	close(attrdirfd);
+
+	return ret;
+}
+
+#else
+
+#error You need to create xattr compatibility functions.
+
+#endif
+
+#endif /* SUPPORT_XATTRS */

lib/sysxattrs.h

@@ -0,0 +1,26 @@
+#ifdef SUPPORT_XATTRS
+
+#if defined HAVE_SYS_XATTR_H
+#include <sys/xattr.h>
+#elif defined HAVE_ATTR_XATTR_H
+#include <attr/xattr.h>
+#elif defined HAVE_SYS_EXTATTR_H
+#include <sys/extattr.h>
+#endif
+
+/* Linux 2.4 does not define this as a distinct errno value: */
+#ifndef ENOATTR
+#define ENOATTR ENODATA
+#endif
+
+ssize_t sys_lgetxattr(const char *path, const char *name, void *value, size_t size);
+ssize_t sys_fgetxattr(int filedes, const char *name, void *value, size_t size);
+int sys_lsetxattr(const char *path, const char *name, const void *value, size_t size);
+int sys_lremovexattr(const char *path, const char *name);
+ssize_t sys_llistxattr(const char *path, char *list, size_t size);
+
+#else
+
+/* No xattrs available */
+
+#endif

lib/wildmatch.c

@@ -57,173 +57,312 @@
 int wildmatch_iteration_count;
 #endif
 
-static int domatch(const uchar *p, const uchar *text)
+static int force_lower_case = 0;
+
+/* Match pattern "p" against the a virtually-joined string consisting
+ * of "text" and any strings in array "a". */
+static int dowild(const uchar *p, const uchar *text, const uchar*const *a)
 {
-    int matched, special;
-    uchar ch, prev;
+    uchar p_ch;
 
 #ifdef WILD_TEST_ITERATIONS
     wildmatch_iteration_count++;
 #endif
 
-    for ( ; (ch = *p) != '\0'; text++, p++) {
-	if (*text == '\0' && ch != '*')
-	    return FALSE;
-	switch (ch) {
+    for ( ; (p_ch = *p) != '\0'; text++, p++) {
+	int matched, special;
+	uchar t_ch, prev_ch;
+	while ((t_ch = *text) == '\0') {
+	    if (*a == NULL) {
+		if (p_ch != '*')
+		    return ABORT_ALL;
+		break;
+	    }
+	    text = *a++;
+	}
+	if (force_lower_case && ISUPPER(t_ch))
+	    t_ch = tolower(t_ch);
+	switch (p_ch) {
 	  case '\\':
 	    /* Literal match with following character.  Note that the test
 	     * in "default" handles the p[1] == '\0' failure case. */
-	    ch = *++p;
+	    p_ch = *++p;
 	    /* FALLTHROUGH */
 	  default:
-	    if (*text != ch)
+	    if (t_ch != p_ch)
 		return FALSE;
 	    continue;
 	  case '?':
 	    /* Match anything but '/'. */
-	    if (*text == '/')
+	    if (t_ch == '/')
 		return FALSE;
 	    continue;
 	  case '*':
 	    if (*++p == '*') {
 		while (*++p == '*') {}
 		special = TRUE;
-	    }
-	    else
+	    } else
 		special = FALSE;
 	    if (*p == '\0') {
 		/* Trailing "**" matches everything.  Trailing "*" matches
 		 * only if there are no more slash characters. */
-		return special? TRUE : strchr((char*)text, '/') == NULL;
+		if (!special) {
+		    do {
+			if (strchr((char*)text, '/') != NULL)
+			    return FALSE;
+		    } while ((text = *a++) != NULL);
+		}
+		return TRUE;
 	    }
-	    for ( ; *text; text++) {
-		if ((matched = domatch(p, text)) != FALSE) {
+	    while (1) {
+		if (t_ch == '\0') {
+		    if ((text = *a++) == NULL)
+			break;
+		    t_ch = *text;
+		    continue;
+		}
+		if ((matched = dowild(p, text, a)) != FALSE) {
 		    if (!special || matched != ABORT_TO_STARSTAR)
 			return matched;
-		}
-		else if (!special && *text == '/')
+		} else if (!special && t_ch == '/')
 		    return ABORT_TO_STARSTAR;
+		t_ch = *++text;
 	    }
 	    return ABORT_ALL;
 	  case '[':
-	    ch = *++p;
+	    p_ch = *++p;
 #ifdef NEGATE_CLASS2
-	    if (ch == NEGATE_CLASS2)
-		ch = NEGATE_CLASS;
+	    if (p_ch == NEGATE_CLASS2)
+		p_ch = NEGATE_CLASS;
 #endif
 	    /* Assign literal TRUE/FALSE because of "matched" comparison. */
-	    special = ch == NEGATE_CLASS? TRUE : FALSE;
+	    special = p_ch == NEGATE_CLASS? TRUE : FALSE;
 	    if (special) {
 		/* Inverted character class. */
-		ch = *++p;
+		p_ch = *++p;
 	    }
-	    prev = 0;
+	    prev_ch = 0;
 	    matched = FALSE;
 	    do {
-		if (!ch)
+		if (!p_ch)
 		    return ABORT_ALL;
-		if (ch == '\\') {
-		    ch = *++p;
-		    if (!ch)
+		if (p_ch == '\\') {
+		    p_ch = *++p;
+		    if (!p_ch)
 			return ABORT_ALL;
-		    if (*text == ch)
+		    if (t_ch == p_ch)
 			matched = TRUE;
-		}
-		else if (ch == '-' && prev && p[1] && p[1] != ']') {
-		    ch = *++p;
-		    if (ch == '\\') {
-			ch = *++p;
-			if (!ch)
+		} else if (p_ch == '-' && prev_ch && p[1] && p[1] != ']') {
+		    p_ch = *++p;
+		    if (p_ch == '\\') {
+			p_ch = *++p;
+			if (!p_ch)
 			    return ABORT_ALL;
 		    }
-		    if (*text <= ch && *text >= prev)
+		    if (t_ch <= p_ch && t_ch >= prev_ch)
 			matched = TRUE;
-		    ch = 0; /* This makes "prev" get set to 0. */
-		}
-		else if (ch == '[' && p[1] == ':') {
+		    p_ch = 0; /* This makes "prev_ch" get set to 0. */
+		} else if (p_ch == '[' && p[1] == ':') {
 		    const uchar *s;
 		    int i;
-		    for (s = p += 2; (ch = *p) && ch != ']'; p++) {}
-		    if (!ch)
+		    for (s = p += 2; (p_ch = *p) && p_ch != ']'; p++) {} /*SHARED ITERATOR*/
+		    if (!p_ch)
 			return ABORT_ALL;
 		    i = p - s - 1;
 		    if (i < 0 || p[-1] != ':') {
 			/* Didn't find ":]", so treat like a normal set. */
 			p = s - 2;
-			ch = '[';
-			if (*text == ch)
+			p_ch = '[';
+			if (t_ch == p_ch)
 			    matched = TRUE;
 			continue;
 		    }
 		    if (CC_EQ(s,i, "alnum")) {
-			if (ISALNUM(*text))
+			if (ISALNUM(t_ch))
 			    matched = TRUE;
-		    }
-		    else if (CC_EQ(s,i, "alpha")) {
-			if (ISALPHA(*text))
+		    } else if (CC_EQ(s,i, "alpha")) {
+			if (ISALPHA(t_ch))
 			    matched = TRUE;
-		    }
-		    else if (CC_EQ(s,i, "blank")) {
-			if (ISBLANK(*text))
+		    } else if (CC_EQ(s,i, "blank")) {
+			if (ISBLANK(t_ch))
 			    matched = TRUE;
-		    }
-		    else if (CC_EQ(s,i, "cntrl")) {
-			if (ISCNTRL(*text))
+		    } else if (CC_EQ(s,i, "cntrl")) {
+			if (ISCNTRL(t_ch))
 			    matched = TRUE;
-		    }
-		    else if (CC_EQ(s,i, "digit")) {
-			if (ISDIGIT(*text))
+		    } else if (CC_EQ(s,i, "digit")) {
+			if (ISDIGIT(t_ch))
 			    matched = TRUE;
-		    }
-		    else if (CC_EQ(s,i, "graph")) {
-			if (ISGRAPH(*text))
+		    } else if (CC_EQ(s,i, "graph")) {
+			if (ISGRAPH(t_ch))
 			    matched = TRUE;
-		    }
-		    else if (CC_EQ(s,i, "lower")) {
-			if (ISLOWER(*text))
+		    } else if (CC_EQ(s,i, "lower")) {
+			if (ISLOWER(t_ch))
 			    matched = TRUE;
-		    }
-		    else if (CC_EQ(s,i, "print")) {
-			if (ISPRINT(*text))
+		    } else if (CC_EQ(s,i, "print")) {
+			if (ISPRINT(t_ch))
 			    matched = TRUE;
-		    }
-		    else if (CC_EQ(s,i, "punct")) {
-			if (ISPUNCT(*text))
+		    } else if (CC_EQ(s,i, "punct")) {
+			if (ISPUNCT(t_ch))
 			    matched = TRUE;
-		    }
-		    else if (CC_EQ(s,i, "space")) {
-			if (ISSPACE(*text))
+		    } else if (CC_EQ(s,i, "space")) {
+			if (ISSPACE(t_ch))
 			    matched = TRUE;
-		    }
-		    else if (CC_EQ(s,i, "upper")) {
-			if (ISUPPER(*text))
+		    } else if (CC_EQ(s,i, "upper")) {
+			if (ISUPPER(t_ch))
 			    matched = TRUE;
-		    }
-		    else if (CC_EQ(s,i, "xdigit")) {
-			if (ISXDIGIT(*text))
+		    } else if (CC_EQ(s,i, "xdigit")) {
+			if (ISXDIGIT(t_ch))
 			    matched = TRUE;
-		    }
-		    else /* malformed [:class:] string */
+		    } else /* malformed [:class:] string */
 			return ABORT_ALL;
-		    ch = 0; /* This makes "prev" get set to 0. */
-		}
-		else if (*text == ch)
+		    p_ch = 0; /* This makes "prev_ch" get set to 0. */
+		} else if (t_ch == p_ch)
 		    matched = TRUE;
-	    } while (prev = ch, (ch = *++p) != ']');
-	    if (matched == special || *text == '/')
+	    } while (prev_ch = p_ch, (p_ch = *++p) != ']');
+	    if (matched == special || t_ch == '/')
 		return FALSE;
 	    continue;
 	}
     }
 
-    return *text == '\0';
+    do {
+	if (*text)
+	    return FALSE;
+    } while ((text = *a++) != NULL);
+
+    return TRUE;
 }
 
-/* Find the pattern (p) in the text string (t). */
-int wildmatch(const char *p, const char *t)
+/* Match literal string "s" against the a virtually-joined string consisting
+ * of "text" and any strings in array "a". */
+static int doliteral(const uchar *s, const uchar *text, const uchar*const *a)
 {
+    for ( ; *s != '\0'; text++, s++) {
+	while (*text == '\0') {
+	    if ((text = *a++) == NULL)
+		return FALSE;
+	}
+	if (*text != *s)
+	    return FALSE;
+    }
+
+    do {
+	if (*text)
+	    return FALSE;
+    } while ((text = *a++) != NULL);
+
+    return TRUE;
+}
+
+/* Return the last "count" path elements from the concatenated string.
+ * We return a string pointer to the start of the string, and update the
+ * array pointer-pointer to point to any remaining string elements. */
+static const uchar *trailing_N_elements(const uchar*const **a_ptr, int count)
+{
+    const uchar*const *a = *a_ptr;
+    const uchar*const *first_a = a;
+
+    while (*a)
+	    a++;
+
+    while (a != first_a) {
+	const uchar *s = *--a;
+	s += strlen((char*)s);
+	while (--s >= *a) {
+	    if (*s == '/' && !--count) {
+		*a_ptr = a+1;
+		return s+1;
+	    }
+	}
+    }
+
+    if (count == 1) {
+	*a_ptr = a+1;
+	return *a;
+    }
+
+    return NULL;
+}
+
+/* Match the "pattern" against the "text" string. */
+int wildmatch(const char *pattern, const char *text)
+{
+    static const uchar *nomore[1]; /* A NULL pointer. */
 #ifdef WILD_TEST_ITERATIONS
     wildmatch_iteration_count = 0;
 #endif
-    return domatch((const uchar*)p, (const uchar*)t) == TRUE;
+    return dowild((const uchar*)pattern, (const uchar*)text, nomore) == TRUE;
+}
+
+/* Match the "pattern" against the forced-to-lower-case "text" string. */
+int iwildmatch(const char *pattern, const char *text)
+{
+    static const uchar *nomore[1]; /* A NULL pointer. */
+    int ret;
+#ifdef WILD_TEST_ITERATIONS
+    wildmatch_iteration_count = 0;
+#endif
+    force_lower_case = 1;
+    ret = dowild((const uchar*)pattern, (const uchar*)text, nomore) == TRUE;
+    force_lower_case = 0;
+    return ret;
+}
+
+/* Match pattern "p" against the a virtually-joined string consisting
+ * of all the pointers in array "texts" (which has a NULL pointer at the
+ * end).  The int "where" can be 0 (normal matching), > 0 (match only
+ * the trailing N slash-separated filename components of "texts"), or < 0
+ * (match the "pattern" at the start or after any slash in "texts"). */
+int wildmatch_array(const char *pattern, const char*const *texts, int where)
+{
+    const uchar *p = (const uchar*)pattern;
+    const uchar*const *a = (const uchar*const*)texts;
+    const uchar *text;
+    int matched;
+
+#ifdef WILD_TEST_ITERATIONS
+    wildmatch_iteration_count = 0;
+#endif
+
+    if (where > 0)
+	text = trailing_N_elements(&a, where);
+    else
+	text = *a++;
+    if (!text)
+	return FALSE;
+
+    if ((matched = dowild(p, text, a)) != TRUE && where < 0
+     && matched != ABORT_ALL) {
+	while (1) {
+	    if (*text == '\0') {
+		if ((text = (uchar*)*a++) == NULL)
+		    return FALSE;
+		continue;
+	    }
+	    if (*text++ == '/' && (matched = dowild(p, text, a)) != FALSE
+	     && matched != ABORT_TO_STARSTAR)
+		break;
+	}
+    }
+    return matched == TRUE;
+}
+
+/* Match literal string "s" against the a virtually-joined string consisting
+ * of all the pointers in array "texts" (which has a NULL pointer at the
+ * end).  The int "where" can be 0 (normal matching), or > 0 (match
+ * only the trailing N slash-separated filename components of "texts"). */
+int litmatch_array(const char *string, const char*const *texts, int where)
+{
+    const uchar *s = (const uchar*)string;
+    const uchar*const *a = (const uchar* const*)texts;
+    const uchar *text;
+
+    if (where > 0)
+	text = trailing_N_elements(&a, where);
+    else
+	text = *a++;
+    if (!text)
+	return FALSE;
+
+    return doliteral(s, text, a) == TRUE;
 }

lib/wildmatch.h

@@ -1,3 +1,6 @@
 /* wildmatch.h */
 
-int wildmatch(const char *p, const char *text);
+int wildmatch(const char *pattern, const char *text);
+int iwildmatch(const char *pattern, const char *text);
+int wildmatch_array(const char *pattern, const char*const *texts, int where);
+int litmatch_array(const char *string, const char*const *texts, int where);

m4/have_type.m4

@@ -0,0 +1,22 @@
+dnl AC_HAVE_TYPE(TYPE,INCLUDES)
+AC_DEFUN([AC_HAVE_TYPE], [
+AC_REQUIRE([AC_HEADER_STDC])
+cv=`echo "$1" | sed 'y%./+- %__p__%'`
+AC_MSG_CHECKING(for $1)
+AC_CACHE_VAL([ac_cv_type_$cv],
+AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
+AC_INCLUDES_DEFAULT
+$2]],
+[[$1 foo;]])],
+[eval "ac_cv_type_$cv=yes"],
+[eval "ac_cv_type_$cv=no"]))dnl
+ac_foo=`eval echo \\$ac_cv_type_$cv`
+AC_MSG_RESULT($ac_foo)
+if test "$ac_foo" = yes; then
+  ac_tr_hdr=HAVE_`echo $1 | sed 'y%abcdefghijklmnopqrstuvwxyz./- %ABCDEFGHIJKLMNOPQRSTUVWXYZ____%'`
+if false; then
+	AC_CHECK_TYPES($1)
+fi
+  AC_DEFINE_UNQUOTED($ac_tr_hdr, 1, [Define if you have type `$1'])
+fi
+])

m4/header_major_fixed.m4

@@ -0,0 +1,27 @@
+AC_DEFUN([AC_HEADER_MAJOR_FIXED],
+[AC_CACHE_CHECK(whether sys/types.h defines makedev,
+		ac_cv_header_sys_types_h_makedev,
+[AC_LINK_IFELSE([AC_LANG_PROGRAM([[@%:@include <sys/types.h>]],
+				 [[return makedev(0, 0);]])],
+		[if grep sys/sysmacros.h conftest.err >/dev/null; then
+		   ac_cv_header_sys_types_h_makedev=no
+		 else
+		   ac_cv_header_sys_types_h_makedev=yes
+		 fi],
+		[ac_cv_header_sys_types_h_makedev=no])
+])
+
+if test $ac_cv_header_sys_types_h_makedev = no; then
+AC_CHECK_HEADER(sys/mkdev.h,
+		[AC_DEFINE(MAJOR_IN_MKDEV, 1,
+			   [Define to 1 if `major', `minor', and `makedev' are
+			    declared in <mkdev.h>.])])
+
+  if test $ac_cv_header_sys_mkdev_h = no; then
+    AC_CHECK_HEADER(sys/sysmacros.h,
+		    [AC_DEFINE(MAJOR_IN_SYSMACROS, 1,
+			       [Define to 1 if `major', `minor', and `makedev'
+				are declared in <sysmacros.h>.])])
+  fi
+fi
+])

m4/socklen_t.m4

@@ -1,27 +1,3 @@
-dnl AC_VALIDATE_CACHE_SYSTEM_TYPE[(cmd)]
-dnl if the cache file is inconsistent with the current host,
-dnl target and build system types, execute CMD or print a default
-dnl error message.
-AC_DEFUN(AC_VALIDATE_CACHE_SYSTEM_TYPE, [
-    AC_REQUIRE([AC_CANONICAL_SYSTEM])
-    AC_MSG_CHECKING([config.cache system type])
-    if { test x"${ac_cv_host_system_type+set}" = x"set" &&
-         test x"$ac_cv_host_system_type" != x"$host"; } ||
-       { test x"${ac_cv_build_system_type+set}" = x"set" &&
-         test x"$ac_cv_build_system_type" != x"$build"; } ||
-       { test x"${ac_cv_target_system_type+set}" = x"set" &&
-         test x"$ac_cv_target_system_type" != x"$target"; }; then
-	AC_MSG_RESULT([different])
-	ifelse($#, 1, [$1],
-		[AC_MSG_ERROR(["you must remove config.cache and restart configure"])])
-    else
-	AC_MSG_RESULT([same])
-    fi
-    ac_cv_host_system_type="$host"
-    ac_cv_build_system_type="$build"
-    ac_cv_target_system_type="$target"
-])
-
 dnl Check for socklen_t: historically on BSD it is an int, and in
 dnl POSIX 1g it is a type of its own, but some platforms use different
 dnl types for the argument to getsockopt, getpeername, etc.  So we
@@ -42,15 +18,15 @@ AC_DEFUN([TYPE_SOCKLEN_T],
          rsync_cv_socklen_t_equiv=
          for arg2 in "struct sockaddr" void; do
             for t in int size_t unsigned long "unsigned long"; do
-               AC_TRY_COMPILE([
+               AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
 #include <sys/types.h>
 #include <sys/socket.h>
 
                   int getpeername (int, $arg2 *, $t *);
-               ],[
+               ]],[[
                   $t len;
                   getpeername(0,0,&len);
-               ],[
+               ]])],[
                   rsync_cv_socklen_t_equiv="$t"
                   break
                ])
@@ -67,5 +43,3 @@ AC_DEFUN([TYPE_SOCKLEN_T],
       [#include <sys/types.h>
 #include <sys/socket.h>])
 ])
-
-

m4/validate_cache_system_type.m4

@@ -0,0 +1,23 @@
+dnl AC_VALIDATE_CACHE_SYSTEM_TYPE[(cmd)]
+dnl if the cache file is inconsistent with the current host,
+dnl target and build system types, execute CMD or print a default
+dnl error message.
+AC_DEFUN([AC_VALIDATE_CACHE_SYSTEM_TYPE], [
+    AC_REQUIRE([AC_CANONICAL_SYSTEM])
+    AC_MSG_CHECKING([config.cache system type])
+    if { test x"${ac_cv_host_system_type+set}" = x"set" &&
+         test x"$ac_cv_host_system_type" != x"$host"; } ||
+       { test x"${ac_cv_build_system_type+set}" = x"set" &&
+         test x"$ac_cv_build_system_type" != x"$build"; } ||
+       { test x"${ac_cv_target_system_type+set}" = x"set" &&
+         test x"$ac_cv_target_system_type" != x"$target"; }; then
+	AC_MSG_RESULT([different])
+	ifelse($#, 1, [$1],
+		[AC_MSG_ERROR(["you must remove config.cache and restart configure"])])
+    else
+	AC_MSG_RESULT([same])
+    fi
+    ac_cv_host_system_type="$host"
+    ac_cv_build_system_type="$build"
+    ac_cv_target_system_type="$target"
+])

match.c

@@ -1,96 +1,95 @@
 /*
-   Copyright (C) Andrew Tridgell 1996
-   Copyright (C) Paul Mackerras 1996
-
-   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., 675 Mass Ave, Cambridge, MA 02139, USA.
-*/
+ * Block matching used by the file-transfer code.
+ *
+ * Copyright (C) 1996 Andrew Tridgell
+ * Copyright (C) 1996 Paul Mackerras
+ * Copyright (C) 2003-2020 Wayne Davison
+ *
+ * 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 3 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, visit the http://fsf.org website.
+ */
 
 #include "rsync.h"
+#include "inums.h"
 
-extern int verbose;
-extern int am_server;
-extern int do_progress;
 extern int checksum_seed;
+extern int append_mode;
+extern int xfersum_type;
 
 int updating_basis_file;
-
-typedef unsigned short tag;
-
-#define TABLESIZE (1<<16)
-#define NULL_TAG (-1)
+char sender_file_sum[MAX_DIGEST_LEN];
 
 static int false_alarms;
-static int tag_hits;
+static int hash_hits;
 static int matches;
 static int64 data_transfer;
 
 static int total_false_alarms;
-static int total_tag_hits;
+static int total_hash_hits;
 static int total_matches;
 
 extern struct stats stats;
 
-struct target {
-	tag t;
-	int32 i;
-};
+#define TRADITIONAL_TABLESIZE (1<<16)
 
-static struct target *targets;
+static uint32 tablesize;
+static int32 *hash_table;
 
-static int32 *tag_table;
-
-#define gettag2(s1,s2) (((s1) + (s2)) & 0xFFFF)
-#define gettag(sum) gettag2((sum)&0xFFFF,(sum)>>16)
-
-static int compare_targets(struct target *t1,struct target *t2)
-{
-	return (int)t1->t - (int)t2->t;
-}
+#define SUM2HASH2(s1,s2) (((s1) + (s2)) & 0xFFFF)
+#define SUM2HASH(sum) SUM2HASH2((sum)&0xFFFF,(sum)>>16)
 
+#define BIG_SUM2HASH(sum) ((sum)%tablesize)
 
 static void build_hash_table(struct sum_struct *s)
 {
+	static uint32 alloc_size;
 	int32 i;
 
-	if (!tag_table)
-		tag_table = new_array(int32, TABLESIZE);
-
-	targets = new_array(struct target, s->count);
-	if (!tag_table || !targets)
-		out_of_memory("build_hash_table");
-
-	for (i = 0; i < s->count; i++) {
-		targets[i].i = i;
-		targets[i].t = gettag(s->sums[i].sum1);
+	/* Dynamically calculate the hash table size so that the hash load
+	 * for big files is about 80%.  A number greater than the traditional
+	 * size must be odd or s2 will not be able to span the entire set. */
+	tablesize = (uint32)(s->count/8) * 10 + 11;
+	if (tablesize < TRADITIONAL_TABLESIZE)
+		tablesize = TRADITIONAL_TABLESIZE;
+	if (tablesize > alloc_size || tablesize < alloc_size - 16*1024) {
+		if (hash_table)
+			free(hash_table);
+		hash_table = new_array(int32, tablesize);
+		alloc_size = tablesize;
 	}
 
-	qsort(targets,s->count,sizeof(targets[0]),(int (*)())compare_targets);
+	memset(hash_table, 0xFF, tablesize * sizeof hash_table[0]);
 
-	for (i = 0; i < TABLESIZE; i++)
-		tag_table[i] = NULL_TAG;
-
-	for (i = s->count; i-- > 0; )
-		tag_table[targets[i].t] = i;
+	if (tablesize == TRADITIONAL_TABLESIZE) {
+		for (i = 0; i < s->count; i++) {
+			uint32 t = SUM2HASH(s->sums[i].sum1);
+			s->sums[i].chain = hash_table[t];
+			hash_table[t] = i;
+		}
+	} else {
+		for (i = 0; i < s->count; i++) {
+			uint32 t = BIG_SUM2HASH(s->sums[i].sum1);
+			s->sums[i].chain = hash_table[t];
+			hash_table[t] = i;
+		}
+	}
 }
 
 
 static OFF_T last_match;
 
 
-/**
- * Transmit a literal and/or match token.
+/* Transmit a literal and/or match token.
  *
  * This delightfully-named function is called either when we find a
  * match and need to transmit all the unmatched data leading up to it,
@@ -98,19 +97,18 @@ static OFF_T last_match;
  * transmit it.  As a result of this second case, it is called even if
  * we have not matched at all!
  *
- * @param i If >0, the number of a matched token.  If 0, indicates we
- * have only literal data.
- **/
-static void matched(int f, struct sum_struct *s, struct map_struct *buf,
-		    OFF_T offset, int32 i)
+ * If i >= 0, the number of a matched token.  If < 0, indicates we have
+ * only literal data.  A -1 will send a 0-token-int too, and a -2 sends
+ * only literal data, w/o any token-int. */
+static void matched(int f, struct sum_struct *s, struct map_struct *buf, OFF_T offset, int32 i)
 {
-	int32 n = offset - last_match; /* max value: block_size (int32) */
+	int32 n = (int32)(offset - last_match); /* max value: block_size (int32) */
 	int32 j;
 
-	if (verbose > 2 && i >= 0) {
+	if (DEBUG_GTE(DELTASUM, 2) && i >= 0) {
 		rprintf(FINFO,
-			"match at %.0f last_match=%.0f j=%d len=%ld n=%ld\n",
-			(double)offset, (double)last_match, i,
+			"match at %s last_match=%s j=%d len=%ld n=%ld\n",
+			big_num(offset), big_num(last_match), i,
 			(long)s->sums[i].len, (long)n);
 	}
 
@@ -127,13 +125,12 @@ static void matched(int f, struct sum_struct *s, struct map_struct *buf,
 		sum_update(map_ptr(buf, last_match + j, n1), n1);
 	}
 
-
 	if (i >= 0)
 		last_match = offset + s->sums[i].len;
 	else
 		last_match = offset;
 
-	if (buf && do_progress)
+	if (buf && INFO_GTE(PROGRESS, 1))
 		show_progress(last_match, buf->file_size);
 }
 
@@ -141,8 +138,8 @@ static void matched(int f, struct sum_struct *s, struct map_struct *buf,
 static void hash_search(int f,struct sum_struct *s,
 			struct map_struct *buf, OFF_T len)
 {
-	OFF_T offset, end, backup;
-	int32 k, want_i;
+	OFF_T offset, aligned_offset, end;
+	int32 k, want_i, aligned_i, backup;
 	char sum2[SUM_LENGTH];
 	uint32 s1, s2, sum;
 	int more;
@@ -152,9 +149,9 @@ static void hash_search(int f,struct sum_struct *s,
 	 * coding of the output to work more efficiently. */
 	want_i = 0;
 
-	if (verbose > 2) {
-		rprintf(FINFO, "hash search b=%ld len=%.0f\n",
-			(long)s->blength, (double)len);
+	if (DEBUG_GTE(DELTASUM, 2)) {
+		rprintf(FINFO, "hash search b=%ld len=%s\n",
+			(long)s->blength, big_num(len));
 	}
 
 	k = (int32)MIN(len, (OFF_T)s->blength);
@@ -164,33 +161,54 @@ static void hash_search(int f,struct sum_struct *s,
 	sum = get_checksum1((char *)map, k);
 	s1 = sum & 0xFFFF;
 	s2 = sum >> 16;
-	if (verbose > 3)
+	if (DEBUG_GTE(DELTASUM, 3))
 		rprintf(FINFO, "sum=%.8x k=%ld\n", sum, (long)k);
 
-	offset = 0;
+	offset = aligned_offset = aligned_i = 0;
 
 	end = len + 1 - s->sums[s->count-1].len;
 
-	if (verbose > 3) {
-		rprintf(FINFO, "hash search s->blength=%ld len=%.0f count=%.0f\n",
-			(long)s->blength, (double)len, (double)s->count);
+	if (DEBUG_GTE(DELTASUM, 3)) {
+		rprintf(FINFO, "hash search s->blength=%ld len=%s count=%s\n",
+			(long)s->blength, big_num(len), big_num(s->count));
 	}
 
 	do {
-		tag t = gettag2(s1,s2);
 		int done_csum2 = 0;
-		int32 j = tag_table[t];
+		uint32 hash_entry;
+		int32 i, *prev;
 
-		if (verbose > 4)
-			rprintf(FINFO,"offset=%.0f sum=%08x\n",(double)offset,sum);
+		if (DEBUG_GTE(DELTASUM, 4)) {
+			rprintf(FINFO, "offset=%s sum=%04x%04x\n",
+				big_num(offset), s2 & 0xFFFF, s1 & 0xFFFF);
+		}
 
-		if (j == NULL_TAG)
-			goto null_tag;
+		if (tablesize == TRADITIONAL_TABLESIZE) {
+			hash_entry = SUM2HASH2(s1,s2);
+			if ((i = hash_table[hash_entry]) < 0)
+				goto null_hash;
+			sum = (s1 & 0xffff) | (s2 << 16);
+		} else {
+			sum = (s1 & 0xffff) | (s2 << 16);
+			hash_entry = BIG_SUM2HASH(sum);
+			if ((i = hash_table[hash_entry]) < 0)
+				goto null_hash;
+		}
+		prev = &hash_table[hash_entry];
 
-		sum = (s1 & 0xffff) | (s2 << 16);
-		tag_hits++;
+		hash_hits++;
 		do {
-			int32 l, i = targets[j].i;
+			int32 l;
+
+			/* When updating in-place, the chunk's offset must be
+			 * either >= our offset or identical data at that offset.
+			 * Remove any bypassed entries that we can never use. */
+			if (updating_basis_file && s->sums[i].offset < offset
+			 && !(s->sums[i].flags & SUMFLG_SAME_OFFSET)) {
+				*prev = s->sums[i].chain;
+				continue;
+			}
+			prev = &s->sums[i].chain;
 
 			if (sum != s->sums[i].sum1)
 				continue;
@@ -200,15 +218,11 @@ static void hash_search(int f,struct sum_struct *s,
 			if (l != s->sums[i].len)
 				continue;
 
-			/* in-place: ensure chunk's offset is either >= our
-			 * offset or that the data didn't move. */
-			if (updating_basis_file && s->sums[i].offset < offset
-			    && !(s->sums[i].flags & SUMFLG_SAME_OFFSET))
-				continue;
-
-			if (verbose > 3)
-				rprintf(FINFO,"potential match at %.0f target=%.0f %.0f sum=%08x\n",
-					(double)offset,(double)j,(double)i,sum);
+			if (DEBUG_GTE(DELTASUM, 3)) {
+				rprintf(FINFO,
+					"potential match at %s i=%ld sum=%08x\n",
+					big_num(offset), (long)i, sum);
+			}
 
 			if (!done_csum2) {
 				map = (schar *)map_ptr(buf,offset,l);
@@ -223,39 +237,62 @@ static void hash_search(int f,struct sum_struct *s,
 
 			/* When updating in-place, the best possible match is
 			 * one with an identical offset, so we prefer that over
-			 * the following want_i optimization. */
+			 * the adjacent want_i optimization. */
 			if (updating_basis_file) {
-				do {
-					int32 i2 = targets[j].i;
-					if (s->sums[i2].offset != offset)
-						continue;
-					if (i2 != i) {
-						if (sum != s->sums[i2].sum1)
-							break;
-						if (memcmp(sum2, s->sums[i2].sum2,
-							   s->s2length) != 0)
-							break;
-						i = i2;
+				/* All the generator's chunks start at blength boundaries. */
+				while (aligned_offset < offset) {
+					aligned_offset += s->blength;
+					aligned_i++;
+				}
+				if ((offset == aligned_offset
+				  || (sum == 0 && l == s->blength && aligned_offset + l <= len))
+				 && aligned_i < s->count) {
+					if (i != aligned_i) {
+						if (sum != s->sums[aligned_i].sum1
+						 || l != s->sums[aligned_i].len
+						 || memcmp(sum2, s->sums[aligned_i].sum2, s->s2length) != 0)
+							goto check_want_i;
+						i = aligned_i;
 					}
-					/* This chunk was at the same offset on
-					 * both the sender and the receiver. */
+					if (offset != aligned_offset) {
+						/* We've matched some zeros in a spot that is also zeros
+						 * further along in the basis file, if we find zeros ahead
+						 * in the sender's file, we'll output enough literal data
+						 * to re-align with the basis file, and get back to seeking
+						 * instead of writing. */
+						backup = (int32)(aligned_offset - last_match);
+						if (backup < 0)
+							backup = 0;
+						map = (schar *)map_ptr(buf, aligned_offset - backup, l + backup)
+						    + backup;
+						sum = get_checksum1((char *)map, l);
+						if (sum != s->sums[i].sum1)
+							goto check_want_i;
+						get_checksum2((char *)map, l, sum2);
+						if (memcmp(sum2, s->sums[i].sum2, s->s2length) != 0)
+							goto check_want_i;
+						/* OK, we have a re-alignment match.  Bump the offset
+						 * forward to the new match point. */
+						offset = aligned_offset;
+					}
+					/* This identical chunk is in the same spot in the old and new file. */
 					s->sums[i].flags |= SUMFLG_SAME_OFFSET;
-					goto set_want_i;
-				} while (++j < s->count && targets[j].t == t);
+					want_i = i;
+				}
 			}
 
+		  check_want_i:
 			/* we've found a match, but now check to see
 			 * if want_i can hint at a better match. */
 			if (i != want_i && want_i < s->count
-			    && (!updating_basis_file || s->sums[want_i].offset >= offset
-			     || s->sums[want_i].flags & SUMFLG_SAME_OFFSET)
-			    && sum == s->sums[want_i].sum1
-			    && memcmp(sum2, s->sums[want_i].sum2, s->s2length) == 0) {
+			 && (!updating_basis_file || s->sums[want_i].offset >= offset
+			  || s->sums[want_i].flags & SUMFLG_SAME_OFFSET)
+			 && sum == s->sums[want_i].sum1
+			 && memcmp(sum2, s->sums[want_i].sum2, s->s2length) == 0) {
 				/* we've found an adjacent match - the RLL coder
 				 * will be happy */
 				i = want_i;
 			}
-		    set_want_i:
 			want_i = i + 1;
 
 			matched(f,s,buf,offset,i);
@@ -267,18 +304,17 @@ static void hash_search(int f,struct sum_struct *s,
 			s2 = sum >> 16;
 			matches++;
 			break;
-		} while (++j < s->count && targets[j].t == t);
+		} while ((i = s->sums[i].chain) >= 0);
 
-	null_tag:
-		backup = offset - last_match;
+	  null_hash:
+		backup = (int32)(offset - last_match);
 		/* We sometimes read 1 byte prior to last_match... */
 		if (backup < 0)
 			backup = 0;
 
 		/* Trim off the first byte from the checksum */
 		more = offset + k < len;
-		map = (schar *)map_ptr(buf, offset - backup, k + more + backup)
-		    + backup;
+		map = (schar *)map_ptr(buf, offset - backup, k + more + backup) + backup;
 		s1 -= map[0] + CHAR_OFFSET;
 		s2 -= k * (map[0]+CHAR_OFFSET);
 
@@ -320,53 +356,78 @@ static void hash_search(int f,struct sum_struct *s,
  **/
 void match_sums(int f, struct sum_struct *s, struct map_struct *buf, OFF_T len)
 {
-	char file_sum[MD4_SUM_LENGTH];
+	int sum_len;
 
 	last_match = 0;
 	false_alarms = 0;
-	tag_hits = 0;
+	hash_hits = 0;
 	matches = 0;
 	data_transfer = 0;
 
-	sum_init(checksum_seed);
+	sum_init(xfersum_type, checksum_seed);
+
+	if (append_mode > 0) {
+		if (append_mode == 2) {
+			OFF_T j = 0;
+			for (j = CHUNK_SIZE; j < s->flength; j += CHUNK_SIZE) {
+				if (buf && INFO_GTE(PROGRESS, 1))
+					show_progress(last_match, buf->file_size);
+				sum_update(map_ptr(buf, last_match, CHUNK_SIZE),
+					   CHUNK_SIZE);
+				last_match = j;
+			}
+			if (last_match < s->flength) {
+				int32 n = (int32)(s->flength - last_match);
+				if (buf && INFO_GTE(PROGRESS, 1))
+					show_progress(last_match, buf->file_size);
+				sum_update(map_ptr(buf, last_match, n), n);
+			}
+		}
+		last_match = s->flength;
+		s->count = 0;
+	}
 
 	if (len > 0 && s->count > 0) {
 		build_hash_table(s);
 
-		if (verbose > 2)
+		if (DEBUG_GTE(DELTASUM, 2))
 			rprintf(FINFO,"built hash table\n");
 
-		hash_search(f,s,buf,len);
+		hash_search(f, s, buf, len);
 
-		if (verbose > 2)
+		if (DEBUG_GTE(DELTASUM, 2))
 			rprintf(FINFO,"done hash search\n");
 	} else {
 		OFF_T j;
 		/* by doing this in pieces we avoid too many seeks */
-		for (j = CHUNK_SIZE; j < len; j += CHUNK_SIZE)
+		for (j = last_match + CHUNK_SIZE; j < len; j += CHUNK_SIZE)
 			matched(f, s, buf, j, -2);
 		matched(f, s, buf, len, -1);
 	}
 
-	sum_end(file_sum);
-	/* If we had a read error, send a bad checksum. */
-	if (buf && buf->status != 0)
-		file_sum[0]++;
+	sum_len = sum_end(sender_file_sum);
 
-	if (verbose > 2)
-		rprintf(FINFO,"sending file_sum\n");
-	write_buf(f,file_sum,MD4_SUM_LENGTH);
-
-	if (targets) {
-		free(targets);
-		targets=NULL;
+	/* If we had a read error, send a bad checksum.  We use all bits
+	 * off as long as the checksum doesn't happen to be that, in
+	 * which case we turn the last 0 bit into a 1. */
+	if (buf && buf->status != 0) {
+		int i;
+		for (i = 0; i < sum_len && sender_file_sum[i] == 0; i++) {}
+		memset(sender_file_sum, 0, sum_len);
+		if (i == sum_len)
+			sender_file_sum[i-1]++;
 	}
 
-	if (verbose > 2)
-		rprintf(FINFO, "false_alarms=%d tag_hits=%d matches=%d\n",
-			false_alarms, tag_hits, matches);
+	if (DEBUG_GTE(DELTASUM, 2))
+		rprintf(FINFO,"sending file_sum\n");
+	write_buf(f, sender_file_sum, sum_len);
 
-	total_tag_hits += tag_hits;
+	if (DEBUG_GTE(DELTASUM, 2)) {
+		rprintf(FINFO, "false_alarms=%d hash_hits=%d matches=%d\n",
+			false_alarms, hash_hits, matches);
+	}
+
+	total_hash_hits += hash_hits;
 	total_false_alarms += false_alarms;
 	total_matches += matches;
 	stats.literal_data += data_transfer;
@@ -374,12 +435,11 @@ void match_sums(int f, struct sum_struct *s, struct map_struct *buf, OFF_T len)
 
 void match_report(void)
 {
-	if (verbose <= 1)
+	if (!DEBUG_GTE(DELTASUM, 1))
 		return;
 
 	rprintf(FINFO,
-		"total: matches=%d  tag_hits=%d  false_alarms=%d data=%.0f\n",
-		total_matches,total_tag_hits,
-		total_false_alarms,
-		(double)stats.literal_data);
+		"total: matches=%d  hash_hits=%d  false_alarms=%d data=%s\n",
+		total_matches, total_hash_hits, total_false_alarms,
+		big_num(stats.literal_data));
 }

maybe-make-man

@@ -0,0 +1,40 @@
+#!/bin/sh
+
+if [ x"$2" = x ]; then
+    echo "Usage: $0 SRC_DIR NAME.NUM.md" 1>&2
+    exit 1
+fi
+
+srcdir="$1"
+inname="$2"
+flagfile="$srcdir/.md2man-works"
+
+if [ ! -d "$srcdir" ]; then
+    echo "The specified SRC_DIR is not a directory: $srcdir" 1>&2
+    exit 1
+fi
+
+if [ ! -f "$flagfile" ]; then
+    # We test our smallest manpage just to see if the python setup works.
+    if "$srcdir/md2man" --test "$srcdir/rsync-ssl.1.md" >/dev/null 2>&1; then
+	touch $flagfile
+    else
+	outname=`echo "$inname" | sed 's/\.md$//'`
+	if [ -f "$outname" ]; then
+	    exit 0
+	elif [ -f "$srcdir/$outname" ]; then
+	    echo "Copying $srcdir/$outname"
+	    cp -p "$srcdir/$outname" .
+	    exit 0
+	else
+	    echo "ERROR: $outname cannot be created."
+	    if [ -f "$HOME/build_farm/build_test.fns" ]; then
+		exit 0 # No exit errorno to avoid a build failure in the samba build farm
+	    else
+		exit 1
+	    fi
+	fi
+    fi
+fi
+
+"$srcdir/md2man" "$srcdir/$inname"

md2man

@@ -0,0 +1,387 @@
+#!/usr/bin/env python3
+
+# This script takes a manpage written in markdown and turns it into an html web
+# page and a nroff man page.  The input file must have the name of the program
+# and the section in this format: NAME.NUM.md.  The output files are written
+# into the current directory named NAME.NUM.html and NAME.NUM.  The input
+# format has one extra extension: if a numbered list starts at 0, it is turned
+# into a description list. The dl's dt tag is taken from the contents of the
+# first tag inside the li, which is usually a p, code, or strong tag.  The
+# cmarkgfm or commonmark lib is used to transforms the input file into html.
+# The html.parser is used as a state machine that both tweaks the html and
+# outputs the nroff data based on the html tags.
+#
+# Copyright (C) 2020 Wayne Davison
+#
+# This program is freely redistributable.
+
+import sys, os, re, argparse, subprocess, time
+from html.parser import HTMLParser
+
+CONSUMES_TXT = set('h1 h2 p li pre'.split())
+
+HTML_START = """\
+<html><head>
+<title>%s</title>
+<link href="https://fonts.googleapis.com/css2?family=Roboto&family=Roboto+Mono&display=swap" rel="stylesheet">
+<style>
+body {
+  max-width: 50em;
+  margin: auto;
+}
+body, b, strong, u {
+  font-family: 'Roboto', sans-serif;
+}
+code {
+  font-family: 'Roboto Mono', monospace;
+  font-weight: bold;
+  white-space: pre;
+}
+pre code {
+  display: block;
+  font-weight: normal;
+}
+blockquote pre code {
+  background: #f1f1f1;
+}
+dd p:first-of-type {
+  margin-block-start: 0em;
+}
+</style>
+</head><body>
+"""
+
+HTML_END = """\
+<div style="float: right"><p><i>%s</i></p></div>
+</body></html>
+"""
+
+MAN_START = r"""
+.TH "%s" "%s" "%s" "%s" "User Commands"
+""".lstrip()
+
+MAN_END = """\
+"""
+
+NORM_FONT = ('\1', r"\fP")
+BOLD_FONT = ('\2', r"\fB")
+UNDR_FONT = ('\3', r"\fI")
+NBR_DASH = ('\4', r"\-")
+NBR_SPACE = ('\xa0', r"\ ")
+
+md_parser = None
+
+def main():
+    fi = re.match(r'^(?P<fn>(?P<srcdir>.+/)?(?P<name>(?P<prog>[^/]+)\.(?P<sect>\d+))\.md)$', args.mdfile)
+    if not fi:
+        die('Failed to parse NAME.NUM.md out of input file:', args.mdfile)
+    fi = argparse.Namespace(**fi.groupdict())
+
+    if not fi.srcdir:
+        fi.srcdir = './'
+
+    fi.title = fi.prog + '(' + fi.sect + ') man page'
+    fi.mtime = 0
+
+    git_dir = fi.srcdir + '.git'
+    if os.path.lexists(git_dir):
+        fi.mtime = int(subprocess.check_output(['git', '--git-dir', git_dir, 'log', '-1', '--format=%at']))
+
+    env_subs = { 'prefix': os.environ.get('RSYNC_OVERRIDE_PREFIX', None) }
+
+    if args.test:
+        env_subs['VERSION'] = '1.0.0'
+        env_subs['libdir'] = '/usr'
+    else:
+        for fn in (fi.srcdir + 'version.h', 'Makefile'):
+            try:
+                st = os.lstat(fn)
+            except:
+                die('Failed to find', fi.srcdir + fn)
+            if not fi.mtime:
+                fi.mtime = st.st_mtime
+
+        with open(fi.srcdir + 'version.h', 'r', encoding='utf-8') as fh:
+            txt = fh.read()
+        m = re.search(r'"(.+?)"', txt)
+        env_subs['VERSION'] = m.group(1)
+
+        with open('Makefile', 'r', encoding='utf-8') as fh:
+            for line in fh:
+                m = re.match(r'^(\w+)=(.+)', line)
+                if not m:
+                    continue
+                var, val = (m.group(1), m.group(2))
+                if var == 'prefix' and env_subs[var] is not None:
+                    continue
+                while re.search(r'\$\{', val):
+                    val = re.sub(r'\$\{(\w+)\}', lambda m: env_subs[m.group(1)], val)
+                env_subs[var] = val
+                if var == 'srcdir':
+                    break
+
+    with open(fi.fn, 'r', encoding='utf-8') as fh:
+        txt = fh.read()
+
+    txt = re.sub(r'@VERSION@', env_subs['VERSION'], txt)
+    txt = re.sub(r'@LIBDIR@', env_subs['libdir'], txt)
+
+    fi.html_in = md_parser(txt)
+    txt = None
+
+    fi.date = time.strftime('%d %b %Y', time.localtime(fi.mtime))
+    fi.man_headings = (fi.prog, fi.sect, fi.date, fi.prog + ' ' + env_subs['VERSION'])
+
+    HtmlToManPage(fi)
+
+    if args.test:
+        print("The test was successful.")
+        return
+
+    for fn, txt in ((fi.name + '.html', fi.html_out), (fi.name, fi.man_out)):
+        print("Wrote:", fn)
+        with open(fn, 'w', encoding='utf-8') as fh:
+            fh.write(txt)
+
+
+def html_via_cmarkgfm(txt):
+    return cmarkgfm.markdown_to_html(txt)
+
+
+def html_via_commonmark(txt):
+    return commonmark.HtmlRenderer().render(commonmark.Parser().parse(txt))
+
+
+class HtmlToManPage(HTMLParser):
+    def __init__(self, fi):
+        HTMLParser.__init__(self, convert_charrefs=True)
+
+        st = self.state = argparse.Namespace(
+                list_state = [ ],
+                p_macro = ".P\n",
+                at_first_tag_in_li = False,
+                at_first_tag_in_dd = False,
+                dt_from = None,
+                in_pre = False,
+                in_code = False,
+                html_out = [ HTML_START % fi.title ],
+                man_out = [ MAN_START % fi.man_headings ],
+                txt = '',
+                )
+
+        self.feed(fi.html_in)
+        fi.html_in = None
+
+        st.html_out.append(HTML_END % fi.date)
+        st.man_out.append(MAN_END)
+
+        fi.html_out = ''.join(st.html_out)
+        st.html_out = None
+
+        fi.man_out = ''.join(st.man_out)
+        st.man_out = None
+
+
+    def handle_starttag(self, tag, attrs_list):
+        st = self.state
+        if args.debug:
+            self.output_debug('START', (tag, attrs_list))
+        if st.at_first_tag_in_li:
+            if st.list_state[-1] == 'dl':
+                st.dt_from = tag
+                if tag == 'p':
+                    tag = 'dt'
+                else:
+                    st.html_out.append('<dt>')
+            elif tag == 'p':
+                st.at_first_tag_in_dd = True # Kluge to suppress a .P at the start of an li.
+            st.at_first_tag_in_li = False
+        if tag == 'p':
+            if not st.at_first_tag_in_dd:
+                st.man_out.append(st.p_macro)
+        elif tag == 'li':
+            st.at_first_tag_in_li = True
+            lstate = st.list_state[-1]
+            if lstate == 'dl':
+                return
+            if lstate == 'o':
+                st.man_out.append(".IP o\n")
+            else:
+                st.man_out.append(".IP " + str(lstate) + ".\n")
+                st.list_state[-1] += 1
+        elif tag == 'blockquote':
+            st.man_out.append(".RS 4\n")
+        elif tag == 'pre':
+            st.in_pre = True
+            st.man_out.append(st.p_macro + ".nf\n")
+        elif tag == 'code' and not st.in_pre:
+            st.in_code = True
+            st.txt += BOLD_FONT[0]
+        elif tag == 'strong' or tag == 'b':
+            st.txt += BOLD_FONT[0]
+        elif tag == 'em' or  tag == 'i':
+            tag = 'u' # Change it into underline to be more like the man page
+            st.txt += UNDR_FONT[0]
+        elif tag == 'ol':
+            start = 1
+            for var, val in attrs_list:
+                if var == 'start':
+                    start = int(val) # We only support integers.
+                    break
+            if st.list_state:
+                st.man_out.append(".RS\n")
+            if start == 0:
+                tag = 'dl'
+                attrs_list = [ ]
+                st.list_state.append('dl')
+            else:
+                st.list_state.append(start)
+            st.man_out.append(st.p_macro)
+            st.p_macro = ".IP\n"
+        elif tag == 'ul':
+            st.man_out.append(st.p_macro)
+            if st.list_state:
+                st.man_out.append(".RS\n")
+                st.p_macro = ".IP\n"
+            st.list_state.append('o')
+        st.html_out.append('<' + tag + ''.join(' ' + var + '="' + htmlify(val) + '"' for var, val in attrs_list) + '>')
+        st.at_first_tag_in_dd = False
+
+
+    def handle_endtag(self, tag):
+        st = self.state
+        if args.debug:
+            self.output_debug('END', (tag,))
+        if tag in CONSUMES_TXT or st.dt_from == tag:
+            txt = st.txt.strip()
+            st.txt = ''
+        else:
+            txt = None
+        add_to_txt = None
+        if tag == 'h1':
+            st.man_out.append(st.p_macro + '.SH "' + manify(txt) + '"\n')
+        elif tag == 'h2':
+            st.man_out.append(st.p_macro + '.SS "' + manify(txt) + '"\n')
+        elif tag == 'p':
+            if st.dt_from == 'p':
+                tag = 'dt'
+                st.man_out.append('.IP "' + manify(txt) + '"\n')
+                st.dt_from = None
+            elif txt != '':
+                st.man_out.append(manify(txt) + "\n")
+        elif tag == 'li':
+            if st.list_state[-1] == 'dl':
+                if st.at_first_tag_in_li:
+                    die("Invalid 0. -> td translation")
+                tag = 'dd'
+            if txt != '':
+                st.man_out.append(manify(txt) + "\n")
+            st.at_first_tag_in_li = False
+        elif tag == 'blockquote':
+            st.man_out.append(".RE\n")
+        elif tag == 'pre':
+            st.in_pre = False
+            st.man_out.append(manify(txt) + "\n.fi\n")
+        elif (tag == 'code' and not st.in_pre):
+            st.in_code = False
+            add_to_txt = NORM_FONT[0]
+        elif tag == 'strong' or tag == 'b':
+            add_to_txt = NORM_FONT[0]
+        elif tag == 'em' or  tag == 'i':
+            tag = 'u' # Change it into underline to be more like the man page
+            add_to_txt = NORM_FONT[0]
+        elif tag == 'ol' or tag == 'ul':
+            if st.list_state.pop() == 'dl':
+                tag = 'dl'
+            if st.list_state:
+                st.man_out.append(".RE\n")
+            else:
+                st.p_macro = ".P\n"
+            st.at_first_tag_in_dd = False
+        st.html_out.append('</' + tag + '>')
+        if add_to_txt:
+            if txt is None:
+                st.txt += add_to_txt
+            else:
+                txt += add_to_txt
+        if st.dt_from == tag:
+            st.man_out.append('.IP "' + manify(txt) + '"\n')
+            st.html_out.append('</dt><dd>')
+            st.at_first_tag_in_dd = True
+            st.dt_from = None
+        elif tag == 'dt':
+            st.html_out.append('<dd>')
+            st.at_first_tag_in_dd = True
+
+
+    def handle_data(self, txt):
+        st = self.state
+        if args.debug:
+            self.output_debug('DATA', (txt,))
+        if st.in_pre:
+            html = htmlify(txt)
+        else:
+            txt = re.sub(r'\s--(\s)', NBR_SPACE[0] + r'--\1', txt).replace('--', NBR_DASH[0]*2)
+            txt = re.sub(r'(^|\W)-', r'\1' + NBR_DASH[0], txt)
+            html = htmlify(txt)
+            if st.in_code:
+                txt = re.sub(r'\s', NBR_SPACE[0], txt)
+                html = html.replace(NBR_DASH[0], '-').replace(NBR_SPACE[0], ' ') # <code> is non-breaking in CSS
+        st.html_out.append(html.replace(NBR_SPACE[0], '&nbsp;').replace(NBR_DASH[0], '-&#8288;'))
+        st.txt += txt
+
+
+    def output_debug(self, event, extra):
+        import pprint
+        st = self.state
+        if args.debug < 2:
+            st = argparse.Namespace(**vars(st))
+            if len(st.html_out) > 2:
+                st.html_out = ['...'] + st.html_out[-2:]
+            if len(st.man_out) > 2:
+                st.man_out = ['...'] + st.man_out[-2:]
+        print(event, extra)
+        pprint.PrettyPrinter(indent=2).pprint(vars(st))
+
+
+def manify(txt):
+    return re.sub(r"^(['.])", r'\&\1', txt.replace('\\', '\\\\')
+            .replace(NBR_SPACE[0], NBR_SPACE[1])
+            .replace(NBR_DASH[0], NBR_DASH[1])
+            .replace(NORM_FONT[0], NORM_FONT[1])
+            .replace(BOLD_FONT[0], BOLD_FONT[1])
+            .replace(UNDR_FONT[0], UNDR_FONT[1]), flags=re.M)
+
+
+def htmlify(txt):
+    return txt.replace('&', '&amp;').replace('<', '&lt;').replace('>', '&gt;').replace('"', '&quot;')
+
+
+def warn(*msg):
+    print(*msg, file=sys.stderr)
+
+
+def die(*msg):
+    warn(*msg)
+    sys.exit(1)
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(description='Transform a NAME.NUM.md markdown file into a NAME.NUM.html web page & a NAME.NUM man page.', add_help=False)
+    parser.add_argument('--test', action='store_true', help='Test if we can parse the input w/o updating any files.')
+    parser.add_argument('--debug', '-D', action='count', default=0, help='Output copious info on the html parsing. Repeat for even more.')
+    parser.add_argument("--help", "-h", action="help", help="Output this help message and exit.")
+    parser.add_argument('mdfile', help="The NAME.NUM.md file to parse.")
+    args = parser.parse_args()
+
+    try:
+        import cmarkgfm
+        md_parser = html_via_cmarkgfm
+    except:
+        try:
+            import commonmark
+            md_parser = html_via_commonmark
+        except:
+            die("Failed to find cmarkgfm or commonmark for python3.")
+
+    main()

mkproto.awk

@@ -1,76 +1,39 @@
-# generate prototypes for Samba C code
-# tridge, June 1996
+#!/usr/bin/awk -f
 
 BEGIN {
-  inheader=0;
-  print "/* This file is automatically generated with \"make proto\". DO NOT EDIT */"
-  print ""
+    while ((getline i < "proto.h") > 0) old_protos = old_protos ? old_protos "\n" i : i
+    protos = "/* This file is automatically generated with \"make proto\". DO NOT EDIT */\n"
 }
 
-{
-  if (inheader) {
-    if (match($0,"[)][ \t]*$")) {
-      inheader = 0;
-      printf "%s;\n",$0;
-    } else {
-      printf "%s\n",$0;
-    }
-    next;
-  }
+inheader {
+    protos = protos "\n" ((inheader = /\)[ \t]*$/ ? 0 : 1) ? $0 : $0 ";")
+    next
 }
 
-/^FN_LOCAL_BOOL/ {
-  split($0,a,"[,()]")
-  printf "BOOL %s(int );\n", a[2]
+/^FN_(LOCAL|GLOBAL)_[^(]+\([^,()]+/ {
+    local = /^FN_LOCAL/
+    gsub(/^FN_(LOC|GLOB)AL_|,.*$/, "")
+    sub(/^BOOL\(/, "BOOL ")
+    sub(/^CHAR\(/, "char ")
+    sub(/^INTEGER\(/, "int ")
+    sub(/^STRING\(/, "char *")
+    protos = protos "\n" $0 (local ? "(int module_id);" : "(void);")
+    next
 }
 
-/^FN_LOCAL_STRING/ {
-  split($0,a,"[,()]")
-  printf "char *%s(int );\n", a[2]
+/^static|^extern|;/||!/^[A-Za-z][A-Za-z0-9_]* / { next }
+
+/\(.*\)[ \t]*$/ {
+    protos = protos "\n" $0 ";"
+    next
 }
 
-/^FN_LOCAL_INT/ {
-  split($0,a,"[,()]")
-  printf "int %s(int );\n", a[2]
+/\(/ {
+    inheader = 1
+    protos = protos "\n" $0
 }
 
-/^FN_LOCAL_CHAR/ {
-  split($0,a,"[,()]")
-  printf "char %s(int );\n", a[2]
+END {
+    if (old_protos != protos) print protos > "proto.h"
+    printf "" > "proto.h-tstamp"
 }
-
-/^FN_GLOBAL_BOOL/ {
-  split($0,a,"[,()]")
-  printf "BOOL %s(void);\n", a[2]
-}
-
-/^FN_GLOBAL_STRING/ {
-  split($0,a,"[,()]")
-  printf "char *%s(void);\n", a[2]
-}
-
-/^FN_GLOBAL_INT/ {
-  split($0,a,"[,()]")
-  printf "int %s(void);\n", a[2]
-}
-
-/^static|^extern/ || !/^[a-zA-Z]/ || /[;]/ {
-  next;
-}
-
-!/^OFF_T|^size_t|^off_t|^pid_t|^unsigned|^mode_t|^DIR|^user|^int|^char|^uint|^uchar|^short|^struct|^BOOL|^void|^time|^const/ {
-  next;
-}
-
-
-/[(].*[)][ \t]*$/ {
-    printf "%s;\n",$0;
-    next;
-}
-
-/[(]/ {
-  inheader=1;
-  printf "%s\n",$0;
-  next;
-}
-

packaging/auto-Makefile

@@ -0,0 +1,8 @@
+TARGETS := all install install-ssl-daemon install-all install-strip conf gen gensend reconfigure restatus \
+	proto man clean cleantests distclean test check check29 check30 installcheck splint doxygen doxygen-upload
+
+.PHONY: $(TARGETS)
+
+$(TARGETS):
+	@if test x`packaging/prep-auto-dir` = x; then echo "auto-build-save is not setup"; exit 1; fi
+	make -C build $@

packaging/branch-from-patch

@@ -0,0 +1,174 @@
+#!/usr/bin/env -S python3 -B
+
+# This script turns one or more diff files in the patches dir (which is
+# expected to be a checkout of the rsync-patches git repo) into a branch
+# in the main rsync git checkout. This allows the applied patch to be
+# merged with the latest rsync changes and tested.  To update the diff
+# with the resulting changes, see the patch-update script.
+
+import os, sys, re, argparse, glob
+
+sys.path = ['packaging'] + sys.path
+
+from pkglib import *
+
+def main():
+    global created, info, local_branch
+
+    cur_branch, args.base_branch = check_git_state(args.base_branch, not args.skip_check, args.patches_dir)
+
+    local_branch = get_patch_branches(args.base_branch)
+
+    if args.delete_local_branches:
+        for name in sorted(local_branch):
+            branch = f"patch/{args.base_branch}/{name}"
+            cmd_chk(['git', 'branch', '-D', branch])
+        local_branch = set()
+
+    if args.add_missing:
+        for fn in sorted(glob.glob(f"{args.patches_dir}/*.diff")):
+            name = re.sub(r'\.diff$', '', re.sub(r'.+/', '', fn))
+            if name not in local_branch and fn not in args.patch_files:
+                args.patch_files.append(fn)
+
+    if not args.patch_files:
+        return
+
+    for fn in args.patch_files:
+        if not fn.endswith('.diff'):
+            die(f"Filename is not a .diff file: {fn}")
+        if not os.path.isfile(fn):
+            die(f"File not found: {fn}")
+
+    scanned = set()
+    info = { }
+
+    patch_list = [ ]
+    for fn in args.patch_files:
+        m = re.match(r'^(?P<dir>.*?)(?P<name>[^/]+)\.diff$', fn)
+        patch = argparse.Namespace(**m.groupdict())
+        if patch.name in scanned:
+            continue
+        patch.fn = fn
+
+        lines = [ ]
+        commit_hash = None
+        with open(patch.fn, 'r', encoding='utf-8') as fh:
+            for line in fh:
+                m = re.match(r'^based-on: (\S+)', line)
+                if m:
+                    commit_hash = m[1]
+                    break
+                if (re.match(r'^index .*\.\..* \d', line)
+                 or re.match(r'^diff --git ', line)
+                 or re.match(r'^--- (old|a)/', line)):
+                    break
+                lines.append(re.sub(r'\s*\Z', "\n", line, 1))
+        info_txt = ''.join(lines).strip() + "\n"
+        lines = None
+
+        parent = args.base_branch
+        patches = re.findall(r'patch -p1 <%s/(\S+)\.diff' % args.patches_dir, info_txt)
+        if patches:
+            last = patches.pop()
+            if last != patch.name:
+                warn(f"No identity patch line in {patch.fn}")
+                patches.append(last)
+            if patches:
+                parent = patches.pop()
+                if parent not in scanned:
+                    diff_fn = patch.dir + parent + '.diff'
+                    if not os.path.isfile(diff_fn):
+                        die(f"Failed to find parent of {patch.fn}: {parent}")
+                    # Add parent to args.patch_files so that we will look for the
+                    # parent's parent.  Any duplicates will be ignored.
+                    args.patch_files.append(diff_fn)
+        else:
+            warn(f"No patch lines found in {patch.fn}")
+
+        info[patch.name] = [ parent, info_txt, commit_hash ]
+
+        patch_list.append(patch)
+
+    created = set()
+    for patch in patch_list:
+        create_branch(patch)
+
+    cmd_chk(['git', 'checkout', args.base_branch])
+
+
+def create_branch(patch):
+    if patch.name in created:
+        return
+    created.add(patch.name)
+
+    parent, info_txt, commit_hash = info[patch.name]
+    parent = argparse.Namespace(dir=patch.dir, name=parent, fn=patch.dir + parent + '.diff')
+
+    if parent.name == args.base_branch:
+        parent_branch = commit_hash if commit_hash else args.base_branch
+    else:
+        create_branch(parent)
+        parent_branch = '/'.join(['patch', args.base_branch, parent.name])
+
+    branch = '/'.join(['patch', args.base_branch, patch.name])
+    print("\n" + '=' * 64)
+    print(f"Processing {branch} ({parent_branch})")
+
+    if patch.name in local_branch:
+        cmd_chk(['git', 'branch', '-D', branch])
+
+    cmd_chk(['git', 'checkout', '-b', branch, parent_branch])
+
+    info_fn = 'PATCH.' + patch.name
+    with open(info_fn, 'w', encoding='utf-8') as fh:
+        fh.write(info_txt)
+    cmd_chk(['git', 'add', info_fn])
+
+    with open(patch.fn, 'r', encoding='utf-8') as fh:
+        patch_txt = fh.read()
+
+    cmd_run('patch -p1'.split(), input=patch_txt)
+
+    for fn in glob.glob('*.orig') + glob.glob('*/*.orig'):
+        os.unlink(fn)
+
+    pos = 0
+    new_file_re = re.compile(r'\nnew file mode (?P<mode>\d+)\s+--- /dev/null\s+\+\+\+ b/(?P<fn>.+)')
+    while True:
+        m = new_file_re.search(patch_txt, pos)
+        if not m:
+            break
+        os.chmod(m['fn'], int(m['mode'], 8))
+        cmd_chk(['git', 'add', m['fn']])
+        pos = m.end()
+
+    while True:
+        cmd_chk('git status'.split())
+        ans = input('Press Enter to commit, Ctrl-C to abort, or type a wild-name to add a new file: ')
+        if ans == '':
+            break
+        cmd_chk("git add " + ans, shell=True)
+
+    while True:
+        s = cmd_run(['git', 'commit', '-a', '-m', f"Creating branch from {patch.name}.diff."])
+        if not s.returncode:
+            break
+        s = cmd_run(['/bin/zsh'])
+        if s.returncode:
+            die('Aborting due to shell error code')
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(description="Create a git patch branch from an rsync patch file.", add_help=False)
+    parser.add_argument('--branch', '-b', dest='base_branch', metavar='BASE_BRANCH', default='master', help="The branch the patch is based on. Default: master.")
+    parser.add_argument('--add-missing', '-a', action='store_true', help="Add a branch for every patches/*.diff that doesn't have a branch.")
+    parser.add_argument('--skip-check', action='store_true', help="Skip the check that ensures starting with a clean branch.")
+    parser.add_argument('--delete', dest='delete_local_branches', action='store_true', help="Delete all the local patch/BASE/* branches, not just the ones that are being recreated.")
+    parser.add_argument('--patches-dir', '-p', metavar='DIR', default='patches', help="Override the location of the rsync-patches dir. Default: patches.")
+    parser.add_argument('patch_files', metavar='patches/DIFF_FILE', nargs='*', help="Specify what patch diff files to process. Default: all of them.")
+    parser.add_argument("--help", "-h", action="help", help="Output this help message and exit.")
+    args = parser.parse_args()
+    main()
+
+# vim: sw=4 et ft=python

packaging/cull_options

@@ -0,0 +1,85 @@
+#!/usr/bin/env perl
+# This script outputs some perl code that parses all possible options
+# that the code in options.c might send to the server.  This perl code
+# is included in the rrsync script.
+use strict;
+
+our %short_no_arg;
+our %short_with_num = ( '@' => 1 );
+our %long_opt = ( # These include some extra long-args that BackupPC uses:
+    'block-size' => 1,
+    'daemon' => -1,
+    'debug' => 1,
+    'fake-super' => 0,
+    'fuzzy' => 0,
+    'group' => 0,
+    'hard-links' => 0,
+    'ignore-times' => 0,
+    'info' => 1,
+    'links' => 0,
+    'log-file' => 3,
+    'one-file-system' => 0,
+    'owner' => 0,
+    'perms' => 0,
+    'recursive' => 0,
+    'times' => 0,
+    'write-devices' => -1,
+);
+our $last_long_opt;
+
+open(IN, '../options.c') or die "Unable to open ../options.c: $!\n";
+
+while (<IN>) {
+    if (/\Qargstr[x++]\E = '([^.ie])'/) {
+	$short_no_arg{$1} = 1;
+	undef $last_long_opt;
+    } elsif (/\Qasprintf(\E[^,]+, "-([a-zA-Z0-9])\%l?[ud]"/) {
+	$short_with_num{$1} = 1;
+	undef $last_long_opt;
+    } elsif (/\Qargs[ac++]\E = "--([^"=]+)"/) {
+	$last_long_opt = $1;
+	$long_opt{$1} = 0 unless exists $long_opt{$1};
+    } elsif (defined($last_long_opt)
+	&& /\Qargs[ac++]\E = ([^["\s]+);/) {
+	$long_opt{$last_long_opt} = 2;
+	undef $last_long_opt;
+    } elsif (/return "--([^"]+-dest)";/) {
+	$long_opt{$1} = 2;
+	undef $last_long_opt;
+    } elsif (/\Qasprintf(\E[^,]+, "--([^"=]+)=/ || /\Qargs[ac++]\E = "--([^"=]+)=/ || /fmt = .*: "--([^"=]+)=/) {
+	$long_opt{$1} = 1;
+	undef $last_long_opt;
+    }
+}
+close IN;
+
+my $short_no_arg = join('', sort keys %short_no_arg);
+my $short_with_num = join('', sort keys %short_with_num);
+
+print <<EOT;
+
+# These options are the only options that rsync might send to the server,
+# and only in the option format that the stock rsync produces.
+
+# To disable a short-named option, add its letter to this string:
+our \$short_disabled = 's';
+
+our \$short_no_arg = '$short_no_arg'; # DO NOT REMOVE ANY
+our \$short_with_num = '$short_with_num'; # DO NOT REMOVE ANY
+
+# To disable a long-named option, change its value to a -1.  The values mean:
+# 0 = the option has no arg; 1 = the arg doesn't need any checking; 2 = only
+# check the arg when receiving; and 3 = always check the arg.
+our \%long_opt = (
+EOT
+
+foreach my $opt (sort keys %long_opt) {
+    my $val = $long_opt{$opt};
+    $val = 1 if $opt =~ /^(max-|min-)/;
+    $val = 3 if $opt eq 'files-from';
+    $val = q"$only eq 'r' ? -1 : " . $val if $opt =~ /^(remove-|log-file)/;
+    $val = q"$only eq 'w' ? -1 : " . $val if $opt eq 'sender';
+    print "  '$opt' => $val,\n";
+}
+
+print ");\n\n";

packaging/lsb/rsync.spec

@@ -1,88 +1,87 @@
-Summary: A program for synchronizing files over a network.
+Summary: A fast, versatile, remote (and local) file-copying tool
 Name: rsync
-Version: 2.6.4
-Release: 1
+Version: 3.2.2
+%define fullversion %{version}pre2
+Release: 0.1.pre2
+%define srcdir src-previews
 Group: Applications/Internet
-Source:	ftp://rsync.samba.org/pub/rsync/rsync-%{version}.tar.gz
-URL: http://rsync.samba.org/
+License: GPL
+Source0: https://rsync.samba.org/ftp/rsync/%{srcdir}/rsync-%{fullversion}.tar.gz
+#Source1: https://rsync.samba.org/ftp/rsync/%{srcdir}/rsync-patches-%{fullversion}.tar.gz
+URL: https://rsync.samba.org/
 
 Prefix: %{_prefix}
 BuildRoot: /var/tmp/%{name}-root
-License: GPL
+
+%package ssl-daemon
+Summary: An stunnel config file to support ssl rsync daemon connections.
+Group: Applications/Internet
+Requires: rsync, stunnel >= 4
 
 %description
-Rsync uses a reliable algorithm to bring remote and host files into
-sync very quickly. Rsync is fast because it just sends the differences
-in the files over the network instead of sending the complete
-files. Rsync is often used as a very powerful mirroring process or
-just as a more capable replacement for the rcp command. A technical
-report which describes the rsync algorithm is included in this
-package.
+Rsync is a fast and extraordinarily versatile file copying tool.  It can
+copy locally, to/from another host over any remote shell, or to/from a
+remote rsync daemon.  It offers a large number of options that control
+every aspect of its behavior and permit very flexible specification of the
+set of files to be copied.  It is famous for its delta-transfer algorithm,
+which reduces the amount of data sent over the network by sending only the
+differences between the source files and the existing files in the
+destination.  Rsync is widely used for backups and mirroring and as an
+improved copy command for everyday use.
+
+%description ssl-daemon
+Provides a config file for stunnel that will (if you start your stunnel
+service) cause stunnel to listen for ssl rsync-daemon connections and run
+"rsync --daemon" to handle them.
 
 %prep
-%setup -q
+# Choose one -- setup source only, or setup source + rsync-patches:
+%setup -q -n rsync-%{fullversion}
+#%setup -q -b1 -n rsync-%{fullversion}
+
+# If you you used "%setup -q -b1 ...", choose the patches you wish to apply:
+#patch -p1 <patches/acls.diff
+#patch -p1 <patches/xattrs.diff
+#patch -p1 <patches/remote-option.diff
+#patch -p1 <patches/db.diff
+
+# Avoid extra perl dependencies for scripts going into doc dir.
+chmod -x support/*
 
 %build
+#./prepare-source
 %configure
 
 make
 
 %install
 rm -rf $RPM_BUILD_ROOT
+make install install-ssl-daemon DESTDIR=$RPM_BUILD_ROOT
 
-%makeinstall
+mkdir -p $RPM_BUILD_ROOT/etc/xinetd.d $RPM_BUILD_ROOT/etc/rsync-ssl/certs
+install -m 644 packaging/lsb/rsync.xinetd $RPM_BUILD_ROOT/etc/xinetd.d/rsync
 
 %clean
 rm -rf $RPM_BUILD_ROOT
 
 %files
 %defattr(-,root,root)
-%doc COPYING README tech_report.tex
+%doc COPYING NEWS.md README.md support/ tech_report.tex
+%config(noreplace) /etc/xinetd.d/rsync
 %{_prefix}/bin/rsync
+%{_prefix}/bin/rsync-ssl
 %{_mandir}/man1/rsync.1*
+%{_mandir}/man1/rsync-ssl.1*
 %{_mandir}/man5/rsyncd.conf.5*
 
+%files ssl-daemon
+%config(noreplace) /etc/stunnel/rsyncd.conf
+%dir /etc/rsync-ssl/certs
+
 %changelog
-* Thu Jan 30 2003 Horst von Brand <vonbrand@inf.utfsm.cl>
-  Fixed "Sept" date in %changelog here
-  Use %{_mandir} to point to manpages
-  Support for compressed manpages (* at end catches them in %files)
-  Add doc/README-SGML and doc/rsync.sgml to %doc
+* Sun Jun 28 2020 Wayne Davison <wayne@opencoder.net>
+Released 3.2.2pre2.
 
-* Mon Sep 11 2000 John H Terpstra <jht@turbolinux.com>
-  Changed target paths to be Linux Standards Base compliant
-
-* Mon Jan 25 1999 Stefan Hornburg <racke@linuxia.de>
-  quoted RPM_OPT_FLAGS for the sake of robustness
-
-* Mon May 18 1998 Andrew Tridgell <tridge@samba.anu.edu.au>
-  reworked for auto-building when I release rsync (tridge@samba.anu.edu.au)
-
-* Sat May 16 1998 John H Terpstra <jht@aquasoft.com.au>
-  Upgraded to Rsync 2.0.6
-    -new feature anonymous rsync
-
-* Mon Apr  6 1998 Douglas N. Arnold <dna@math.psu.edu>
-
-Upgrade to rsync version 1.7.2.
-
-* Sun Mar  1 1998 Douglas N. Arnold <dna@math.psu.edu>
-
-Built 1.6.9-1 based on the 1.6.3-2 spec file of John A. Martin.
-Changes from 1.6.3-2 packaging: added latex and dvips commands
-to create tech_report.ps.
-
-* Mon Aug 25 1997 John A. Martin <jam@jamux.com>
-
-Built 1.6.3-2 after finding no rsync-1.6.3-1.src.rpm although there
-was an ftp://ftp.redhat.com/pub/contrib/alpha/rsync-1.6.3-1.alpha.rpm
-showing no packager nor signature but giving 
-"Source RPM: rsync-1.6.3-1.src.rpm".
-
-Changes from 1.6.2-1 packaging: added '$RPM_OPT_FLAGS' to make, strip
-to '%build', removed '%prefix'.
-
-* Thu Apr 10 1997 Michael De La Rue <miked@ed.ac.uk>
-
-rsync-1.6.2-1 packaged.  (This entry by jam to credit Michael for the
-previous package(s).)
+* Fri Mar 21 2008 Wayne Davison <wayne@opencoder.net>
+Added installation of /etc/xinetd.d/rsync file and some commented-out
+lines that demonstrate how to use the rsync-patches tar file.

packaging/lsb/rsync.xinetd

@@ -0,0 +1,13 @@
+# default: off
+# description: The rsync server is a good addition to an ftp server, as it
+#	allows crc checksumming etc.
+service rsync
+{
+	disable	= yes
+	socket_type     = stream
+	wait            = no
+	user            = root
+	server          = /usr/bin/rsync
+	server_args     = --daemon
+	log_on_failure  += USERID
+}

packaging/md2html

@@ -0,0 +1,99 @@
+#!/usr/bin/env python3
+
+# Copyright (C) 2020 Wayne Davison
+#
+# This program is freely redistributable.
+
+import os, re, argparse
+
+HTML_START = """\
+<html><head>
+<title>%s</title>
+<link href="https://fonts.googleapis.com/css2?family=Roboto&family=Roboto+Mono&display=swap" rel="stylesheet">
+<style>
+body {
+  max-width: 50em;
+  margin: auto;
+}
+body, b, strong, u {
+  font-family: 'Roboto', sans-serif;
+}
+code {
+  font-family: 'Roboto Mono', monospace;
+  font-weight: bold;
+}
+pre code {
+  display: block;
+  font-weight: normal;
+}
+blockquote pre code {
+  background: #f1f1f1;
+}
+dd p:first-of-type {
+  margin-block-start: 0em;
+}
+</style>
+</head><body>
+"""
+
+HTML_END = """\
+</body></html>
+"""
+
+md_parser = None
+
+def main():
+    for mdfn in args.mdfiles:
+        if not mdfn.endswith('.md'):
+            print('Ignoring non-md input file:', mdfn)
+            continue
+        title = re.sub(r'.*/', '', mdfn).replace('.md', '')
+        htfn = mdfn.replace('.md', '.html')
+
+        print("Parsing", mdfn, '->', htfn)
+
+        with open(mdfn, 'r', encoding='utf-8') as fh:
+            txt = fh.read()
+
+        txt = re.sub(r'\s--\s', '\xa0-- ', txt)
+
+        html = md_parser(txt)
+
+        html = re.sub(r'(<code>)([\s\S]*?)(</code>)', lambda m: m[1] + re.sub(r'\s', '\xa0', m[2]) + m[3], html)
+        html = html.replace('--', '&#8209;&#8209;').replace("\xa0-", '&nbsp;&#8209;').replace("\xa0", '&nbsp;')
+        html = re.sub(r'(\W)-', r'\1&#8209;', html)
+
+        if os.path.lexists(htfn):
+            os.unlink(htfn)
+
+        with open(htfn, 'w', encoding='utf-8') as fh:
+            fh.write(HTML_START % title)
+            fh.write(html)
+            fh.write(HTML_END)
+
+
+def html_via_cmarkgfm(txt):
+    return cmarkgfm.markdown_to_html(txt)
+
+
+def html_via_commonmark(txt):
+    return commonmark.HtmlRenderer().render(commonmark.Parser().parse(txt))
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(description='Output html for md pages.', add_help=False)
+    parser.add_argument("--help", "-h", action="help", help="Output this help message and exit.")
+    parser.add_argument("mdfiles", nargs='+', help="The .md files to turn into .html files.")
+    args = parser.parse_args()
+
+    try:
+        import cmarkgfm
+        md_parser = html_via_cmarkgfm
+    except:
+        try:
+            import commonmark
+            md_parser = html_via_commonmark
+        except:
+            die("Failed to find cmarkgfm or commonmark for python3.")
+
+    main()

packaging/patch-update

@@ -0,0 +1,218 @@
+#!/usr/bin/env -S python3 -B
+
+# This script is used to turn one or more of the "patch/BASE/*" branches
+# into one or more diffs in the "patches" directory.  Pass the option
+# --gen if you want generated files in the diffs.  Pass the name of
+# one or more diffs if you want to just update a subset of all the
+# diffs.
+
+import os, sys, re, argparse, time, shutil
+
+sys.path = ['packaging'] + sys.path
+
+from pkglib import *
+
+MAKE_GEN_CMDS = [
+        './prepare-source'.split(),
+        'cd build && if test -f config.status ; then ./config.status ; else ../configure ; fi',
+        'make -C build gen'.split(),
+        ]
+TMP_DIR = "patches.gen"
+
+os.environ['GIT_MERGE_AUTOEDIT'] = 'no'
+
+def main():
+    global master_commit, parent_patch, description, completed, last_touch
+
+    if not os.path.isdir(args.patches_dir):
+        die(f'No "{args.patches_dir}" directory was found.')
+    if not os.path.isdir('.git'):
+        die('No ".git" directory present in the current dir.')
+
+    starting_branch, args.base_branch = check_git_state(args.base_branch, not args.skip_check, args.patches_dir)
+
+    master_commit = latest_git_hash(args.base_branch)
+
+    if cmd_txt_chk(['packaging/prep-auto-dir']) == '':
+        die('You must setup an auto-build-save dir to use this script.')
+
+    if args.gen:
+        if os.path.lexists(TMP_DIR):
+            die(f'"{TMP_DIR}" must not exist in the current directory.')
+        gen_files = get_gen_files()
+        os.mkdir(TMP_DIR, 0o700)
+        for cmd in MAKE_GEN_CMDS:
+            cmd_chk(cmd)
+        cmd_chk(['rsync', '-a', *gen_files, f'{TMP_DIR}/master/'])
+
+    last_touch = int(time.time())
+
+    # Start by finding all patches so that we can load all possible parents.
+    patches = sorted(list(get_patch_branches(args.base_branch)))
+
+    parent_patch = { }
+    description = { }
+
+    for patch in patches:
+        branch = f"patch/{args.base_branch}/{patch}"
+        desc = ''
+        proc = cmd_pipe(['git', 'diff', '-U1000', f"{args.base_branch}...{branch}", '--', f"PATCH.{patch}"])
+        in_diff = False
+        for line in proc.stdout:
+            if in_diff:
+                if not re.match(r'^[ +]', line):
+                    continue
+                line = line[1:]
+                m = re.search(r'patch -p1 <patches/(\S+)\.diff', line)
+                if m and m[1] != patch:
+                    parpat = parent_patch[patch] = m[1]
+                    if not parpat in patches:
+                        die(f"Parent of {patch} is not a local branch: {parpat}")
+                desc += line
+            elif re.match(r'^@@ ', line):
+                in_diff = True
+        description[patch] = desc
+        proc.communicate()
+
+    if args.patch_files: # Limit the list of patches to actually process
+        valid_patches = patches
+        patches = [ ]
+        for fn in args.patch_files:
+            name = re.sub(r'\.diff$', '', re.sub(r'.+/', '', fn))
+            if name not in valid_patches:
+                die(f"Local branch not available for patch: {name}")
+            patches.append(name)
+
+    completed = set()
+
+    for patch in patches:
+        if patch in completed:
+            continue
+        if not update_patch(patch):
+            break
+
+    if args.gen:
+        shutil.rmtree(TMP_DIR)
+
+    while last_touch >= int(time.time()):
+        time.sleep(1)
+    cmd_chk(['git', 'checkout', starting_branch])
+    cmd_chk(['packaging/prep-auto-dir'], discard='output')
+
+
+def update_patch(patch):
+    global last_touch
+
+    completed.add(patch) # Mark it as completed early to short-circuit any (bogus) dependency loops.
+
+    parent = parent_patch.get(patch, None)
+    if parent:
+        if parent not in completed:
+            if not update_patch(parent):
+                return 0
+        based_on = parent = f"patch/{args.base_branch}/{parent}"
+    else:
+        parent = args.base_branch
+        based_on = master_commit
+
+    print(f"======== {patch} ========")
+
+    while args.gen and last_touch >= int(time.time()):
+        time.sleep(1)
+
+    branch = f"patch/{args.base_branch}/{patch}"
+    s = cmd_run(['git', 'checkout', branch])
+    if s.returncode != 0:
+        return 0
+
+    s = cmd_run(['git', 'merge', based_on])
+    ok = s.returncode == 0
+    if not ok or args.shell:
+        cmd_chk(['packaging/prep-auto-dir'], discard='output')
+        m = re.search(r'([^/]+)$', parent)
+        parent_dir = m[1]
+        if not ok:
+            print(f'"git merge {based_on}" incomplete -- please fix.')
+        os.environ['PS1'] = f"[{parent_dir}] {patch}: "
+        while True:
+            s = cmd_run([os.environ.get('SHELL', '/bin/sh')])
+            if s.returncode != 0:
+                ans = input("Abort? [n/y] ")
+                if re.match(r'^y', ans, flags=re.I):
+                    return 0
+                continue
+            cur_branch, is_clean, status_txt = check_git_status(0)
+            if is_clean:
+                break
+            print(status_txt, end='')
+        cmd_run('rm -f build/*.o build/*/*.o')
+
+    with open(f"{args.patches_dir}/{patch}.diff", 'w', encoding='utf-8') as fh:
+        fh.write(description[patch])
+        fh.write(f"\nbased-on: {based_on}\n")
+
+        if args.gen:
+            gen_files = get_gen_files()
+            for cmd in MAKE_GEN_CMDS:
+                cmd_chk(cmd)
+            cmd_chk(['rsync', '-a', *gen_files, f"{TMP_DIR}/{patch}/"])
+        else:
+            gen_files = [ ]
+        last_touch = int(time.time())
+
+        proc = cmd_pipe(['git', 'diff', based_on])
+        skipping = False
+        for line in proc.stdout:
+            if skipping:
+                if not re.match(r'^diff --git a/', line):
+                    continue
+                skipping = False
+            elif re.match(r'^diff --git a/PATCH', line):
+                skipping = True
+                continue
+            if not re.match(r'^index ', line):
+                fh.write(line)
+        proc.communicate()
+
+        if args.gen:
+            e_tmp_dir = re.escape(TMP_DIR)
+            diff_re  = re.compile(r'^(diff -Nurp) %s/[^/]+/(.*?) %s/[^/]+/(.*)' % (e_tmp_dir, e_tmp_dir))
+            minus_re = re.compile(r'^\-\-\- %s/[^/]+/([^\t]+)\t.*' % e_tmp_dir)
+            plus_re  = re.compile(r'^\+\+\+ %s/[^/]+/([^\t]+)\t.*' % e_tmp_dir)
+
+            if parent == args.base_branch:
+                parent_dir = 'master'
+            else:
+                m = re.search(r'([^/]+)$', parent)
+                parent_dir = m[1]
+
+            proc = cmd_pipe(['diff', '-Nurp', f"{TMP_DIR}/{parent_dir}", f"{TMP_DIR}/{patch}"])
+            for line in proc.stdout:
+                line = diff_re.sub(r'\1 a/\2 b/\3', line)
+                line = minus_re.sub(r'--- a/\1', line)
+                line =  plus_re.sub(r'+++ b/\1', line)
+                fh.write(line)
+            proc.communicate()
+            for fn in gen_files:
+                os.unlink(fn)
+
+    return 1
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(description="Turn a git branch back into a diff files in the patches dir.", add_help=False)
+    parser.add_argument('--branch', '-b', dest='base_branch', metavar='BASE_BRANCH', default='master', help="The branch the patch is based on. Default: master.")
+    parser.add_argument('--skip-check', action='store_true', help="Skip the check that ensures starting with a clean branch.")
+    parser.add_argument('--shell', '-s', action='store_true', help="Launch a shell for every patch/BASE/* branch updated, not just when a conflict occurs.")
+    parser.add_argument('--gen', metavar='DIR', nargs='?', const='', help='Include generated files. Optional DIR value overrides the default of using the "patches" dir.')
+    parser.add_argument('--patches-dir', '-p', metavar='DIR', default='patches', help="Override the location of the rsync-patches dir. Default: patches.")
+    parser.add_argument('patch_files', metavar='patches/DIFF_FILE', nargs='*', help="Specify what patch diff files to process. Default: all of them.")
+    parser.add_argument("--help", "-h", action="help", help="Output this help message and exit.")
+    args = parser.parse_args()
+    if args.gen == '':
+        args.gen = args.patches_dir
+    elif args.gen is not None:
+        args.patches_dir = args.gen
+    main()
+
+# vim: sw=4 et ft=python

packaging/pkglib.py

@@ -0,0 +1,266 @@
+import os, sys, re, subprocess
+
+# This python3 library provides a few helpful routines that are
+# used by the latest packaging scripts.
+
+default_encoding = 'utf-8'
+
+# Output the msg args to stderr.  Accepts all the args that print() accepts.
+def warn(*msg):
+    print(*msg, file=sys.stderr)
+
+
+# Output the msg args to stderr and die with a non-zero return-code.
+# Accepts all the args that print() accepts.
+def die(*msg):
+    warn(*msg)
+    sys.exit(1)
+
+
+# Set this to an encoding name or set it to None to avoid the default encoding idiom.
+def set_default_encoding(enc):
+    default_encoding = enc
+
+
+# Set shell=True if the cmd is a string; sets a default encoding unless raw=True was specified.
+def _tweak_opts(cmd, opts, **maybe_set):
+    # This sets any maybe_set value that isn't already set AND creates a copy of opts for us.
+    opts = {**maybe_set, **opts}
+
+    if type(cmd) == str:
+        opts = {'shell': True, **opts}
+
+    want_raw = opts.pop('raw', False)
+    if default_encoding and not want_raw:
+        opts = {'encoding': default_encoding, **opts}
+
+    capture = opts.pop('capture', None)
+    if capture:
+        if capture == 'stdout':
+            opts = {'stdout': subprocess.PIPE, **opts}
+        elif capture == 'stderr':
+            opts = {'stderr': subprocess.PIPE, **opts}
+        elif capture == 'output':
+            opts = {'stdout': subprocess.PIPE, 'stderr': subprocess.PIPE, **opts}
+        elif capture == 'combined':
+            opts = {'stdout': subprocess.PIPE, 'stderr': subprocess.STDOUT, **opts}
+
+    discard = opts.pop('discard', None)
+    if discard:
+        # We DO want to override any already set stdout|stderr values (unlike above).
+        if discard == 'stdout' or discard == 'output':
+            opts['stdout'] = subprocess.DEVNULL
+        if discard == 'stderr' or discard == 'output':
+            opts['stderr'] = subprocess.DEVNULL
+
+    return opts
+
+
+# This does a normal subprocess.run() with some auto-args added to make life easier.
+def cmd_run(cmd, **opts):
+    return subprocess.run(cmd, **_tweak_opts(cmd, opts))
+
+
+# Like cmd_run() with a default check=True specified.
+def cmd_chk(cmd, **opts):
+    return subprocess.run(cmd, **_tweak_opts(cmd, opts, check=True))
+
+
+# Capture stdout in a string and return the (output, return_code) tuple.
+# Use capture='combined' opt to get both stdout and stderr together.
+def cmd_txt_status(cmd, **opts):
+    input = opts.pop('input', None)
+    if input is not None:
+        opts['stdin'] = subprocess.PIPE
+    proc = subprocess.Popen(cmd, **_tweak_opts(cmd, opts, capture='stdout'))
+    out = proc.communicate(input=input)[0]
+    return (out, proc.returncode)
+
+
+# Like cmd_txt_status() but just return the output.
+def cmd_txt(cmd, **opts):
+    return cmd_txt_status(cmd, **opts)[0]
+
+
+# Capture stdout in a string and return the output if the command has a 0 return code.
+# Otherwise it throws an exception that indicates the return code and the output.
+def cmd_txt_chk(cmd, **opts):
+    out, rc = cmd_txt_status(cmd, **opts)
+    if rc != 0:
+        cmd_err = f'Command "{cmd}" returned non-zero exit status "{rc}" and output:\n{out}'
+        raise Exception(cmd_err)
+    return out
+
+
+# Starts a piped-output command of stdout (by default) and leaves it up to you to read
+# the output and call communicate() on the returned object.
+def cmd_pipe(cmd, **opts):
+    return subprocess.Popen(cmd, **_tweak_opts(cmd, opts, capture='stdout'))
+
+
+# Runs a "git status" command and dies if the checkout is not clean (the
+# arg fatal_unless_clean can be used to make that non-fatal.  Returns a
+# tuple of the current branch, the is_clean flag, and the status text.
+def check_git_status(fatal_unless_clean=True, subdir='.'):
+    status_txt = cmd_txt_chk(f"cd '{subdir}' && git status")
+    is_clean = re.search(r'\nnothing to commit.+working (directory|tree) clean', status_txt) != None
+
+    if not is_clean and fatal_unless_clean:
+        if subdir == '.':
+            subdir = ''
+        else:
+            subdir = f" *{subdir}*"
+        die(f"The{subdir} checkout is not clean:\n" + status_txt)
+
+    m = re.match(r'^(?:# )?On branch (.+)\n', status_txt)
+    cur_branch = m[1] if m else None
+
+    return (cur_branch, is_clean, status_txt)
+
+
+# Calls check_git_status() on the current git checkout and (optionally) a subdir path's
+# checkout. Use fatal_unless_clean to indicate if an unclean checkout is fatal or not.
+# The master_branch arg indicates what branch we want both checkouts to be using, and
+# if the branch is wrong the user is given the option of either switching to the right
+# branch or aborting.
+def check_git_state(master_branch, fatal_unless_clean=True, check_extra_dir=None):
+    cur_branch = check_git_status(fatal_unless_clean)[0]
+    branch = re.sub(r'^patch/([^/]+)/[^/]+$', r'\1', cur_branch) # change patch/BRANCH/PATCH_NAME into BRANCH
+    if branch != master_branch:
+        print(f"The checkout is not on the {master_branch} branch.")
+        if master_branch != 'master':
+            sys.exit(1)
+        ans = input(f"Do you want me to continue with --branch={branch}? [n] ")
+        if not ans or not re.match(r'^y', ans, flags=re.I):
+            sys.exit(1)
+        master_branch = branch
+
+    if check_extra_dir and os.path.isdir(os.path.join(check_extra_dir, '.git')):
+        branch = check_git_status(fatal_unless_clean, check_extra_dir)[0]
+        if branch != master_branch:
+            print(f"The *{check_extra_dir}* checkout is on branch {branch}, not branch {master_branch}.")
+            ans = input(f"Do you want to change it to branch {master_branch}? [n] ")
+            if not ans or not re.match(r'^y', ans, flags=re.I):
+                sys.exit(1)
+            subdir.check_call(f"cd {check_extra_dir} && git checkout '{master_branch}'", shell=True)
+
+    return (cur_branch, master_branch)
+
+
+# Return the git hash of the most recent commit.
+def latest_git_hash(branch):
+    out = cmd_txt_chk(['git', 'log', '-1', '--no-color', branch])
+    m = re.search(r'^commit (\S+)', out, flags=re.M)
+    if not m:
+        die(f"Unable to determine commit hash for master branch: {branch}")
+    return m[1]
+
+
+# Return a set of all branch names that have the format "patch/BASE_BRANCH/NAME"
+# for the given base_branch string.  Just the NAME portion is put into the set.
+def get_patch_branches(base_branch):
+    branches = set()
+    proc = cmd_pipe('git branch -l'.split())
+    for line in proc.stdout:
+        m = re.search(r' patch/([^/]+)/(.+)', line)
+        if m and m[1] == base_branch:
+            branches.add(m[2])
+    proc.communicate()
+    return branches
+
+
+def mandate_gensend_hook():
+    hook = '.git/hooks/pre-push'
+    if not os.path.exists(hook):
+        print('Creating hook file:', hook)
+        cmd_chk(['./rsync', '-a', 'packaging/pre-push', hook])
+    else:
+        out, rc = cmd_txt_status(['fgrep', 'make gensend', hook], discard='output')
+        if rc:
+            die('Please add a "make gensend" into your', hook, 'script.')
+
+
+# Snag the GENFILES values out of the Makefile.in file and return them as a list.
+def get_gen_files(want_dir_plus_list=False):
+    cont_re = re.compile(r'\\\n')
+
+    gen_files = [ ]
+
+    auto_dir = os.path.join('auto-build-save', cmd_txt('git rev-parse --abbrev-ref HEAD').strip().replace('/', '%'))
+
+    with open('Makefile.in', 'r', encoding='utf-8') as fh:
+        for line in fh:
+            if not gen_files:
+                chk = re.sub(r'^GENFILES=', '', line)
+                if line == chk:
+                    continue
+                line = chk
+            m = re.search(r'\\$', line)
+            line = re.sub(r'^\s+|\s*\\\n?$|\s+$', '', line)
+            gen_files += line.split()
+            if not m:
+                break
+
+    if want_dir_plus_list:
+        return (auto_dir, gen_files)
+
+    return [ os.path.join(auto_dir, fn) for fn in gen_files ]
+
+
+def get_rsync_version():
+    with open('version.h', 'r', encoding='utf-8') as fh:
+        txt = fh.read()
+    m = re.match(r'^#define\s+RSYNC_VERSION\s+"(\d.+?)"', txt)
+    if m:
+        return m[1]
+    die("Unable to find RSYNC_VERSION define in version.h")
+
+
+def get_NEWS_version_info():
+    rel_re = re.compile(r'^\| \d{2} \w{3} \d{4}\s+\|\s+(?P<ver>\d+\.\d+\.\d+)\s+\|\s+(?P<pdate>\d{2} \w{3} \d{4}\s+)?\|\s+(?P<pver>\d+)\s+\|')
+    last_version = last_protocol_version = None
+    pdate = { }
+
+    with open('NEWS.md', 'r', encoding='utf-8') as fh:
+        for line in fh:
+            if not last_version: # Find the first non-dev|pre version with a release date.
+                m = re.search(r'rsync (\d+\.\d+\.\d+) .*\d\d\d\d', line)
+                if m:
+                    last_version = m[1]
+            m = rel_re.match(line)
+            if m:
+                if m['pdate']:
+                    pdate[m['ver']] = m['pdate']
+                if m['ver'] == last_version:
+                    last_protocol_version = m['pver']
+                    break
+
+    if not last_protocol_version:
+        die(f"Unable to determine protocol_version for {last_version}.")
+
+    return last_version, last_protocol_version
+
+
+def get_protocol_versions():
+    protocol_version = subprotocol_version = None
+
+    with open('rsync.h', 'r', encoding='utf-8') as fh:
+        for line in fh:
+            m = re.match(r'^#define\s+PROTOCOL_VERSION\s+(\d+)', line)
+            if m:
+                protocol_version = m[1]
+                continue
+            m = re.match(r'^#define\s+SUBPROTOCOL_VERSION\s+(\d+)', line)
+            if m:
+                subprotocol_version = m[1]
+                break
+
+    if not protocol_version:
+        die("Unable to determine the current PROTOCOL_VERSION.")
+
+    if not subprotocol_version:
+        die("Unable to determine the current SUBPROTOCOL_VERSION.")
+
+    return protocol_version, subprotocol_version
+
+# vim: sw=4 et

packaging/pre-push

@@ -0,0 +1,3 @@
+#!/bin/sh
+cat >/dev/null # Just discard stdin data
+make gensend

packaging/prep-auto-dir

@@ -0,0 +1,39 @@
+#!/bin/sh -e
+
+# This script will setup the build dir based on the current git branch and the
+# directory auto-build-save/$BRANCH.  We don't use a symlink for the build dir
+# because we want to maximize the ccache reuse, so all builds must happen in
+# the same real dir.  When a dir is moved out of auto-build-save/$BRANCH to the
+# build dir, it is replaced with a symlink so that it can still be found under
+# that dir.  The build dir also gets a .branch -> $BRANCH symlink so that we
+# can figure out the current build dir's branch.
+
+# To get started, just clone the rsync git repo and create the auto-build-save
+# dir.  If you have an existing git checkout and it is not in a pristine state,
+# run "make distclean" before creating the auto-build-save dir.
+
+auto_top='auto-build-save'
+if test -d $auto_top -a -d .git; then
+    desired_branch=`git rev-parse --abbrev-ref HEAD | tr / %`
+    auto_dir="$auto_top/$desired_branch"
+    if test -d build; then
+	cur_branch=`readlink build/.branch`
+    else
+	cur_branch='/'
+    fi
+    if test "$desired_branch" != "$cur_branch"; then
+	if test "$cur_branch" != /; then
+	    rm -f "$auto_top/$cur_branch"
+	    mv build "$auto_top/$cur_branch"
+	fi
+	test -d "$auto_dir" || mkdir "$auto_dir"
+	test -h "$auto_dir/.branch" || ln -s "$desired_branch" "$auto_dir/.branch"
+	mv "$auto_dir" build
+	ln -s ../build "$auto_dir"
+    fi
+    if test ! -h Makefile; then
+	rm -f Makefile
+	ln -s packaging/auto-Makefile Makefile
+    fi
+    echo $desired_branch
+fi