diff --git a/LICENSE-3RD-PARTY.txt b/LICENSE-3RD-PARTY.txt new file mode 100644 index 000000000..fb63dc59c --- /dev/null +++ b/LICENSE-3RD-PARTY.txt @@ -0,0 +1,1703 @@ +-------------------------------------------------------------------------------- + 3-Clause BSD License + applies to: + - MiniUPnPc + Copyright (c) 2005-2011, Thomas BERNARD + All rights reserved. + http://miniupnp.free.fr + - libwebm + Copyright (c) 2010, Google Inc. + All rights reserved. + https://chromium.googlesource.com/webm/libwebm/ + - libyuv + Copyright 2011 The LibYuv Project Authors. + All rights reserved. + https://chromium.googlesource.com/libyuv/libyuv/ + - libvpx + Copyright (c) 2010, The WebM Project authors. All rights reserved. + All rights reserved. + https://chromium.googlesource.com/webm/libvpx + - libvorbis + libvorbisenc + Copyright (c) 2002-2020 Xiph.org Foundation + https://github.com/xiph/vorbis +-------------------------------------------------------------------------------- + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: + + * Redistributions of source code must retain the above copyright notice, + this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above copyright notice, + this list of conditions and the following disclaimer in the documentation + and/or other materials provided with the distribution. + * The name of the author may not be used to endorse or promote products + derived from this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" +AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE +LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR +CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF +SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS +INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN +CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) +ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE +POSSIBILITY OF SUCH DAMAGE. + +-------------------------------------------------------------------------------- + curl License + applies to: + - curl + Copyright (c) 1996 - 2018, Daniel Stenberg, daniel@haxx.se, + and many contributors, see the THANKS file. + https://curl.haxx.se +-------------------------------------------------------------------------------- + +All rights reserved. + +Permission to use, copy, modify, and distribute this software for any purpose +with or without fee is hereby granted, provided that the above copyright notice +and this permission notice appear in all copies. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT OF THIRD PARTY RIGHTS. IN +NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, +DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR +OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE +OR OTHER DEALINGS IN THE SOFTWARE. + +Except as contained in this notice, the name of a copyright holder shall not be +used in advertising or otherwise to promote the sale, use or other dealings in +this Software without prior written authorization of the copyright holder. + +-------------------------------------------------------------------------------- + GCC Runtime Library Exception, Version 3.1 + applies to: + - GCC Runtime Library + Copyright (C) Free Software Foundation, Inc. + https://www.gnu.org/software/gcc/ +-------------------------------------------------------------------------------- + +GCC RUNTIME LIBRARY EXCEPTION Version 3.1, 31 March 2009 + +Copyright © 2009 Free Software Foundation, Inc. + +Everyone is permitted to copy and distribute verbatim copies of this license +document, but changing it is not allowed. + +This GCC Runtime Library Exception ("Exception") is an additional permission +under section 7 of the GNU General Public License, version 3 ("GPLv3"). It +applies to a given file (the "Runtime Library") that bears a notice placed by +the copyright holder of the file stating that the file is governed by GPLv3 +along with this Exception. + +When you use GCC to compile a program, GCC may combine portions of certain GCC +header files and runtime libraries with the compiled program. The purpose of +this Exception is to allow compilation of non-GPL (including proprietary) +programs to use, in this way, the header files and runtime libraries covered by +this Exception. + +0. Definitions. A file is an "Independent Module" if it either requires the +Runtime Library for execution after a Compilation Process, or makes use of an +interface provided by the Runtime Library, but is not otherwise based on the +Runtime Library. + +"GCC" means a version of the GNU Compiler Collection, with or without +modifications, governed by version 3 (or a specified later version) of the GNU +General Public License (GPL) with the option of using any subsequent versions +published by the FSF. + +"GPL-compatible Software" is software whose conditions of propagation, +modification and use would permit combination with GCC in accord with the +license of GCC. + +"Target Code" refers to output from any compiler for a real or virtual target +processor architecture, in executable form or suitable for input to an +assembler, loader, linker and/or execution phase. Notwithstanding that, Target +Code does not include data in any format that is used as a compiler +intermediate representation, or used for producing a compiler intermediate +representation. + +The "Compilation Process" transforms code entirely represented in +non-intermediate languages designed for human-written code, and/or in Java +Virtual Machine byte code, into Target Code. Thus, for example, use of source +code generators and preprocessors need not be considered part of the +Compilation Process, since the Compilation Process can be understood as +starting with the output of the generators or preprocessors. + +A Compilation Process is "Eligible" if it is done using GCC, alone or with +other GPL-compatible software, or if it is done without using any work based on +GCC. For example, using non-GPL-compatible Software to optimize any GCC +intermediate representations would not qualify as an Eligible Compilation +Process. + +1. Grant of Additional Permission. You have permission to propagate a work of +Target Code formed by combining the Runtime Library with Independent Modules, +even if such propagation would otherwise violate the terms of GPLv3, provided +that all Target Code was generated by Eligible Compilation Processes. You may +then convey such a combination under terms of your choice, consistent with the +licensing of the Independent Modules. + +2. No Weakening of GCC Copyleft. The availability of this Exception does not +imply any general presumption that third-party software is unaffected by the +copyleft requirements of the license of GCC. + +-------------------------------------------------------------------------------- + GNU General Public License, Version 3 + applies to: + - GCC Runtime Library + Copyright (C) Free Software Foundation, Inc. + https://www.gnu.org/software/gcc/ +-------------------------------------------------------------------------------- + + GNU GENERAL PUBLIC LICENSE + Version 3, 29 June 2007 + + Copyright (C) 2007 Free Software Foundation, Inc. + Everyone is permitted to copy and distribute verbatim copies + of this license document, but changing it is not allowed. + + Preamble + + 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 +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 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 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. + + 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. + + 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. + + 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. + + TERMS AND CONDITIONS + + 0. Definitions. + + "This License" refers to version 3 of the GNU General Public License. + + "Copyright" also means copyright-like laws that apply to other kinds of +works, such as semiconductor masks. + + "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. + + 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. + + A "covered work" means either the unmodified Program or a work based +on the Program. + + 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. + + 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. + + 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. + + 1. Source Code. + + 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. + + 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. + + 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 "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. + + The Corresponding Source need not include anything that users +can regenerate automatically from other parts of the Corresponding +Source. + + The Corresponding Source for a work in source code form is that +same work. + + 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. + + 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 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. + + 13. Use with the GNU Affero General Public License. + + 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. + + 14. Revised Versions of this License. + + 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 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. + + 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. + + 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. + + 15. Disclaimer of Warranty. + + 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. + + 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 +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 +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. + + + Copyright (C) + + 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, see . + +Also add information on how to contact you by electronic and paper mail. + + If the program does terminal interaction, make it output a short +notice like this when it starts in an interactive mode: + + Copyright (C) + 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, 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 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 +. + + 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 +. + +-------------------------------------------------------------------------------- + GNU Lesser General Public License, Version 2.1 + applies to: + - Game_Music_Emu + Shay Green + http://www.slack.net/~ant/ + + - libintl + Copyright (C) 1995-2018 Free Software Foundation, Inc. + https://www.gnu.org/software/gettext/ + + - mpg123 + Copyright (c) 1995-2013 by Michael Hipp and others, + free software under the terms of the LGPL v2.1 + https://www.mpg123.de + + - libxmp + https://github.com/libxmp/libxmp/blob/master/docs/CREDITS + https://github.com/libxmp/libxmp + + - ACSVM + https://github.com/DavidPH/ACSVM + +-------------------------------------------------------------------------------- + + GNU LESSER GENERAL PUBLIC LICENSE + Version 2.1, February 1999 + + Copyright (C) 1991, 1999 Free Software Foundation, Inc. + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + Everyone is permitted to copy and distribute verbatim copies + of this license document, but changing it is not allowed. + +[This is the first released version of the Lesser GPL. It also counts + as the successor of the GNU Library Public License, version 2, hence + the version number 2.1.] + + Preamble + + The licenses for most software are designed to take away your +freedom to share and change it. By contrast, the GNU General Public +Licenses are intended to guarantee your freedom to share and change +free software--to make sure the software is free for all its users. + + This license, the Lesser General Public License, applies to some +specially designated software packages--typically libraries--of the +Free Software Foundation and other authors who decide to use it. You +can use it too, but we suggest you first think carefully about whether +this license or the ordinary General Public License is the better +strategy to use in any particular case, based on the explanations below. + + When we speak of free software, we are referring to freedom of use, +not price. Our General Public Licenses are designed to make sure that +you have the freedom to distribute copies of free software (and charge +for this service if you wish); that you receive source code or can get +it if you want it; that you can change the software and use pieces of +it in new free programs; and that you are informed that you can do +these things. + + To protect your rights, we need to make restrictions that forbid +distributors to deny you these rights or to ask you to surrender these +rights. These restrictions translate to certain responsibilities for +you if you distribute copies of the library or if you modify it. + + For example, if you distribute copies of the library, whether gratis +or for a fee, you must give the recipients all the rights that we gave +you. You must make sure that they, too, receive or can get the source +code. If you link other code with the library, you must provide +complete object files to the recipients, so that they can relink them +with the library after making changes to the library and recompiling +it. And you must show them these terms so they know their rights. + + We protect your rights with a two-step method: (1) we copyright the +library, and (2) we offer you this license, which gives you legal +permission to copy, distribute and/or modify the library. + + To protect each distributor, we want to make it very clear that +there is no warranty for the free library. Also, if the library is +modified by someone else and passed on, the recipients should know +that what they have is not the original version, so that the original +author's reputation will not be affected by problems that might be +introduced by others. + + Finally, software patents pose a constant threat to the existence of +any free program. We wish to make sure that a company cannot +effectively restrict the users of a free program by obtaining a +restrictive license from a patent holder. Therefore, we insist that +any patent license obtained for a version of the library must be +consistent with the full freedom of use specified in this license. + + Most GNU software, including some libraries, is covered by the +ordinary GNU General Public License. This license, the GNU Lesser +General Public License, applies to certain designated libraries, and +is quite different from the ordinary General Public License. We use +this license for certain libraries in order to permit linking those +libraries into non-free programs. + + When a program is linked with a library, whether statically or using +a shared library, the combination of the two is legally speaking a +combined work, a derivative of the original library. The ordinary +General Public License therefore permits such linking only if the +entire combination fits its criteria of freedom. The Lesser General +Public License permits more lax criteria for linking other code with +the library. + + We call this license the "Lesser" General Public License because it +does Less to protect the user's freedom than the ordinary General +Public License. It also provides other free software developers Less +of an advantage over competing non-free programs. These disadvantages +are the reason we use the ordinary General Public License for many +libraries. However, the Lesser license provides advantages in certain +special circumstances. + + For example, on rare occasions, there may be a special need to +encourage the widest possible use of a certain library, so that it becomes +a de-facto standard. To achieve this, non-free programs must be +allowed to use the library. A more frequent case is that a free +library does the same job as widely used non-free libraries. In this +case, there is little to gain by limiting the free library to free +software only, so we use the Lesser General Public License. + + In other cases, permission to use a particular library in non-free +programs enables a greater number of people to use a large body of +free software. For example, permission to use the GNU C Library in +non-free programs enables many more people to use the whole GNU +operating system, as well as its variant, the GNU/Linux operating +system. + + Although the Lesser General Public License is Less protective of the +users' freedom, it does ensure that the user of a program that is +linked with the Library has the freedom and the wherewithal to run +that program using a modified version of the Library. + + The precise terms and conditions for copying, distribution and +modification follow. Pay close attention to the difference between a +"work based on the library" and a "work that uses the library". The +former contains code derived from the library, whereas the latter must +be combined with the library in order to run. + + GNU LESSER GENERAL PUBLIC LICENSE + TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION + + 0. This License Agreement applies to any software library or other +program which contains a notice placed by the copyright holder or +other authorized party saying it may be distributed under the terms of +this Lesser General Public License (also called "this License"). +Each licensee is addressed as "you". + + A "library" means a collection of software functions and/or data +prepared so as to be conveniently linked with application programs +(which use some of those functions and data) to form executables. + + The "Library", below, refers to any such software library or work +which has been distributed under these terms. A "work based on the +Library" means either the Library or any derivative work under +copyright law: that is to say, a work containing the Library or a +portion of it, either verbatim or with modifications and/or translated +straightforwardly into another language. (Hereinafter, translation is +included without limitation in the term "modification".) + + "Source code" for a work means the preferred form of the work for +making modifications to it. For a library, complete source code means +all the source code for all modules it contains, plus any associated +interface definition files, plus the scripts used to control compilation +and installation of the library. + + Activities other than copying, distribution and modification are not +covered by this License; they are outside its scope. The act of +running a program using the Library is not restricted, and output from +such a program is covered only if its contents constitute a work based +on the Library (independent of the use of the Library in a tool for +writing it). Whether that is true depends on what the Library does +and what the program that uses the Library does. + + 1. You may copy and distribute verbatim copies of the Library's +complete source code as you receive it, in any medium, provided that +you conspicuously and appropriately publish on each copy an +appropriate copyright notice and disclaimer of warranty; keep intact +all the notices that refer to this License and to the absence of any +warranty; and distribute a copy of this License along with the +Library. + + You may charge a fee for the physical act of transferring a copy, +and you may at your option offer warranty protection in exchange for a +fee. + + 2. You may modify your copy or copies of the Library or any portion +of it, thus forming a work based on the Library, and copy and +distribute such modifications or work under the terms of Section 1 +above, provided that you also meet all of these conditions: + + a) The modified work must itself be a software library. + + b) You must cause the files modified to carry prominent notices + stating that you changed the files and the date of any change. + + c) You must cause the whole of the work to be licensed at no + charge to all third parties under the terms of this License. + + d) If a facility in the modified Library refers to a function or a + table of data to be supplied by an application program that uses + the facility, other than as an argument passed when the facility + is invoked, then you must make a good faith effort to ensure that, + in the event an application does not supply such function or + table, the facility still operates, and performs whatever part of + its purpose remains meaningful. + + (For example, a function in a library to compute square roots has + a purpose that is entirely well-defined independent of the + application. Therefore, Subsection 2d requires that any + application-supplied function or table used by this function must + be optional: if the application does not supply it, the square + root function must still compute square roots.) + +These requirements apply to the modified work as a whole. If +identifiable sections of that work are not derived from the Library, +and can be reasonably considered independent and separate works in +themselves, then this License, and its terms, do not apply to those +sections when you distribute them as separate works. But when you +distribute the same sections as part of a whole which is a work based +on the Library, the distribution of the whole must be on the terms of +this License, whose permissions for other licensees extend to the +entire whole, and thus to each and every part regardless of who wrote +it. + +Thus, it is not the intent of this section to claim rights or contest +your rights to work written entirely by you; rather, the intent is to +exercise the right to control the distribution of derivative or +collective works based on the Library. + +In addition, mere aggregation of another work not based on the Library +with the Library (or with a work based on the Library) on a volume of +a storage or distribution medium does not bring the other work under +the scope of this License. + + 3. You may opt to apply the terms of the ordinary GNU General Public +License instead of this License to a given copy of the Library. To do +this, you must alter all the notices that refer to this License, so +that they refer to the ordinary GNU General Public License, version 2, +instead of to this License. (If a newer version than version 2 of the +ordinary GNU General Public License has appeared, then you can specify +that version instead if you wish.) Do not make any other change in +these notices. + + Once this change is made in a given copy, it is irreversible for +that copy, so the ordinary GNU General Public License applies to all +subsequent copies and derivative works made from that copy. + + This option is useful when you wish to copy part of the code of +the Library into a program that is not a library. + + 4. You may copy and distribute the Library (or a portion or +derivative of it, under Section 2) in object code or executable form +under the terms of Sections 1 and 2 above provided that you accompany +it with the complete corresponding machine-readable source code, which +must be distributed under the terms of Sections 1 and 2 above on a +medium customarily used for software interchange. + + If distribution of object code is made by offering access to copy +from a designated place, then offering equivalent access to copy the +source code from the same place satisfies the requirement to +distribute the source code, even though third parties are not +compelled to copy the source along with the object code. + + 5. A program that contains no derivative of any portion of the +Library, but is designed to work with the Library by being compiled or +linked with it, is called a "work that uses the Library". Such a +work, in isolation, is not a derivative work of the Library, and +therefore falls outside the scope of this License. + + However, linking a "work that uses the Library" with the Library +creates an executable that is a derivative of the Library (because it +contains portions of the Library), rather than a "work that uses the +library". The executable is therefore covered by this License. +Section 6 states terms for distribution of such executables. + + When a "work that uses the Library" uses material from a header file +that is part of the Library, the object code for the work may be a +derivative work of the Library even though the source code is not. +Whether this is true is especially significant if the work can be +linked without the Library, or if the work is itself a library. The +threshold for this to be true is not precisely defined by law. + + If such an object file uses only numerical parameters, data +structure layouts and accessors, and small macros and small inline +functions (ten lines or less in length), then the use of the object +file is unrestricted, regardless of whether it is legally a derivative +work. (Executables containing this object code plus portions of the +Library will still fall under Section 6.) + + Otherwise, if the work is a derivative of the Library, you may +distribute the object code for the work under the terms of Section 6. +Any executables containing that work also fall under Section 6, +whether or not they are linked directly with the Library itself. + + 6. As an exception to the Sections above, you may also combine or +link a "work that uses the Library" with the Library to produce a +work containing portions of the Library, and distribute that work +under terms of your choice, provided that the terms permit +modification of the work for the customer's own use and reverse +engineering for debugging such modifications. + + You must give prominent notice with each copy of the work that the +Library is used in it and that the Library and its use are covered by +this License. You must supply a copy of this License. If the work +during execution displays copyright notices, you must include the +copyright notice for the Library among them, as well as a reference +directing the user to the copy of this License. Also, you must do one +of these things: + + a) Accompany the work with the complete corresponding + machine-readable source code for the Library including whatever + changes were used in the work (which must be distributed under + Sections 1 and 2 above); and, if the work is an executable linked + with the Library, with the complete machine-readable "work that + uses the Library", as object code and/or source code, so that the + user can modify the Library and then relink to produce a modified + executable containing the modified Library. (It is understood + that the user who changes the contents of definitions files in the + Library will not necessarily be able to recompile the application + to use the modified definitions.) + + b) Use a suitable shared library mechanism for linking with the + Library. A suitable mechanism is one that (1) uses at run time a + copy of the library already present on the user's computer system, + rather than copying library functions into the executable, and (2) + will operate properly with a modified version of the library, if + the user installs one, as long as the modified version is + interface-compatible with the version that the work was made with. + + c) Accompany the work with a written offer, valid for at + least three years, to give the same user the materials + specified in Subsection 6a, above, for a charge no more + than the cost of performing this distribution. + + d) If distribution of the work is made by offering access to copy + from a designated place, offer equivalent access to copy the above + specified materials from the same place. + + e) Verify that the user has already received a copy of these + materials or that you have already sent this user a copy. + + For an executable, the required form of the "work that uses the +Library" must include any data and utility programs needed for +reproducing the executable from it. However, as a special exception, +the materials to be distributed need not include anything that is +normally distributed (in either source or binary form) with the major +components (compiler, kernel, and so on) of the operating system on +which the executable runs, unless that component itself accompanies +the executable. + + It may happen that this requirement contradicts the license +restrictions of other proprietary libraries that do not normally +accompany the operating system. Such a contradiction means you cannot +use both them and the Library together in an executable that you +distribute. + + 7. You may place library facilities that are a work based on the +Library side-by-side in a single library together with other library +facilities not covered by this License, and distribute such a combined +library, provided that the separate distribution of the work based on +the Library and of the other library facilities is otherwise +permitted, and provided that you do these two things: + + a) Accompany the combined library with a copy of the same work + based on the Library, uncombined with any other library + facilities. This must be distributed under the terms of the + Sections above. + + b) Give prominent notice with the combined library of the fact + that part of it is a work based on the Library, and explaining + where to find the accompanying uncombined form of the same work. + + 8. You may not copy, modify, sublicense, link with, or distribute +the Library except as expressly provided under this License. Any +attempt otherwise to copy, modify, sublicense, link with, or +distribute the Library is void, and will automatically terminate your +rights under this License. However, parties who have received copies, +or rights, from you under this License will not have their licenses +terminated so long as such parties remain in full compliance. + + 9. You are not required to accept this License, since you have not +signed it. However, nothing else grants you permission to modify or +distribute the Library or its derivative works. These actions are +prohibited by law if you do not accept this License. Therefore, by +modifying or distributing the Library (or any work based on the +Library), you indicate your acceptance of this License to do so, and +all its terms and conditions for copying, distributing or modifying +the Library or works based on it. + + 10. Each time you redistribute the Library (or any work based on the +Library), the recipient automatically receives a license from the +original licensor to copy, distribute, link with or modify the Library +subject to these terms and conditions. You may not impose any further +restrictions on the recipients' exercise of the rights granted herein. +You are not responsible for enforcing compliance by third parties with +this License. + + 11. If, as a consequence of a court judgment or allegation of patent +infringement or for any other reason (not limited to patent issues), +conditions are imposed on you (whether by court order, agreement or +otherwise) that contradict the conditions of this License, they do not +excuse you from the conditions of this License. If you cannot +distribute so as to satisfy simultaneously your obligations under this +License and any other pertinent obligations, then as a consequence you +may not distribute the Library at all. For example, if a patent +license would not permit royalty-free redistribution of the Library by +all those who receive copies directly or indirectly through you, then +the only way you could satisfy both it and this License would be to +refrain entirely from distribution of the Library. + +If any portion of this section is held invalid or unenforceable under any +particular circumstance, the balance of the section is intended to apply, +and the section as a whole is intended to apply in other circumstances. + +It is not the purpose of this section to induce you to infringe any +patents or other property right claims or to contest validity of any +such claims; this section has the sole purpose of protecting the +integrity of the free software distribution system which is +implemented by public license practices. Many people have made +generous contributions to the wide range of software distributed +through that system in reliance on consistent application of that +system; it is up to the author/donor to decide if he or she is willing +to distribute software through any other system and a licensee cannot +impose that choice. + +This section is intended to make thoroughly clear what is believed to +be a consequence of the rest of this License. + + 12. If the distribution and/or use of the Library is restricted in +certain countries either by patents or by copyrighted interfaces, the +original copyright holder who places the Library under this License may add +an explicit geographical distribution limitation excluding those countries, +so that distribution is permitted only in or among countries not thus +excluded. In such case, this License incorporates the limitation as if +written in the body of this License. + + 13. The Free Software Foundation may publish revised and/or new +versions of the Lesser General Public License from time to time. +Such new versions will be similar in spirit to the present version, +but may differ in detail to address new problems or concerns. + +Each version is given a distinguishing version number. If the Library +specifies a version number of this License which applies to it and +"any later version", you have the option of following the terms and +conditions either of that version or of any later version published by +the Free Software Foundation. If the Library does not specify a +license version number, you may choose any version ever published by +the Free Software Foundation. + + 14. If you wish to incorporate parts of the Library into other free +programs whose distribution conditions are incompatible with these, +write to the author to ask for permission. For software which is +copyrighted by the Free Software Foundation, write to the Free +Software Foundation; we sometimes make exceptions for this. Our +decision will be guided by the two goals of preserving the free status +of all derivatives of our free software and of promoting the sharing +and reuse of software generally. + + NO WARRANTY + + 15. BECAUSE THE LIBRARY IS LICENSED FREE OF CHARGE, THERE IS NO +WARRANTY FOR THE LIBRARY, TO THE EXTENT PERMITTED BY APPLICABLE LAW. +EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR +OTHER PARTIES PROVIDE THE LIBRARY "AS IS" WITHOUT WARRANTY OF ANY +KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE +IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR +PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE +LIBRARY IS WITH YOU. SHOULD THE LIBRARY PROVE DEFECTIVE, YOU ASSUME +THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION. + + 16. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN +WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY +AND/OR REDISTRIBUTE THE LIBRARY AS PERMITTED ABOVE, BE LIABLE TO YOU +FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR +CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE +LIBRARY (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING +RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A +FAILURE OF THE LIBRARY TO OPERATE WITH ANY OTHER SOFTWARE), EVEN IF +SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH +DAMAGES. + + END OF TERMS AND CONDITIONS + + How to Apply These Terms to Your New Libraries + + If you develop a new library, and you want it to be of the greatest +possible use to the public, we recommend making it free software that +everyone can redistribute and change. You can do so by permitting +redistribution under these terms (or, alternatively, under the terms of the +ordinary General Public License). + + To apply these terms, attach the following notices to the library. It is +safest to attach them to the start of each source file to most effectively +convey the exclusion of warranty; and each file should have at least the +"copyright" line and a pointer to where the full notice is found. + + + Copyright (C) + + This library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with this library; if not, write to the Free Software + Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + +Also add information on how to contact you by electronic and paper mail. + +You should also get your employer (if you work as a programmer) or your +school, if any, to sign a "copyright disclaimer" for the library, if +necessary. Here is a sample; alter the names: + + Yoyodyne, Inc., hereby disclaims all copyright interest in the + library `Frob' (a library for tweaking knobs) written by James Random Hacker. + + , 1 April 1990 + Ty Coon, President of Vice + +That's all there is to it! + +-------------------------------------------------------------------------------- + libpng License + applies to: + - libpng + Copyright (c) 2000-2002, 2004, 2006-2018 Glenn Randers-Pehrson + http://www.libpng.org/pub/png/libpng.html +-------------------------------------------------------------------------------- + +This copy of the libpng notices is provided for your convenience. In case of +any discrepancy between this copy and the notices in the file png.h that is +included in the libpng distribution, the latter shall prevail. + +COPYRIGHT NOTICE, DISCLAIMER, and LICENSE: + +If you modify libpng you may insert additional notices immediately following +this sentence. + +This code is released under the libpng license. + +libpng versions 1.0.7, July 1, 2000 through 1.6.35, July 15, 2018 are +Copyright (c) 2000-2002, 2004, 2006-2018 Glenn Randers-Pehrson, are +derived from libpng-1.0.6, and are distributed according to the same +disclaimer and license as libpng-1.0.6 with the following individuals +added to the list of Contributing Authors: + + Simon-Pierre Cadieux + Eric S. Raymond + Mans Rullgard + Cosmin Truta + Gilles Vollant + James Yu + Mandar Sahastrabuddhe + Google Inc. + Vadim Barkov + +and with the following additions to the disclaimer: + + There is no warranty against interference with your enjoyment of the + library or against infringement. There is no warranty that our + efforts or the library will fulfill any of your particular purposes + or needs. This library is provided with all faults, and the entire + risk of satisfactory quality, performance, accuracy, and effort is with + the user. + +Some files in the "contrib" directory and some configure-generated +files that are distributed with libpng have other copyright owners and +are released under other open source licenses. + +libpng versions 0.97, January 1998, through 1.0.6, March 20, 2000, are +Copyright (c) 1998-2000 Glenn Randers-Pehrson, are derived from +libpng-0.96, and are distributed according to the same disclaimer and +license as libpng-0.96, with the following individuals added to the list +of Contributing Authors: + + Tom Lane + Glenn Randers-Pehrson + Willem van Schaik + +libpng versions 0.89, June 1996, through 0.96, May 1997, are +Copyright (c) 1996-1997 Andreas Dilger, are derived from libpng-0.88, +and are distributed according to the same disclaimer and license as +libpng-0.88, with the following individuals added to the list of +Contributing Authors: + + John Bowler + Kevin Bracey + Sam Bushell + Magnus Holmgren + Greg Roelofs + Tom Tanner + +Some files in the "scripts" directory have other copyright owners +but are released under this license. + +libpng versions 0.5, May 1995, through 0.88, January 1996, are +Copyright (c) 1995-1996 Guy Eric Schalnat, Group 42, Inc. + +For the purposes of this copyright and license, "Contributing Authors" +is defined as the following set of individuals: + + Andreas Dilger + Dave Martindale + Guy Eric Schalnat + Paul Schmidt + Tim Wegner + +The PNG Reference Library is supplied "AS IS". The Contributing Authors +and Group 42, Inc. disclaim all warranties, expressed or implied, +including, without limitation, the warranties of merchantability and of +fitness for any purpose. The Contributing Authors and Group 42, Inc. +assume no liability for direct, indirect, incidental, special, exemplary, +or consequential damages, which may result from the use of the PNG +Reference Library, even if advised of the possibility of such damage. + +Permission is hereby granted to use, copy, modify, and distribute this +source code, or portions hereof, for any purpose, without fee, subject +to the following restrictions: + + 1. The origin of this source code must not be misrepresented. + + 2. Altered versions must be plainly marked as such and must not + be misrepresented as being the original source. + + 3. This Copyright notice may not be removed or altered from any + source or altered source distribution. + +The Contributing Authors and Group 42, Inc. specifically permit, without +fee, and encourage the use of this source code as a component to +supporting the PNG file format in commercial products. If you use this +source code in a product, acknowledgment is not required but would be +appreciated. + +END OF COPYRIGHT NOTICE, DISCLAIMER, and LICENSE. + +TRADEMARK: + +The name "libpng" has not been registered by the Copyright owner +as a trademark in any jurisdiction. However, because libpng has +been distributed and maintained world-wide, continually since 1995, +the Copyright owner claims "common-law trademark protection" in any +jurisdiction where common-law trademark is recognized. + +OSI CERTIFICATION: + +Libpng is OSI Certified Open Source Software. OSI Certified Open Source is +a certification mark of the Open Source Initiative. OSI has not addressed +the additional disclaimers inserted at version 1.0.7. + +EXPORT CONTROL: + +The Copyright owner believes that the Export Control Classification +Number (ECCN) for libpng is EAR99, which means not subject to export +controls or International Traffic in Arms Regulations (ITAR) because +it is open source, publicly available software, that does not contain +any encryption software. See the EAR, paragraphs 734.3(b)(3) and +734.7(b). + +Glenn Randers-Pehrson +glennrp at users.sourceforge.net +July 15, 2018 + +-------------------------------------------------------------------------------- + New BSD License + applies to: + - FLAC + Copyright (C) 2000-2009 Josh Coalson + Copyright (C) 2011-2016 Xiph.Org Foundation + https://xiph.org/flac/api/ + + - Vorbis + Copyright (c) 2002-2008 Xiph.org Foundation + https://xiph.org/vorbis/ + + - Opus + Copyright 2001-2011 Xiph.Org, Skype Limited, Octasic, + Jean-Marc Valin, Timothy B. Terriberry, + CSIRO, Gregory Maxwell, Mark Borgerding, + Erik de Castro Lopo + https://opus-codec.org + + - Opus File + Copyright (c) 1994-2013 Xiph.Org Foundation and contributors + https://opus-codec.org +-------------------------------------------------------------------------------- + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions +are met: + +- Redistributions of source code must retain the above copyright +notice, this list of conditions and the following disclaimer. + +- Redistributions in binary form must reproduce the above copyright +notice, this list of conditions and the following disclaimer in the +documentation and/or other materials provided with the distribution. + +[ For FLAC, Vorbis, and Opus File +- Neither the name of the Xiph.org Foundation nor the names of its +contributors may be used to endorse or promote products derived from +this software without specific prior written permission. +] + +[ For Opus +- Neither the name of Internet Society, IETF or IETF Trust, nor the +names of specific contributors, may be used to endorse or promote +products derived from this software without specific prior written +permission. +] + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR +CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, +EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, +PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR +PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF +LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING +NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS +SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +-------------------------------------------------------------------------------- + Public Domain + applies to: + - win_iconv + Yukihiro Nakadaira + win_iconv is placed in the public domain. + https://github.com/win-iconv/win-iconv + + - libmodplug + ModPlug-XMMS and libmodplug are now in the public domain. + http://modplug-xmms.sourceforge.net +-------------------------------------------------------------------------------- + +-------------------------------------------------------------------------------- + zlib License + applies to: + - Simple DirectMedia Layer + Copyright (C) 1997-2018 Sam Lantinga + https://www.libsdl.org/hg.php + + - SDL_mixer: An audio mixer library based on the SDL library + Copyright (C) 1997-2018 Sam Lantinga + https://www.libsdl.org/projects/SDL_mixer/ + + - zlib + Copyright (C) 1995-2017 Jean-loup Gailly and Mark Adler + https://zlib.net +-------------------------------------------------------------------------------- + +This software is provided 'as-is', without any express or implied +warranty. In no event will the authors be held liable for any damages +arising from the use of this software. + +Permission is granted to anyone to use this software for any purpose, +including commercial applications, and to alter it and redistribute it +freely, subject to the following restrictions: + +1. The origin of this software must not be misrepresented; you must not + claim that you wrote the original software. If you use this software + in a product, an acknowledgment in the product documentation would be + appreciated but is not required. +2. Altered source versions must be plainly marked as such, and must not be + misrepresented as being the original software. +3. This notice may not be removed or altered from any source distribution. + +-------------------------------------------------------------------------------- + 2-Clause BSD License + applies to: + - monocypher + Copyright (c) 2017-2020, Loup Vaillant + All rights reserved. + https://monocypher.org/ +-------------------------------------------------------------------------------- + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + +1. Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. +2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in the + documentation and/or other materials provided with the + distribution. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +-------------------------------------------------------------------------------- + MIT License + applies to: + - fmt + Copyright (c) 2012 - present, Victor Zverovich and {fmt} contributors + https://github.com/fmtlib/fmt + - imgui + Copyright (c) 2014-2023 Omar Cornut + https://github.com/ocornut/imgui + - stb_vorbis + stb_rect_pack + Copyright (c) 2017 Sean Barrett + https://github.com/nothings/stb +-------------------------------------------------------------------------------- + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. + +-------------------------------------------------------------------------------- + Boost Software License - Version 1.0 + applies to: + - tcbrindle span + https://www.boost.org/ +-------------------------------------------------------------------------------- + +Boost Software License - Version 1.0 - August 17th, 2003 + +Permission is hereby granted, free of charge, to any person or organization +obtaining a copy of the software and accompanying documentation covered by +this license (the "Software") to use, reproduce, display, distribute, +execute, and transmit the Software, and to prepare derivative works of the +Software, and to permit third-parties to whom the Software is furnished to +do so, all subject to the following: + +The copyright notices in the Software and this entire statement, including +the above license grant, this restriction and the following disclaimer, +must be included in all copies of the Software, in whole or in part, and +all derivative works of the Software, unless such copies or derivative +works are solely in the form of machine-executable object code generated by +a source language processor. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE, TITLE AND NON-INFRINGEMENT. IN NO EVENT +SHALL THE COPYRIGHT HOLDERS OR ANYONE DISTRIBUTING THE SOFTWARE BE LIABLE +FOR ANY DAMAGES OR OTHER LIABILITY, WHETHER IN CONTRACT, TORT OR OTHERWISE, +ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER +DEALINGS IN THE SOFTWARE. + +------------------------------------------------------------------------------- + +In addition to the above, we make use of the following public-domain software: + +* glad - https://github.com/Dav1dde/glad \ No newline at end of file diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt index 042ccda03..8fd0c12ac 100644 --- a/src/CMakeLists.txt +++ b/src/CMakeLists.txt @@ -559,6 +559,7 @@ add_subdirectory(sdl) add_subdirectory(objects) add_subdirectory(acs) add_subdirectory(rhi) +add_subdirectory(monocypher) if(SRB2_CONFIG_ENABLE_TESTS) add_subdirectory(tests) endif() diff --git a/src/d_clisrv.c b/src/d_clisrv.c index 298c6a5bc..6e1066db5 100644 --- a/src/d_clisrv.c +++ b/src/d_clisrv.c @@ -47,6 +47,8 @@ #include "lua_hook.h" #include "md5.h" #include "m_perfstats.h" +#include "monocypher/monocypher.h" +#include "stun.h" // SRB2Kart #include "k_kart.h" @@ -121,6 +123,7 @@ SINT8 nodetoplayer3[MAXNETNODES]; // say the numplayer for this node if any (spl SINT8 nodetoplayer4[MAXNETNODES]; // say the numplayer for this node if any (splitscreen == 3) UINT8 playerpernode[MAXNETNODES]; // used specialy for splitscreen boolean nodeingame[MAXNETNODES]; // set false as nodes leave game +boolean nodeneedsauth[MAXNETNODES]; tic_t servermaxping = 20; // server's max delay, in frames. Defaults to 20 static tic_t nettics[MAXNETNODES]; // what tic the client have received @@ -156,9 +159,34 @@ char connectedservername[MAXSERVERNAME]; /// \todo WORK! boolean acceptnewnode = true; +UINT32 ourIP; // Used when populating PT_SERVERCHALLENGE (guards against signature reuse) +uint8_t lastReceivedKey[MAXNETNODES][MAXSPLITSCREENPLAYERS][PUBKEYLENGTH]; // Player's public key (join process only! active players have it on player_t) +uint8_t lastSentChallenge[MAXNETNODES][CHALLENGELENGTH]; // The random message we asked them to sign in PT_SERVERCHALLENGE, check it in PT_CLIENTJOIN +uint8_t awaitingChallenge[CHALLENGELENGTH]; // The message the server asked our client to sign when joining +uint8_t lastChallengeAll[CHALLENGELENGTH]; // The message we asked EVERYONE to sign for client-to-client identity proofs +uint8_t lastReceivedSignature[MAXPLAYERS][SIGNATURELENGTH]; // Everyone's response to lastChallengeAll +uint8_t knownWhenChallenged[MAXPLAYERS][PUBKEYLENGTH]; // Everyone a client saw at the moment a challenge should be initiated +boolean expectChallenge = false; // Were we in-game before a client-to-client challenge should have been sent? + +uint8_t priorKeys[MAXPLAYERS][PUBKEYLENGTH]; // Make a note of keys before consuming a new gamestate, and if the server tries to send us a gamestate where keys differ, assume shenanigans + boolean serverisfull = false; //lets us be aware if the server was full after we check files, but before downloading, so we can ask if the user still wants to download or not tic_t firstconnectattempttime = 0; +consvar_t cv_allowguests = CVAR_INIT ("allowguests", "On", CV_SAVE, CV_OnOff, NULL); + +#ifdef DEVELOP + consvar_t cv_badjoin = CVAR_INIT ("badjoin", "0", 0, CV_Unsigned, NULL); + consvar_t cv_badtraffic = CVAR_INIT ("badtraffic", "0", 0, CV_Unsigned, NULL); + consvar_t cv_badresponse = CVAR_INIT ("badresponse", "0", 0, CV_Unsigned, NULL); + consvar_t cv_noresponse = CVAR_INIT ("noresponse", "0", 0, CV_Unsigned, NULL); + consvar_t cv_nochallenge = CVAR_INIT ("nochallenge", "0", 0, CV_Unsigned, NULL); + consvar_t cv_badresults = CVAR_INIT ("badresults", "0", 0, CV_Unsigned, NULL); + consvar_t cv_noresults = CVAR_INIT ("noresults", "0", 0, CV_Unsigned, NULL); + consvar_t cv_badtime = CVAR_INIT ("badtime", "0", 0, CV_Unsigned, NULL); + consvar_t cv_badip = CVAR_INIT ("badip", "0", 0, CV_Unsigned, NULL); +#endif + // engine // Must be a power of two @@ -191,6 +219,60 @@ consvar_t cv_httpsource = CVAR_INIT ("http_source", "", CV_SAVE, NULL, NULL); consvar_t cv_kicktime = CVAR_INIT ("kicktime", "10", CV_SAVE, CV_Unsigned, NULL); +// Generate a message for an authenticating client to sign, with some guarantees about who we are. +void GenerateChallenge(uint8_t *buf) +{ + #ifndef SRB2_LITTLE_ENDIAN + #error "FIXME: 64-bit timestamp field is not supported on Big Endian" + #endif + + UINT64 now = time(NULL); + csprng(buf, CHALLENGELENGTH); // Random noise as a baseline, but... + memcpy(buf, &now, sizeof(now)); // Timestamp limits the reuse window. + memcpy(buf + sizeof(now), &ourIP, sizeof(ourIP)); // IP prevents captured signatures from being used elsewhere. + + #ifdef DEVELOP + if (cv_badtime.value) + { + CV_AddValue(&cv_badtime, -1); + CONS_Alert(CONS_WARNING, "cv_badtime enabled, trashing time in auth message\n"); + memset(buf, 0, sizeof(now)); + } + + if (cv_badip.value) + { + CV_AddValue(&cv_badip, -1); + CONS_Alert(CONS_WARNING, "cv_badip enabled, trashing IP in auth message\n"); + memset(buf + sizeof(now), 0, sizeof(ourIP)); + } + #endif +} + +// Modified servers can throw softballs or reuse challenges. +// Don't sign anything that wasn't generated just for us! +shouldsign_t ShouldSignChallenge(uint8_t *message) +{ + #ifndef SRB2_LITTLE_ENDIAN + #error "FIXME: 64-bit timestamp field is not supported on Big Endian" + #endif + + UINT64 then, now; + UINT32 claimedIP, realIP; + + now = time(NULL); + memcpy(&then, message, sizeof(then)); + memcpy(&claimedIP, message + sizeof(then), sizeof(claimedIP)); + realIP = I_GetNodeAddressInt(servernode); + + if ((max(now, then) - min(now, then)) > 60*5) + return SIGN_BADTIME; + + if (realIP != claimedIP && I_IsExternalAddress(&realIP)) + return SIGN_BADIP; + + return SIGN_OK; +} + static inline void *G_DcpyTiccmd(void* dest, const ticcmd_t* src, const size_t n) { const size_t d = n / sizeof(ticcmd_t); @@ -553,6 +635,8 @@ typedef enum CL_PREPAREHTTPFILES, CL_DOWNLOADHTTPFILES, #endif + CL_SENDKEY, + CL_WAITCHALLENGE, } cl_mode_t; static void GetPackets(void); @@ -834,9 +918,80 @@ static boolean CL_SendJoin(void) memcpy(&netbuffer->u.clientcfg.availabilities, R_GetSkinAvailabilities(false, false), MAXAVAILABILITY*sizeof(UINT8)); + // Don't leak old signatures from prior sessions. + memset(&netbuffer->u.clientcfg.challengeResponse, 0, sizeof(((clientconfig_pak *)0)->challengeResponse)); + + if (client && netgame) + { + shouldsign_t safe = ShouldSignChallenge(awaitingChallenge); + + if (safe != SIGN_OK) + { + if (safe == SIGN_BADIP) + { + I_Error("External server IP didn't match the message it sent."); + } + else if (safe == SIGN_BADTIME) + { + I_Error("External server sent a message with an unusual timestamp.\nCheck your clocks!"); + } + else + { + I_Error("External server asked for a signature on something strange.\nPlease notify a developer if you've seen this more than once."); + } + return false; + } + } + + for (i = 0; i <= splitscreen; i++) + { + uint8_t signature[SIGNATURELENGTH]; + profile_t *localProfile = PR_GetLocalPlayerProfile(i); + + if (PR_IsLocalPlayerGuest(i)) // GUESTS don't have keys + { + memset(signature, 0, sizeof(signature)); + } + else + { + // If our keys are garbage (corrupted profile?), fail here instead of when the server boots us, so the player knows what's going on. + crypto_eddsa_sign(signature, localProfile->secret_key, awaitingChallenge, sizeof(awaitingChallenge)); + if (crypto_eddsa_check(signature, localProfile->public_key, awaitingChallenge, sizeof(awaitingChallenge)) != 0) + I_Error("Couldn't self-verify key associated with player %d, profile %d.\nProfile data may be corrupted.", i, cv_lastprofile[i].value); // I guess this is the most reasonable way to catch a malformed key. + } + + #ifdef DEVELOP + if (cv_badjoin.value) + { + CV_AddValue(&cv_badjoin, -1); + CONS_Alert(CONS_WARNING, "cv_badjoin enabled, scrubbing signature from CL_SendJoin\n"); + memset(signature, 0, sizeof(signature)); + } + #endif + + // Testing + // memset(signature, 0, sizeof(signature)); + + memcpy(&netbuffer->u.clientcfg.challengeResponse[i], signature, sizeof(signature)); + } + return HSendPacket(servernode, false, 0, sizeof (clientconfig_pak)); } +static boolean CL_SendKey(void) +{ + int i; + netbuffer->packettype = PT_CLIENTKEY; + + memset(netbuffer->u.clientkey.key, 0, sizeof(((clientkey_pak *)0)->key)); + for (i = 0; i <= splitscreen; i++) + { + // GUEST profiles have all-zero keys. This will be handled at the end of the challenge process, don't worry about it. + memcpy(netbuffer->u.clientkey.key[i], PR_GetProfile(cv_lastprofile[i].value)->public_key, PUBKEYLENGTH); + } + return HSendPacket(servernode, false, 0, sizeof (clientkey_pak) ); +} + static void CopyCaretColors (char *p, const char *s, int n) { @@ -1320,6 +1475,19 @@ static void CL_LoadReceivedSavegame(boolean reloading) // so they know they can resume the game netbuffer->packettype = PT_RECEIVEDGAMESTATE; HSendPacket(servernode, true, 0, 0); + + if (reloading) + { + int i; + for (i = 0; i < MAXPLAYERS; i++) + { + if (memcmp(priorKeys[i], players[i].public_key, sizeof(priorKeys[i])) != 0) + { + HandleSigfail("Gamestate reload contained new keys"); + break; + } + } + } } static void CL_ReloadReceivedSavegame(void) @@ -1896,7 +2064,7 @@ static boolean CL_ServerConnectionTicker(const char *tmpsave, tic_t *oldtic, tic { *asksent = 0; //This ensure the first join ask is right away firstconnectattempttime = I_GetTime(); - cl_mode = CL_ASKJOIN; + cl_mode = CL_SENDKEY; } break; case CL_ASKJOIN: @@ -1932,6 +2100,19 @@ static boolean CL_ServerConnectionTicker(const char *tmpsave, tic_t *oldtic, tic cl_mode = CL_ASKJOIN; } break; + case CL_SENDKEY: + if (I_GetTime() >= *asksent && CL_SendKey()) + { + *asksent = I_GetTime() + NEWTICRATE*3; + cl_mode = CL_WAITCHALLENGE; + } + break; + case CL_WAITCHALLENGE: + if (I_GetTime() >= *asksent) + { + cl_mode = CL_SENDKEY; + } + break; case CL_DOWNLOADSAVEGAME: // At this state, the first (and only) needed file is the gamestate if (fileneeded[0].status == FS_FOUND) @@ -2683,6 +2864,8 @@ void CL_Reset(void) serverisfull = false; connectiontimeout = (tic_t)cv_nettimeout.value; //reset this temporary hack + expectChallenge = false; + #ifdef HAVE_CURL curl_failedwebdownload = false; curl_transfers = 0; @@ -2766,6 +2949,8 @@ static void Command_Nodes(void) CONS_Printf(" - %s", address); } + CONS_Printf(" [RRID-%s] ", GetPrettyRRID(players[i].public_key, true)); + if (IsPlayerAdmin(i)) CONS_Printf(M_GetText(" (verified admin)")); @@ -3121,6 +3306,10 @@ static void Got_KickCmd(UINT8 **p, INT32 playernum) HU_AddChatText(va("\x82*%s left the game (Connection timeout)", player_names[pnum]), false); kickreason = KR_TIMEOUT; break; + case KICK_MSG_SIGFAIL: + HU_AddChatText(va("\x82*%s left the game (Invalid signature)", player_names[pnum]), false); + kickreason = KR_TIMEOUT; + break; case KICK_MSG_PLAYER_QUIT: if (netgame) // not splitscreen/bots HU_AddChatText(va("\x82*%s left the game", player_names[pnum]), false); @@ -3177,6 +3366,8 @@ static void Got_KickCmd(UINT8 **p, INT32 playernum) M_StartMessage(va(M_GetText("You have been kicked\n(%s)\nPress (B)\n"), reason), NULL, MM_NOTHING); else if (msg == KICK_MSG_CUSTOM_BAN) M_StartMessage(va(M_GetText("You have been banned\n(%s)\nPress (B)\n"), reason), NULL, MM_NOTHING); + else if (msg == KICK_MSG_SIGFAIL) + M_StartMessage(M_GetText("Server closed connection\n(Invalid signature)\nPress (B)\n"), NULL, MM_NOTHING); else M_StartMessage(M_GetText("You have been kicked by the server\n\nPress (B)\n"), NULL, MM_NOTHING); } @@ -3390,9 +3581,7 @@ void D_ClientServerInit(void) COM_AddCommand("drop", Command_Drop); COM_AddCommand("droprate", Command_Droprate); #endif -#ifdef _DEBUG COM_AddCommand("numnodes", Command_Numnodes); -#endif RegisterNetXCmd(XD_KICK, Got_KickCmd); RegisterNetXCmd(XD_ADDPLAYER, Got_AddPlayer); @@ -3417,6 +3606,7 @@ static void ResetNode(INT32 node) { nodeingame[node] = false; nodewaiting[node] = 0; + nodeneedsauth[node] = false; nettics[node] = gametic; supposedtics[node] = gametic; @@ -3497,6 +3687,8 @@ void SV_ResetServer(void) for (i = 0; i < MAXUNLOCKABLES; i++) netUnlocked[i] = (dedicated || gamedata->unlocked[i]); + expectChallenge = false; + DEBFILE("\n-=-=-=-=-=-=-= Server Reset =-=-=-=-=-=-=-\n\n"); } @@ -3581,6 +3773,8 @@ static inline void SV_AddNode(INT32 node) // nodeingame when connected not here if (node) nodeingame[node] = true; + + nodeneedsauth[node] = false; } // Xcmd XD_ADDPLAYER @@ -3622,6 +3816,7 @@ static void Got_AddPlayer(UINT8 **p, INT32 playernum) newplayer->jointime = 0; READSTRINGN(*p, player_names[newplayernum], MAXPLAYERNAME); + READMEM(*p, players[newplayernum].public_key, PUBKEYLENGTH); console = READUINT8(*p); splitscreenplayer = READUINT8(*p); @@ -3781,10 +3976,12 @@ static void Got_AddBot(UINT8 **p, INT32 playernum) LUA_HookInt(newplayernum, HOOK(PlayerJoin)); } -static boolean SV_AddWaitingPlayers(SINT8 node, UINT8 *availabilities, const char *name, const char *name2, const char *name3, const char *name4) +static boolean SV_AddWaitingPlayers(SINT8 node, UINT8 *availabilities, +const char *name, uint8_t *key, const char *name2, uint8_t *key2, +const char *name3, uint8_t *key3, const char *name4, uint8_t *key4) { INT32 n, newplayernum, i; - UINT8 buf[4 + MAXPLAYERNAME + MAXAVAILABILITY]; + UINT8 buf[4 + MAXPLAYERNAME + PUBKEYLENGTH + MAXAVAILABILITY]; UINT8 *buf_p = buf; boolean newplayer = false; @@ -3847,21 +4044,25 @@ static boolean SV_AddWaitingPlayers(SINT8 node, UINT8 *availabilities, const cha { nodetoplayer[node] = newplayernum; WRITESTRINGN(buf_p, name, MAXPLAYERNAME); + WRITEMEM(buf_p, key, PUBKEYLENGTH); } else if (playerpernode[node] < 2) { nodetoplayer2[node] = newplayernum; WRITESTRINGN(buf_p, name2, MAXPLAYERNAME); + WRITEMEM(buf_p, key2, PUBKEYLENGTH); } else if (playerpernode[node] < 3) { nodetoplayer3[node] = newplayernum; WRITESTRINGN(buf_p, name3, MAXPLAYERNAME); + WRITEMEM(buf_p, key3, PUBKEYLENGTH); } else if (playerpernode[node] < 4) { nodetoplayer4[node] = newplayernum; WRITESTRINGN(buf_p, name4, MAXPLAYERNAME); + WRITEMEM(buf_p, key4, PUBKEYLENGTH); } WRITEUINT8(buf_p, nodetoplayer[node]); // consoleplayer @@ -3896,6 +4097,15 @@ void CL_RemoveSplitscreenPlayer(UINT8 p) SendKick(p, KICK_MSG_PLAYER_QUIT); } +static void GotOurIP(UINT32 address) +{ + const unsigned char * p = (const unsigned char *)&address; + #ifdef DEVELOP + CONS_Printf("Got IP of %u.%u.%u.%u\n", p[0], p[1], p[2], p[3]); + #endif + ourIP = address; +} + // is there a game running boolean Playing(void) { @@ -3932,13 +4142,18 @@ boolean SV_SpawnServer(void) else doomcom->numslots = 1; } + ourIP = 0; + if (netgame && server) + STUN_bind(GotOurIP); + // strictly speaking, i'm not convinced the following is necessary // but I'm not confident enough to remove it entirely in case it breaks something { UINT8 *availabilitiesbuffer = R_GetSkinAvailabilities(false, false); SINT8 node = 0; for (; node < MAXNETNODES; node++) - result |= SV_AddWaitingPlayers(node, availabilitiesbuffer, cv_playername[0].zstring, cv_playername[1].zstring, cv_playername[2].zstring, cv_playername[3].zstring); + result |= SV_AddWaitingPlayers(node, availabilitiesbuffer, cv_playername[0].zstring, PR_GetLocalPlayerProfile(0)->public_key, cv_playername[1].zstring, PR_GetLocalPlayerProfile(1)->public_key, + cv_playername[2].zstring, PR_GetLocalPlayerProfile(2)->public_key, cv_playername[3].zstring, PR_GetLocalPlayerProfile(3)->public_key); } return result; #endif @@ -4012,6 +4227,34 @@ static size_t TotalTextCmdPerTic(tic_t tic) return total; } +#ifdef SIGNGAMETRAFFIC + static boolean IsSplitPlayerOnNodeGuest(int node, int split) + { + char allZero[PUBKEYLENGTH]; + memset(allZero, 0, PUBKEYLENGTH); + + if (split == 0) + return (memcmp(players[nodetoplayer[node]].public_key, allZero, PUBKEYLENGTH) == 0); + else if (split == 1) + return (memcmp(players[nodetoplayer2[node]].public_key, allZero, PUBKEYLENGTH) == 0); + else if (split == 2) + return (memcmp(players[nodetoplayer3[node]].public_key, allZero, PUBKEYLENGTH) == 0); + else if (split == 3) + return (memcmp(players[nodetoplayer4[node]].public_key, allZero, PUBKEYLENGTH) == 0); + else + I_Error("IsSplitPlayerOnNodeGuest: Out of bounds"); + return false; // unreachable + } +#endif + +static boolean IsPlayerGuest(int player) +{ + char allZero[PUBKEYLENGTH]; + memset(allZero, 0, PUBKEYLENGTH); + + return (memcmp(players[player].public_key, allZero, PUBKEYLENGTH) == 0); +} + /** Called when a PT_CLIENTJOIN packet is received * * \param node The packet sender @@ -4029,6 +4272,7 @@ static void HandleConnect(SINT8 node) UINT8 maxplayers = min((dedicated ? MAXPLAYERS-1 : MAXPLAYERS), cv_maxconnections.value); UINT8 connectedplayers = 0; + for (i = dedicated ? 1 : 0; i < MAXPLAYERS; i++) if (playernode[i] != UINT8_MAX) // We use this to count players because it is affected by SV_AddWaitingPlayers when more than one client joins on the same tic, unlike playeringame and D_NumPlayers. UINT8_MAX denotes no node for that player connectedplayers++; @@ -4116,6 +4360,7 @@ static void HandleConnect(SINT8 node) } else { + int sigcheck; boolean newnode = false; for (i = 0; i < netbuffer->u.clientcfg.localplayers - playerpernode[node]; i++) @@ -4126,6 +4371,37 @@ static void HandleConnect(SINT8 node) SV_SendRefuse(node, "Bad player name"); return; } + + if (node == 0) // Hey, that's us. We're always allowed to do what we want. + { + memcpy(lastReceivedKey[node][i], PR_GetLocalPlayerProfile(i)->public_key, sizeof(lastReceivedKey[node][i])); + } + else // Remote player, gotta check their signature. + { + char allZero[PUBKEYLENGTH]; + memset(allZero, 0, sizeof(allZero)); + + if (memcmp(lastReceivedKey[node][i], allZero, PUBKEYLENGTH) == 0) // IsSplitPlayerOnNodeGuest isn't appropriate here, they're not in-game yet! + { + if (!cv_allowguests.value) + { + SV_SendRefuse(node, M_GetText("The server doesn't allow GUESTs.\nCreate a profile to join!")); + return; + } + + sigcheck = 0; // Always succeeds. Yes, this is a success response. C R Y P T O + } + else + { + sigcheck = crypto_eddsa_check(netbuffer->u.clientcfg.challengeResponse[i], lastReceivedKey[node][i], lastSentChallenge[node], CHALLENGELENGTH); + } + + if (netgame && sigcheck != 0) + { + SV_SendRefuse(node, M_GetText("Signature verification failed.")); + return; + } + } } memcpy(availabilitiesbuffer, netbuffer->u.clientcfg.availabilities, sizeof(availabilitiesbuffer)); @@ -4165,7 +4441,8 @@ static void HandleConnect(SINT8 node) DEBFILE("send savegame\n"); } - SV_AddWaitingPlayers(node, availabilitiesbuffer, names[0], names[1], names[2], names[3]); + SV_AddWaitingPlayers(node, availabilitiesbuffer, names[0], lastReceivedKey[node][0], names[1], lastReceivedKey[node][1], + names[2], lastReceivedKey[node][2], names[3], lastReceivedKey[node][3]); joindelay += cv_joindelay.value * TICRATE; player_joining = true; } @@ -4204,6 +4481,23 @@ static void HandleTimeout(SINT8 node) M_StartMessage(M_GetText("Server Timeout\n\nPress (B)\n"), NULL, MM_NOTHING); } +// Called when a signature check fails and we suspect the server is playing games. +void HandleSigfail(const char *string) +{ + if (server) // This situation is basically guaranteed to be nonsense. + { + CONS_Alert(CONS_ERROR, "Auth error! %s\n", string); + return; // Keep the game running, you're probably testing. + } + + LUA_HookBool(false, HOOK(GameQuit)); + D_QuitNetGame(); + CL_Reset(); + D_ClearState(); + M_StartControlPanel(); + M_StartMessage(va(M_GetText("Signature check failed.\n(%s)\nPress (B)\n"), string), NULL, MM_NOTHING); +} + /** Called when a PT_SERVERINFO packet is received * * \param node The packet sender @@ -4236,6 +4530,14 @@ static void PT_WillResendGamestate(void) if (server || cl_redownloadinggamestate) return; + // Don't let the server pull a fast one with everyone's identity! + // Save the public keys we see, so if the server tries to swap one, we'll know. + int i; + for (i = 0; i < MAXPLAYERS; i++) + { + memcpy(priorKeys[i], players[i].public_key, sizeof(priorKeys[i])); + } + // Send back a PT_CANRECEIVEGAMESTATE packet to the server // so they know they can start sending the game state netbuffer->packettype = PT_CANRECEIVEGAMESTATE; @@ -4490,7 +4792,10 @@ static void HandlePacketFromAwayNode(SINT8 node) case PT_NODETIMEOUT: case PT_CLIENTQUIT: if (server) + { Net_CloseConnection(node); + nodeneedsauth[node] = false; + } break; case PT_CLIENTCMD: @@ -4501,7 +4806,28 @@ static void HandlePacketFromAwayNode(SINT8 node) if (node == servernode) break; /* FALLTHRU */ + case PT_CLIENTKEY: + if (server) + { + PT_ClientKey(node); + // Client's not in the server yet, but we still need to lock up the node. + // Otherwise, someone else could request a challenge on the same node and trash it. + nodeneedsauth[node] = true; + freezetimeout[node] = I_GetTime() + jointimeout; + } + break; + case PT_SERVERCHALLENGE: + if (server && serverrunning && node != servernode) + { + Net_CloseConnection(node); + break; + } + if (cl_mode != CL_WAITCHALLENGE) + break; + memcpy(awaitingChallenge, netbuffer->u.serverchallenge.secret, sizeof(awaitingChallenge)); + cl_mode = CL_ASKJOIN; + break; default: DEBFILE(va("unknown packet received (%d) from unknown host\n",netbuffer->packettype)); Net_CloseConnection(node); @@ -4534,6 +4860,19 @@ static boolean CheckForSpeedHacks(UINT8 p) return false; } +static char NodeToSplitPlayer(int node, int split) +{ + if (split == 0) + return nodetoplayer[node]; + else if (split == 1) + return nodetoplayer2[node]; + else if (split == 2) + return nodetoplayer3[node]; + else if (split == 3) + return nodetoplayer4[node]; + return -1; +} + /** Handles a packet received from a node that is in game * * \param node The packet sender @@ -4561,6 +4900,49 @@ static void HandlePacketFromPlayer(SINT8 node) if (netconsole >= MAXPLAYERS) I_Error("bad table nodetoplayer: node %d player %d", doomcom->remotenode, netconsole); #endif + + +#ifdef SIGNGAMETRAFFIC + if (server) + { + + int splitnodes; + if (IsPacketSigned(netbuffer->packettype)) + { + for (splitnodes = 0; splitnodes < MAXSPLITSCREENPLAYERS; splitnodes++) + { + int targetplayer = NodeToSplitPlayer(node, splitnodes); + if (targetplayer == -1) + continue; + + const void* message = &netbuffer->u; + if (IsSplitPlayerOnNodeGuest(node, splitnodes) || demo.playback) + { + //CONS_Printf("Throwing out a guest signature from node %d player %d\n", node, splitnodes); + } + else + { + if (crypto_eddsa_check(netbuffer->signature[splitnodes], players[targetplayer].public_key, message, doomcom->datalength - BASEPACKETSIZE)) + { + CONS_Alert(CONS_ERROR, "SIGFAIL! Packet type %d from node %d player %d\nkey %s size %d netconsole %d\n", + netbuffer->packettype, node, splitnodes, + GetPrettyRRID(players[targetplayer].public_key, true), doomcom->datalength - BASEPACKETSIZE, netconsole); + + // Something scary can happen when multiple kicks that resolve to the same node are processed in quick succession. + // Sometimes, a kick will still be left to process after the player's been disposed, and that causes the kick to resolve on the server instead! + // This sucks, so we check for a stale/misfiring kick beforehand. + if (netconsole != -1) + SendKick(netconsole, KICK_MSG_SIGFAIL); + // Net_CloseConnection(node); + // nodeingame[node] = false; + return; + } + } + + } + } + } +#endif switch (netbuffer->packettype) { @@ -4851,6 +5233,7 @@ static void HandlePacketFromPlayer(SINT8 node) } Net_CloseConnection(node); nodeingame[node] = false; + nodeneedsauth[node] = false; break; case PT_CANRECEIVEGAMESTATE: PT_CanReceiveGamestate(node); @@ -4980,6 +5363,140 @@ static void HandlePacketFromPlayer(SINT8 node) if (client) CL_PrepareDownloadLuaFile(); break; + case PT_CHALLENGEALL: + if (demo.playback || node != servernode) // SERVER should still respond to this to prove its own identity, just not from clients. + break; + + int challengeplayers; + + memcpy(lastChallengeAll, netbuffer->u.challengeall.secret, sizeof(lastChallengeAll)); + + shouldsign_t safe = ShouldSignChallenge(lastChallengeAll); + if (safe != SIGN_OK) + { + if (safe == SIGN_BADIP) + HandleSigfail("External server sent the wrong IP"); + else if (safe == SIGN_BADTIME) + HandleSigfail("Bad timestamp - check your clocks"); + else + HandleSigfail("Unknown auth error - contact a developer"); + break; + } + + netbuffer->packettype = PT_RESPONSEALL; + + #ifdef DEVELOP + if (cv_noresponse.value) + { + CV_AddValue(&cv_noresponse, -1); + CONS_Alert(CONS_WARNING, "cv_noresponse enabled, not sending PT_RESPONSEALL\n"); + break; + } + #endif + + // Don't leak uninitialized memory. + memset(&netbuffer->u.responseall, 0, sizeof(netbuffer->u.responseall)); + + for (challengeplayers = 0; challengeplayers <= splitscreen; challengeplayers++) + { + uint8_t signature[SIGNATURELENGTH]; + profile_t *localProfile = PR_GetLocalPlayerProfile(challengeplayers); + if (!PR_IsLocalPlayerGuest(challengeplayers)) // GUESTS don't have keys + { + crypto_eddsa_sign(signature, localProfile->secret_key, lastChallengeAll, sizeof(lastChallengeAll)); + + // If our keys are garbage (corrupted profile?), fail here instead of when the server boots us, so the player knows what's going on. + if (crypto_eddsa_check(signature, localProfile->public_key, lastChallengeAll, sizeof(lastChallengeAll)) != 0) + I_Error("Couldn't self-verify key associated with player %d, profile %d.\nProfile data may be corrupted.", challengeplayers, cv_lastprofile[challengeplayers].value); + } + + #ifdef DEVELOP + if (cv_badresponse.value) + { + CV_AddValue(&cv_badresponse, -1); + CONS_Alert(CONS_WARNING, "cv_badresponse enabled, scrubbing signature from PT_RESPONSEALL\n"); + memset(signature, 0, sizeof(signature)); + } + #endif + + memcpy(netbuffer->u.responseall.signature[challengeplayers], signature, sizeof(signature)); + } + + HSendPacket(servernode, true, 0, sizeof(netbuffer->u.responseall)); + break; + case PT_RESPONSEALL: + if (demo.playback || client) + break; + + int responseplayer; + for (responseplayer = 0; responseplayer < MAXSPLITSCREENPLAYERS; responseplayer++) + { + int targetplayer = NodeToSplitPlayer(node, responseplayer); + if (targetplayer == -1) + continue; + + if (!IsPlayerGuest(targetplayer)) + { + if (crypto_eddsa_check(netbuffer->u.responseall.signature[responseplayer], players[targetplayer].public_key, lastChallengeAll, sizeof(lastChallengeAll))) + { + // Something scary can happen when multiple kicks that resolve to the same node are processed in quick succession. + // Sometimes, a kick will still be left to process after the player's been disposed, and that causes the kick to resolve on the server instead! + // This sucks, so we check for a stale/misfiring kick beforehand. + if (playernode[targetplayer] != 0) + SendKick(targetplayer, KICK_MSG_SIGFAIL); + break; + } + else + { + memcpy(lastReceivedSignature[targetplayer], netbuffer->u.responseall.signature[responseplayer], sizeof(lastReceivedSignature[targetplayer])); + } + } + } + break; + case PT_RESULTSALL: + if (demo.playback || server || node != servernode || !expectChallenge) + break; + + int resultsplayer; + uint8_t allZero[PUBKEYLENGTH]; + memset(allZero, 0, sizeof(PUBKEYLENGTH)); + + for (resultsplayer = 0; resultsplayer < MAXPLAYERS; resultsplayer++) + { + if (!playeringame[resultsplayer]) + { + continue; + } + else if (IsPlayerGuest(resultsplayer)) + { + continue; + } + else if (memcmp(knownWhenChallenged[resultsplayer], allZero, sizeof(PUBKEYLENGTH)) == 0) + { + // Wasn't here for the challenge. + continue; + } + else if (memcmp(knownWhenChallenged[resultsplayer], players[resultsplayer].public_key, sizeof(knownWhenChallenged[resultsplayer])) != 0) + { + // A player left after the challenge process started, and someone else took their place. + // That means they haven't received a challenge either. + continue; + } + else + { + if (crypto_eddsa_check(netbuffer->u.resultsall.signature[resultsplayer], + knownWhenChallenged[resultsplayer], lastChallengeAll, sizeof(lastChallengeAll))) + { + CONS_Alert(CONS_WARNING, "PT_RESULTSALL had invalid signature %s for node %d player %d split %d, something doesn't add up!\n", + GetPrettyRRID(netbuffer->u.resultsall.signature[resultsplayer], true), playernode[resultsplayer], resultsplayer, players[resultsplayer].splitscreenindex); + HandleSigfail("Server sent invalid client signature."); + break; + } + } + } + csprng(lastChallengeAll, sizeof(lastChallengeAll)); + expectChallenge = false; + break; default: DEBFILE(va("UNKNOWN PACKET TYPE RECEIVED %d from host %d\n", netbuffer->packettype, node)); @@ -5814,6 +6331,168 @@ static void UpdatePingTable(void) } } +// It's that time again! Send everyone a safe message to sign, so we can show off their signature and prove we're playing fair. +static void SendChallenges(void) +{ + int i; + netbuffer->packettype = PT_CHALLENGEALL; + + #ifdef DEVELOP + if (cv_nochallenge.value) + { + CV_AddValue(&cv_nochallenge, -1); + CONS_Alert(CONS_WARNING, "cv_nochallenge enabled, not sending PT_CHALLENGEALL\n"); + return; + } + #endif + + memset(knownWhenChallenged, 0, sizeof(knownWhenChallenged)); + memset(lastReceivedSignature, 0, sizeof(lastReceivedSignature)); + + GenerateChallenge(netbuffer->u.challengeall.secret); + memcpy(lastChallengeAll, netbuffer->u.challengeall.secret, sizeof(lastChallengeAll)); + + // Take note of everyone's current key, so that players who disconnect and are replaced aren't held to the old player's challenge. + for (i = 0; i < MAXPLAYERS; i++) + { + if (playeringame[i]) + memcpy(knownWhenChallenged[i], players[i].public_key, sizeof(knownWhenChallenged[i])); + } + + for (i = 0; i < MAXNETNODES; i++) + { + if (nodeingame[i]) + HSendPacket(i, true, 0, sizeof(challengeall_pak)); + } +} + +// Before we start sending out the results, we need to kick everyone who didn't respond. +// (If we try to do both at once, clients will still see players who failled in-game when the results arrive...) +static void KickUnverifiedPlayers(void) +{ + int i; + uint8_t allZero[SIGNATURELENGTH]; + memset(allZero, 0, SIGNATURELENGTH); + + for (i = 0; i < MAXPLAYERS; i++) + { + if (!playeringame[i]) + continue; + if (memcmp(lastReceivedSignature[i], allZero, SIGNATURELENGTH) == 0) // We never got a response! + { + if (!IsPlayerGuest(i) && memcmp(&knownWhenChallenged[i], &players[i].public_key, sizeof(knownWhenChallenged[i])) == 0) + { + if (playernode[i] != servernode) + SendKick(i, KICK_MSG_SIGFAIL); + } + } + } +} + +// +static void SendChallengeResults(void) +{ + int i; + netbuffer->packettype = PT_RESULTSALL; + + #ifdef DEVELOP + if (cv_noresults.value) + { + CV_AddValue(&cv_noresults, -1); + CONS_Alert(CONS_WARNING, "cv_noresults enabled, not sending PT_RESULTSALL\n"); + return; + } + #endif + + uint8_t allZero[SIGNATURELENGTH]; + memset(allZero, 0, sizeof(allZero)); + + memset(&netbuffer->u.resultsall, 0, sizeof(netbuffer->u.resultsall)); + + for (i = 0; i < MAXPLAYERS; i++) + { + if (!playeringame[i]) + continue; + + // Don't try to transmit signatures for players who didn't get here in time to send one. + // (Everyone who had their chance should have been kicked by KickUnverifiedPlayers by now.) + if (memcmp(lastReceivedSignature[i], allZero, SIGNATURELENGTH) == 0) + continue; + + memcpy(netbuffer->u.resultsall.signature[i], lastReceivedSignature[i], sizeof(netbuffer->u.resultsall.signature[i])); + #ifdef DEVELOP + if (cv_badresults.value) + { + CV_AddValue(&cv_badresults, -1); + CONS_Alert(CONS_WARNING, "cv_badresults enabled, scrubbing signature from PT_RESULTSALL\n"); + memset(netbuffer->u.resultsall.signature[i], 0, sizeof(netbuffer->u.resultsall.signature[i])); + } + #endif + } + + for (i = 0; i < MAXNETNODES; i++) + { + if (nodeingame[i]) + HSendPacket(i, true, 0, sizeof(resultsall_pak)); + } +} + +// Who should we try to verify when results come in? +// Store a public key for every active slot, so if players shuffle during challenge leniency, +// we don't incorrectly try to verify someone who didn't even get a challenge, throw a tantrum, and bail. +static void CheckPresentPlayers(void) +{ + int i; + memset(knownWhenChallenged, 0, sizeof(knownWhenChallenged)); + + for (i = 0; i < MAXPLAYERS; i++) + { + if (!playeringame[i]) + { + continue; + } + else if (IsPlayerGuest(i)) + { + continue; + } + else + { + memcpy(knownWhenChallenged[i], players[i].public_key, sizeof(knownWhenChallenged[i])); + } + } +} + +// Handle "client-to-client" auth challenge flow. +static void UpdateChallenges(void) +{ + if (!(Playing() && netgame)) + return; + + if (server) + { + if (leveltime == CHALLENGEALL_START) + SendChallenges(); + + if (leveltime == CHALLENGEALL_KICKUNRESPONSIVE) + KickUnverifiedPlayers(); + + if (leveltime == CHALLENGEALL_SENDRESULTS) + SendChallengeResults(); + + } + else // client + { + if (leveltime <= CHALLENGEALL_START) + expectChallenge = true; + + if (leveltime == CHALLENGEALL_START) + CheckPresentPlayers(); + + if (leveltime > CHALLENGEALL_CLIENTCUTOFF && expectChallenge) + HandleSigfail("Didn't receive client signatures."); + } +} + static void RenewHolePunch(void) { static time_t past; @@ -5835,7 +6514,7 @@ static void HandleNodeTimeouts(void) if (server) { for (i = 1; i < MAXNETNODES; i++) - if (nodeingame[i] && freezetimeout[i] < I_GetTime()) + if ((nodeingame[i] || nodeneedsauth[i]) && freezetimeout[i] < I_GetTime()) Net_ConnectionTimeout(i); // In case the cvar value was lowered @@ -5965,6 +6644,8 @@ void NetUpdate(void) UpdatePingTable(); + UpdateChallenges(); + if (client) maketic = neededtic; diff --git a/src/d_clisrv.h b/src/d_clisrv.h index 0b51846ca..c50dce4ba 100644 --- a/src/d_clisrv.h +++ b/src/d_clisrv.h @@ -54,6 +54,11 @@ applications may follow different packet versions. // This just works as a quick implementation. #define MAXGENTLEMENDELAY TICRATE +#define PUBKEYLENGTH 32 // Enforced by Monocypher EdDSA +#define PRIVKEYLENGTH 64 // Enforced by Monocypher EdDSA +#define SIGNATURELENGTH 64 // Enforced by Monocypher EdDSA +#define CHALLENGELENGTH 64 // Servers verify client identity by giving them messages to sign. How long are these messages? + // // Packet structure // @@ -119,16 +124,29 @@ typedef enum PT_LOGIN, // Login attempt from the client. PT_PING, // Packet sent to tell clients the other client's latency to server. + + PT_CLIENTKEY, // "Here's my public key" + PT_SERVERCHALLENGE, // "Prove it" + + PT_CHALLENGEALL, // Prove to the other clients you are who you say you are, sign this random bullshit! + PT_RESPONSEALL, // OK, here is my signature on that random bullshit + PT_RESULTSALL, // Here's what everyone responded to PT_CHALLENGEALL with, if this is wrong or you don't receive it disconnect + NUMPACKETTYPE } packettype_t; +typedef enum +{ + SIGN_OK, + SIGN_BADTIME, // Timestamp differs by too much, suspect reuse of an old challenge. + SIGN_BADIP // Asked to sign the wrong IP by an external host, suspect reuse of another server's challenge. +} shouldsign_t; + #ifdef PACKETDROP void Command_Drop(void); void Command_Droprate(void); #endif -#ifdef _DEBUG void Command_Numnodes(void); -#endif #if defined(_MSC_VER) #pragma pack(1) @@ -252,6 +270,7 @@ struct clientconfig_pak UINT8 mode; char names[MAXSPLITSCREENPLAYERS][MAXPLAYERNAME]; UINT8 availabilities[MAXAVAILABILITY]; + uint8_t challengeResponse[MAXSPLITSCREENPLAYERS][SIGNATURELENGTH]; } ATTRPACK; #define SV_SPEEDMASK 0x03 // used to send kartspeed @@ -346,6 +365,31 @@ struct filesneededconfig_pak UINT8 files[MAXFILENEEDED]; // is filled with writexxx (byteptr.h) } ATTRPACK; +struct clientkey_pak +{ + uint8_t key[MAXSPLITSCREENPLAYERS][PUBKEYLENGTH]; +} ATTRPACK; + +struct serverchallenge_pak +{ + uint8_t secret[CHALLENGELENGTH]; +} ATTRPACK; + +struct challengeall_pak +{ + uint8_t secret[CHALLENGELENGTH]; +} ATTRPACK; + +struct responseall_pak +{ + uint8_t signature[MAXSPLITSCREENPLAYERS][SIGNATURELENGTH]; +} ATTRPACK; + +struct resultsall_pak +{ + uint8_t signature[MAXPLAYERS][SIGNATURELENGTH]; +} ATTRPACK; + // // Network packet data // @@ -356,6 +400,9 @@ struct doomdata_t UINT8 ackreturn; // The return of the ack number UINT8 packettype; +#ifdef SIGNGAMETRAFFIC + uint8_t signature[MAXSPLITSCREENPLAYERS][SIGNATURELENGTH]; +#endif UINT8 reserved; // Padding union { @@ -380,6 +427,11 @@ struct doomdata_t INT32 filesneedednum; // 4 bytes filesneededconfig_pak filesneededcfg; // ??? bytes UINT32 pingtable[MAXPLAYERS+1]; // 68 bytes + clientkey_pak clientkey; // 32 bytes + serverchallenge_pak serverchallenge; // 256 bytes + challengeall_pak challengeall; // 256 bytes + responseall_pak responseall; // 256 bytes + resultsall_pak resultsall; // 1024 bytes. Also, you really shouldn't trust anything here. } u; // This is needed to pack diff packet types data together } ATTRPACK; @@ -419,6 +471,7 @@ extern consvar_t cv_playbackspeed; #define KICK_MSG_PING_HIGH 6 #define KICK_MSG_CUSTOM_KICK 7 #define KICK_MSG_CUSTOM_BAN 8 +#define KICK_MSG_SIGFAIL 9 typedef enum { @@ -443,6 +496,20 @@ extern UINT16 software_MAXPACKETLENGTH; extern boolean acceptnewnode; extern SINT8 servernode; extern char connectedservername[MAXSERVERNAME]; +extern UINT32 ourIP; +extern uint8_t lastReceivedKey[MAXNETNODES][MAXSPLITSCREENPLAYERS][PUBKEYLENGTH]; +extern uint8_t lastSentChallenge[MAXNETNODES][CHALLENGELENGTH]; +extern uint8_t lastChallengeAll[CHALLENGELENGTH]; +extern uint8_t lastReceivedSignature[MAXPLAYERS][SIGNATURELENGTH]; +extern uint8_t knownWhenChallenged[MAXPLAYERS][PUBKEYLENGTH]; +extern boolean expectChallenge; + +// We give clients a chance to verify each other once per race. +// When is that challenge sent, and when should clients bail if they don't receive the responses? +#define CHALLENGEALL_START (TICRATE*5) // Server sends challenges here. +#define CHALLENGEALL_KICKUNRESPONSIVE (TICRATE*10) // Server kicks players that haven't submitted signatures here. +#define CHALLENGEALL_SENDRESULTS (TICRATE*15) // Server waits for kicks to process until here. (Failing players shouldn't be in-game when results are received, or clients get spooked.) +#define CHALLENGEALL_CLIENTCUTOFF (TICRATE*20) // If challenge process hasn't completed by now, clients who were in-game for CHALLENGEALL_START should leave. void Command_Ping_f(void); extern tic_t connectiontimeout; @@ -465,10 +532,27 @@ extern consvar_t cv_joinnextround; extern consvar_t cv_discordinvites; +extern consvar_t cv_allowguests; + +#ifdef DEVELOP + extern consvar_t cv_badjoin; + extern consvar_t cv_badtraffic; + extern consvar_t cv_badresponse; + extern consvar_t cv_noresponse; + extern consvar_t cv_nochallenge; + extern consvar_t cv_badresults; + extern consvar_t cv_noresults; + extern consvar_t cv_badtime; + extern consvar_t cv_badip; +#endif + // Used in d_net, the only dependence tic_t ExpandTics(INT32 low, tic_t basetic); void D_ClientServerInit(void); +void GenerateChallenge(uint8_t *buf); +shouldsign_t ShouldSignChallenge(uint8_t *message); + // Initialise the other field void RegisterNetXCmd(netxcmd_t id, void (*cmd_f)(UINT8 **p, INT32 playernum)); void SendNetXCmdForPlayer(UINT8 playerid, netxcmd_t id, const void *param, size_t nparam); @@ -549,6 +633,8 @@ void CL_ClearRewinds(void); rewind_t *CL_SaveRewindPoint(size_t demopos); rewind_t *CL_RewindToTime(tic_t time); +void HandleSigfail(const char *string); + #ifdef __cplusplus } // extern "C" #endif diff --git a/src/d_main.c b/src/d_main.c index a206148ef..3dcb8c4c2 100644 --- a/src/d_main.c +++ b/src/d_main.c @@ -71,6 +71,9 @@ #include "g_input.h" // tutorial mode control scheming #include "m_perfstats.h" +#include "monocypher/monocypher.h" +#include "stun.h" + // SRB2Kart #include "k_grandprix.h" #include "doomstat.h" diff --git a/src/d_net.c b/src/d_net.c index 3799ecd13..917b181fe 100644 --- a/src/d_net.c +++ b/src/d_net.c @@ -29,6 +29,8 @@ #include "z_zone.h" #include "i_tcp.h" #include "d_main.h" // srb2home +#include "stun.h" +#include "monocypher/monocypher.h" // // NETWORKING @@ -81,6 +83,7 @@ boolean (*I_NetOpenSocket)(void) = NULL; boolean (*I_Ban) (INT32 node) = NULL; void (*I_ClearBans)(void) = NULL; const char *(*I_GetNodeAddress) (INT32 node) = NULL; +UINT32 (*I_GetNodeAddressInt) (INT32 node) = NULL; const char *(*I_GetBanAddress) (size_t ban) = NULL; const char *(*I_GetBanMask) (size_t ban) = NULL; const char *(*I_GetBanUsername) (size_t ban) = NULL; @@ -90,6 +93,7 @@ boolean (*I_SetBanAddress) (const char *address, const char *mask) = NULL; boolean (*I_SetBanUsername) (const char *username) = NULL; boolean (*I_SetBanReason) (const char *reason) = NULL; boolean (*I_SetUnbanTime) (time_t timestamp) = NULL; +boolean (*I_IsExternalAddress) (const void *p) = NULL; bannednode_t *bannednode = NULL; @@ -629,7 +633,12 @@ static void InitAck(void) ackpak[i].acknum = 0; for (i = 0; i < MAXNETNODES; i++) + { InitNode(&nodes[i]); + + csprng(lastSentChallenge[i], sizeof(lastSentChallenge[i])); + csprng(lastReceivedKey[i], sizeof(lastReceivedKey[i])); + } } /** Removes all acks of a given packet type @@ -699,6 +708,9 @@ void Net_CloseConnection(INT32 node) if (server) SV_AbortLuaFileTransfer(node); I_NetFreeNodenum(node); + + csprng(lastSentChallenge[node], sizeof(lastSentChallenge[node])); + csprng(lastReceivedKey[node], sizeof(lastReceivedKey[node])); } // @@ -803,7 +815,14 @@ static const char *packettypename[NUMPACKETTYPE] = "LOGIN", - "PING" + "PING", + + "CLIENTKEY", + "SERVERCHALLENGE", + + "CHALLENGEALL", + "RESPONSEALL", + "RESULTSALL" }; static void DebugPrintpacket(const char *header) @@ -983,12 +1002,72 @@ static boolean ShouldDropPacket(void) } #endif +// Unused because Eidolon correctly pointed out that +512b on every packet was scary. +#ifdef SIGNGAMETRAFFIC + boolean IsPacketSigned(int packettype) + { + switch (packettype) + { + case PT_CLIENTCMD: + case PT_CLIENT2CMD: + case PT_CLIENT3CMD: + case PT_CLIENT4CMD: + case PT_CLIENTMIS: + case PT_CLIENT2MIS: + case PT_CLIENT3MIS: + case PT_CLIENT4MIS: + case PT_TEXTCMD: + case PT_TEXTCMD2: + case PT_TEXTCMD3: + case PT_TEXTCMD4: + case PT_LOGIN: + case PT_ASKLUAFILE: + case PT_SENDINGLUAFILE: + return true; + default: + return false; + } + } +#endif + // // HSendPacket // boolean HSendPacket(INT32 node, boolean reliable, UINT8 acknum, size_t packetlength) { doomcom->datalength = (INT16)(packetlength + BASEPACKETSIZE); + +#ifdef SIGNGAMETRAFFIC + if (IsPacketSigned(netbuffer->packettype)) + { + int i; + + for (i = 0; i < MAXSPLITSCREENPLAYERS; i++) + { + const void* message = &netbuffer->u; + //CONS_Printf("Signing packet type %d of length %d\n", netbuffer->packettype, packetlength); + if (PR_IsLocalPlayerGuest(i)) + memset(netbuffer->signature[i], 0, sizeof(netbuffer->signature[i])); + else + crypto_eddsa_sign(netbuffer->signature[i], PR_GetLocalPlayerProfile(i)->secret_key, message, packetlength); + } + + #ifdef DEVELOP + if (cv_badtraffic.value) + { + CV_AddValue(&cv_badtraffic, -1); + CONS_Alert(CONS_WARNING, "cv_badtraffic enabled, scrubbing signature from HSendPacket\n"); + memset(netbuffer->signature, 0, sizeof(netbuffer->signature)); + } + #endif + } + else + { + //CONS_Printf("NOT signing PT_%d of length %d, it doesn't need to be\n", netbuffer->packettype, packetlength); + memset(netbuffer->signature, 0, sizeof(netbuffer->signature)); + } +#endif + if (node == 0) // Packet is to go back to us { if ((rebound_head+1) % MAXREBOUND == rebound_tail) diff --git a/src/d_net.h b/src/d_net.h index 12ab72693..5cf4b9f77 100644 --- a/src/d_net.h +++ b/src/d_net.h @@ -46,6 +46,7 @@ extern SINT8 nodetoplayer3[MAXNETNODES]; // Say the numplayer for this node if a extern SINT8 nodetoplayer4[MAXNETNODES]; // Say the numplayer for this node if any (splitscreen == 3) extern UINT8 playerpernode[MAXNETNODES]; // Used specially for splitscreen extern boolean nodeingame[MAXNETNODES]; // Set false as nodes leave game +extern boolean nodeneedsauth[MAXNETNODES]; extern boolean serverrunning; @@ -68,6 +69,8 @@ void Net_AbortPacketType(UINT8 packettype); void Net_SendAcks(INT32 node); void Net_WaitAllAckReceived(UINT32 timeout); +boolean IsPacketSigned(int packettype); + #ifdef __cplusplus } // extern "C" #endif diff --git a/src/d_netcmd.c b/src/d_netcmd.c index 1efb49238..46c8f230e 100644 --- a/src/d_netcmd.c +++ b/src/d_netcmd.c @@ -946,6 +946,20 @@ void D_RegisterClientCommands(void) CV_RegisterVar(&cv_netticbuffer); CV_RegisterVar(&cv_mindelay); + CV_RegisterVar(&cv_allowguests); + + #ifdef DEVELOP + CV_RegisterVar(&cv_badjoin); + CV_RegisterVar(&cv_badtraffic); + CV_RegisterVar(&cv_badresponse); + CV_RegisterVar(&cv_noresponse); + CV_RegisterVar(&cv_nochallenge); + CV_RegisterVar(&cv_badresults); + CV_RegisterVar(&cv_noresults); + CV_RegisterVar(&cv_badtime); + CV_RegisterVar(&cv_badip); + #endif + // HUD CV_RegisterVar(&cv_alttitle); CV_RegisterVar(&cv_itemfinder); diff --git a/src/d_netfil.c b/src/d_netfil.c index ef99a2321..93d7e8eba 100644 --- a/src/d_netfil.c +++ b/src/d_netfil.c @@ -53,6 +53,7 @@ #include "k_menu.h" #include "md5.h" #include "filesrch.h" +#include "stun.h" #include @@ -1313,6 +1314,22 @@ void PT_FileReceived(void) SV_EndFileSend(doomcom->remotenode); } +// Someone knocked on the door with their public key. +// Give them a challenge to sign in their PT_CLIENTJOIN. +void PT_ClientKey(INT32 node) +{ + clientkey_pak *packet = (void*)&netbuffer->u.clientkey; + + memcpy(lastReceivedKey[node], packet->key, sizeof(lastReceivedKey[node])); + + netbuffer->packettype = PT_SERVERCHALLENGE; + + GenerateChallenge(lastSentChallenge[node]); + + memcpy(&netbuffer->u.serverchallenge, lastSentChallenge[node], sizeof(serverchallenge_pak)); + HSendPacket(node, false, 0, sizeof (serverchallenge_pak)); +} + static void SendAckPacket(fileack_pak *packet, UINT8 fileid) { size_t packetsize; diff --git a/src/d_netfil.h b/src/d_netfil.h index 26bffbc18..fb5a6d29a 100644 --- a/src/d_netfil.h +++ b/src/d_netfil.h @@ -106,6 +106,8 @@ boolean CL_CheckDownloadable(void); boolean CL_SendFileRequest(void); boolean PT_RequestFile(INT32 node); +void PT_ClientKey(INT32 node); + typedef enum { LFTNS_NONE, // This node is not connected diff --git a/src/d_player.h b/src/d_player.h index e362687fb..f1638024d 100644 --- a/src/d_player.h +++ b/src/d_player.h @@ -713,6 +713,8 @@ struct player_t mobj_t *stumbleIndicator; mobj_t *sliptideZipIndicator; + uint8_t public_key[PUBKEYLENGTH]; + #ifdef HWRENDER fixed_t fovadd; // adjust FOV for hw rendering #endif @@ -721,6 +723,9 @@ struct player_t roundconditions_t roundconditions; }; +// WARNING FOR ANYONE ABOUT TO ADD SOMETHING TO THE PLAYER STRUCT, G_PlayerReborn WANTS YOU TO SUFFER +// If data on player_t needs to persist between rounds or during the join process, modify G_PlayerReborn to preserve it. + #ifdef __cplusplus } // extern "C" #endif diff --git a/src/g_game.c b/src/g_game.c index d0b2b1066..1a4389c67 100644 --- a/src/g_game.c +++ b/src/g_game.c @@ -2436,6 +2436,8 @@ void G_PlayerReborn(INT32 player, boolean betweenmaps) SINT8 xtralife; + uint8_t public_key[PUBKEYLENGTH]; + // SRB2kart itemroulette_t itemRoulette; respawnvars_t respawn; @@ -2509,6 +2511,7 @@ void G_PlayerReborn(INT32 player, boolean betweenmaps) // SRB2kart memcpy(&itemRoulette, &players[player].itemRoulette, sizeof (itemRoulette)); memcpy(&respawn, &players[player].respawn, sizeof (respawn)); + memcpy(&public_key, &players[player].public_key, sizeof(public_key)); if (betweenmaps || leveltime < introtime) { @@ -2678,6 +2681,8 @@ void G_PlayerReborn(INT32 player, boolean betweenmaps) memcpy(&p->itemRoulette, &itemRoulette, sizeof (p->itemRoulette)); memcpy(&p->respawn, &respawn, sizeof (p->respawn)); + memcpy(&p->public_key, &public_key, sizeof(p->public_key)); + if (saveroundconditions) memcpy(&p->roundconditions, &roundconditions, sizeof (p->roundconditions)); diff --git a/src/i_net.h b/src/i_net.h index bcd07eb17..5755f6169 100644 --- a/src/i_net.h +++ b/src/i_net.h @@ -166,6 +166,7 @@ extern void (*I_NetRegisterHolePunch)(void); extern boolean (*I_Ban) (INT32 node); extern void (*I_ClearBans)(void); extern const char *(*I_GetNodeAddress) (INT32 node); +extern UINT32 (*I_GetNodeAddressInt) (INT32 node); extern const char *(*I_GetBanAddress) (size_t ban); extern const char *(*I_GetBanMask) (size_t ban); extern const char *(*I_GetBanUsername) (size_t ban); @@ -175,6 +176,7 @@ extern boolean (*I_SetBanAddress) (const char *address,const char *mask); extern boolean (*I_SetBanUsername) (const char *username); extern boolean (*I_SetBanReason) (const char *reason); extern boolean (*I_SetUnbanTime) (time_t timestamp); +extern boolean (*I_IsExternalAddress) (const void *p); struct bannednode_t { diff --git a/src/i_tcp.c b/src/i_tcp.c index d99029b44..18e79fbd6 100644 --- a/src/i_tcp.c +++ b/src/i_tcp.c @@ -387,6 +387,20 @@ static const char *SOCK_GetNodeAddress(INT32 node) return SOCK_AddrToStr(&clientaddress[node]); } +static UINT32 SOCK_GetNodeAddressInt(INT32 node) +{ + if (nodeconnected[node] && clientaddress[node].any.sa_family == AF_INET) + { + return clientaddress[node].ip4.sin_addr.s_addr; + } + else + { + I_Error("SOCK_GetNodeAddressInt: Node %d is not IPv4!\n", node); + } + + return 0; +} + static const char *SOCK_GetBanAddress(size_t ban) { if (ban >= numbans) @@ -459,7 +473,7 @@ static void cleanupnodes(void) // Why can't I start at zero? for (j = 1; j < MAXNETNODES; j++) - if (!(nodeingame[j] || SendingFile(j))) + if (!(nodeingame[j] || nodeneedsauth[j] || SendingFile(j))) nodeconnected[j] = false; } @@ -489,7 +503,6 @@ static SINT8 getfreenode(void) return -1; } -#ifdef _DEBUG void Command_Numnodes(void) { INT32 connected = 0; @@ -527,7 +540,6 @@ void Command_Numnodes(void) "Ingame: %d\n", connected, ingame); } -#endif static boolean hole_punch(ssize_t c) { @@ -1510,6 +1522,31 @@ static void SOCK_ClearBans(void) banned = NULL; } +// https://github.com/jameds/holepunch/blob/master/holepunch.c#L75 +static int SOCK_IsExternalAddress (const void *p) +{ + const int a = ((const unsigned char*)p)[0]; + const int b = ((const unsigned char*)p)[1]; + + if (*(const UINT32*)p == (UINT32)~0)/* 255.255.255.255 */ + return 0; + + switch (a) + { + case 0: + case 10: + case 127: + return 0; + case 172: + return (b & ~15) != 16;/* 16 - 31 */ + case 192: + return b != 168; + default: + return 1; + } +} + + boolean I_InitTcpNetwork(void) { char serverhostname[255]; @@ -1600,6 +1637,7 @@ boolean I_InitTcpNetwork(void) I_Ban = SOCK_Ban; I_ClearBans = SOCK_ClearBans; I_GetNodeAddress = SOCK_GetNodeAddress; + I_GetNodeAddressInt = SOCK_GetNodeAddressInt; I_GetBanAddress = SOCK_GetBanAddress; I_GetBanMask = SOCK_GetBanMask; I_GetBanUsername = SOCK_GetBanUsername; @@ -1609,6 +1647,8 @@ boolean I_InitTcpNetwork(void) I_SetBanUsername = SOCK_SetBanUsername; I_SetBanReason = SOCK_SetBanReason; I_SetUnbanTime = SOCK_SetUnbanTime; + I_IsExternalAddress = SOCK_IsExternalAddress; + bannednode = SOCK_bannednode; return ret; diff --git a/src/k_menudraw.c b/src/k_menudraw.c index abf0f7e1f..175422662 100644 --- a/src/k_menudraw.c +++ b/src/k_menudraw.c @@ -1768,7 +1768,8 @@ static void M_DrawProfileCard(INT32 x, INT32 y, boolean greyedout, profile_t *p) if (p != NULL) { V_DrawProfileNum(x + 37 + 10, y + 131, 0, PR_GetProfileNum(p)); - V_DrawCenteredThinString(x, y + 151, V_GRAYMAP|V_6WIDTHSPACE, p->playername); + V_DrawCenteredThinString(x, y + 141, V_GRAYMAP|V_6WIDTHSPACE, p->playername); + V_DrawCenteredThinString(x, y + 151, V_GRAYMAP|V_6WIDTHSPACE, GetPrettyRRID(p->public_key, true)); } } diff --git a/src/k_profiles.c b/src/k_profiles.c index 75f2e77b4..aad3fd221 100644 --- a/src/k_profiles.c +++ b/src/k_profiles.c @@ -18,6 +18,8 @@ #include "k_profiles.h" #include "z_zone.h" #include "r_skins.h" +#include "monocypher/monocypher.h" +#include "stun.h" // List of all the profiles. static profile_t *profilesList[MAXPROFILES+1]; // +1 because we're gonna add a default "GUEST' profile. @@ -28,6 +30,13 @@ INT32 PR_GetNumProfiles(void) return numprofiles; } +static void PR_GenerateProfileKeys(profile_t *new) +{ + static uint8_t seed[32]; + csprng(seed, 32); + crypto_eddsa_key_pair(new->secret_key, new->public_key, seed); +} + profile_t* PR_MakeProfile( const char *prname, const char *pname, @@ -41,6 +50,14 @@ profile_t* PR_MakeProfile( new->version = PROFILEVER; + memset(new->secret_key, 0, sizeof(new->secret_key)); + memset(new->public_key, 0, sizeof(new->public_key)); + + if (!guest) + { + PR_GenerateProfileKeys(new); + } + strcpy(new->profilename, prname); new->profilename[sizeof new->profilename - 1] = '\0'; @@ -238,8 +255,10 @@ void PR_SaveProfiles(void) for (i = 1; i < numprofiles; i++) { - // Names. + // Names and keys, all the string data up front WRITESTRINGN(save.p, profilesList[i]->profilename, PROFILENAMELEN); + WRITEMEM(save.p, profilesList[i]->public_key, sizeof(((profile_t *)0)->public_key)); + WRITEMEM(save.p, profilesList[i]->secret_key, sizeof(((profile_t *)0)->secret_key)); WRITESTRINGN(save.p, profilesList[i]->playername, MAXPLAYERNAME); // Character and colour. @@ -329,8 +348,21 @@ void PR_LoadProfiles(void) // Version. (We always update this on successful forward step) profilesList[i]->version = PROFILEVER; - // Names. + // Names and keys, all the identity stuff up front READSTRINGN(save.p, profilesList[i]->profilename, PROFILENAMELEN); + + // Profile update 2-->3: Add profile keys. + if (version < 3) + { + // Generate missing keys. + PR_GenerateProfileKeys(profilesList[i]); + } + else + { + READMEM(save.p, profilesList[i]->public_key, sizeof(((profile_t *)0)->public_key)); + READMEM(save.p, profilesList[i]->secret_key, sizeof(((profile_t *)0)->secret_key)); + } + READSTRINGN(save.p, profilesList[i]->playername, MAXPLAYERNAME); // Character and colour. @@ -550,3 +582,39 @@ profile_t *PR_GetPlayerProfile(player_t *player) return NULL; } + +profile_t *PR_GetLocalPlayerProfile(INT32 player) +{ + if (player >= MAXSPLITSCREENPLAYERS) + return NULL; + return PR_GetProfile(cv_lastprofile[player].value); +} + +boolean PR_IsLocalPlayerGuest(INT32 player) +{ + return !(cv_lastprofile[player].value); +} + +static char rrid_buf[256]; + +char *GetPrettyRRID(const unsigned char *bin, boolean brief) +{ + size_t i; + size_t len = PUBKEYLENGTH; + + if (brief) + len = 8; + + if (bin == NULL || len == 0) + return NULL; + + for (i=0; i> 4]; + rrid_buf[i*2+1] = "0123456789ABCDEF"[bin[i] & 0x0F]; + } + + rrid_buf[len*2] = '\0'; + + return rrid_buf; +} diff --git a/src/k_profiles.h b/src/k_profiles.h index f0ae7e287..c5e49bab9 100644 --- a/src/k_profiles.h +++ b/src/k_profiles.h @@ -31,7 +31,7 @@ extern "C" { #define SKINNAMESIZE 16 #define PROFILENAMELEN 6 -#define PROFILEVER 2 +#define PROFILEVER 3 #define MAXPROFILES 16 #define PROFILESFILE "ringprofiles.prf" #define PROFILE_GUEST 0 @@ -59,6 +59,9 @@ struct profile_t // Profile header char profilename[PROFILENAMELEN+1]; // Profile name (not to be confused with player name) + uint8_t public_key[PUBKEYLENGTH]; // Netgame authentication + uint8_t secret_key[PRIVKEYLENGTH]; + // Player data char playername[MAXPLAYERNAME+1]; // Player name char skinname[SKINNAMESIZE+1]; // Default Skin @@ -156,6 +159,12 @@ SINT8 PR_ProfileUsedBy(profile_t *p); profile_t *PR_GetPlayerProfile(player_t *player); +profile_t *PR_GetLocalPlayerProfile(INT32 player); + +boolean PR_IsLocalPlayerGuest(INT32 player); + +char *GetPrettyRRID(const unsigned char *bin, boolean brief); + #ifdef __cplusplus } // extern "C" #endif diff --git a/src/lua_playerlib.c b/src/lua_playerlib.c index 76325ebc9..6e93fd974 100644 --- a/src/lua_playerlib.c +++ b/src/lua_playerlib.c @@ -22,6 +22,7 @@ #include "lua_libs.h" #include "lua_hud.h" // hud_running errors #include "lua_hook.h" // hook_cmd_running errors +#include "k_profiles.h" // GetPrettyRRID static int lib_iteratePlayers(lua_State *L) { @@ -507,6 +508,8 @@ static int player_get(lua_State *L) #endif else if (fastcmp(field,"ping")) lua_pushinteger(L, playerpingtable[( plr - players )]); + else if (fastcmp(field, "public_key")) + lua_pushstring(L, GetPrettyRRID(plr->public_key, false)); else { lua_getfield(L, LUA_REGISTRYINDEX, LREG_EXTVARS); I_Assert(lua_istable(L, -1)); @@ -876,7 +879,7 @@ static int player_set(lua_State *L) else if (fastcmp(field,"bot")) return NOSET; else if (fastcmp(field,"jointime")) - plr->jointime = (tic_t)luaL_checkinteger(L, 3); + return NOSET; else if (fastcmp(field,"splitscreenindex")) return NOSET; #ifdef HWRENDER diff --git a/src/monocypher/CMakeLists.txt b/src/monocypher/CMakeLists.txt new file mode 100644 index 000000000..8122e03ba --- /dev/null +++ b/src/monocypher/CMakeLists.txt @@ -0,0 +1,4 @@ +target_sources(SRB2SDL2 PRIVATE +monocypher.c +monocypher.h +) diff --git a/src/monocypher/monocypher.c b/src/monocypher/monocypher.c new file mode 100644 index 000000000..0d0fa5495 --- /dev/null +++ b/src/monocypher/monocypher.c @@ -0,0 +1,2938 @@ +// Monocypher version 4.0.0 +// +// Copyright (c) 2017-2020, Loup Vaillant +// All rights reserved. +// +// +// Redistribution and use in source and binary forms, with or without +// modification, are permitted provided that the following conditions are +// met: +// +// 1. Redistributions of source code must retain the above copyright +// notice, this list of conditions and the following disclaimer. +// +// 2. Redistributions in binary form must reproduce the above copyright +// notice, this list of conditions and the following disclaimer in the +// documentation and/or other materials provided with the +// distribution. +// +// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +// HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +#include "monocypher.h" + +#ifdef MONOCYPHER_CPP_NAMESPACE +namespace MONOCYPHER_CPP_NAMESPACE { +#endif + +///////////////// +/// Utilities /// +///////////////// +#define FOR_T(type, i, start, end) for (type i = (start); i < (end); i++) +#define FOR(i, start, end) FOR_T(size_t, i, start, end) +#define COPY(dst, src, size) FOR(_i_, 0, size) (dst)[_i_] = (src)[_i_] +#define ZERO(buf, size) FOR(_i_, 0, size) (buf)[_i_] = 0 +#define WIPE_CTX(ctx) crypto_wipe(ctx , sizeof(*(ctx))) +#define WIPE_BUFFER(buffer) crypto_wipe(buffer, sizeof(buffer)) +#define MIN(a, b) ((a) <= (b) ? (a) : (b)) +#define MAX(a, b) ((a) >= (b) ? (a) : (b)) + +typedef int8_t i8; +typedef uint8_t u8; +typedef int16_t i16; +typedef uint32_t u32; +typedef int32_t i32; +typedef int64_t i64; +typedef uint64_t u64; + +static const u8 zero[128] = {0}; + +// returns the smallest positive integer y such that +// (x + y) % pow_2 == 0 +// Basically, it's how many bytes we need to add to "align" x. +// Only works when pow_2 is a power of 2. +// Note: we use ~x+1 instead of -x to avoid compiler warnings +static size_t align(size_t x, size_t pow_2) +{ + return (~x + 1) & (pow_2 - 1); +} + +static u32 load24_le(const u8 s[3]) +{ + return + ((u32)s[0] << 0) | + ((u32)s[1] << 8) | + ((u32)s[2] << 16); +} + +static u32 load32_le(const u8 s[4]) +{ + return + ((u32)s[0] << 0) | + ((u32)s[1] << 8) | + ((u32)s[2] << 16) | + ((u32)s[3] << 24); +} + +static u64 load64_le(const u8 s[8]) +{ + return load32_le(s) | ((u64)load32_le(s+4) << 32); +} + +static void store32_le(u8 out[4], u32 in) +{ + out[0] = in & 0xff; + out[1] = (in >> 8) & 0xff; + out[2] = (in >> 16) & 0xff; + out[3] = (in >> 24) & 0xff; +} + +static void store64_le(u8 out[8], u64 in) +{ + store32_le(out , (u32)in ); + store32_le(out + 4, in >> 32); +} + +static void load32_le_buf (u32 *dst, const u8 *src, size_t size) { + FOR(i, 0, size) { dst[i] = load32_le(src + i*4); } +} +static void load64_le_buf (u64 *dst, const u8 *src, size_t size) { + FOR(i, 0, size) { dst[i] = load64_le(src + i*8); } +} +static void store32_le_buf(u8 *dst, const u32 *src, size_t size) { + FOR(i, 0, size) { store32_le(dst + i*4, src[i]); } +} +static void store64_le_buf(u8 *dst, const u64 *src, size_t size) { + FOR(i, 0, size) { store64_le(dst + i*8, src[i]); } +} + +static u64 rotr64(u64 x, u64 n) { return (x >> n) ^ (x << (64 - n)); } +static u32 rotl32(u32 x, u32 n) { return (x << n) ^ (x >> (32 - n)); } + +static int neq0(u64 diff) +{ + // constant time comparison to zero + // return diff != 0 ? -1 : 0 + u64 half = (diff >> 32) | ((u32)diff); + return (1 & ((half - 1) >> 32)) - 1; +} + +static u64 x16(const u8 a[16], const u8 b[16]) +{ + return (load64_le(a + 0) ^ load64_le(b + 0)) + | (load64_le(a + 8) ^ load64_le(b + 8)); +} +static u64 x32(const u8 a[32],const u8 b[32]){return x16(a,b)| x16(a+16, b+16);} +static u64 x64(const u8 a[64],const u8 b[64]){return x32(a,b)| x32(a+32, b+32);} +int crypto_verify16(const u8 a[16], const u8 b[16]){ return neq0(x16(a, b)); } +int crypto_verify32(const u8 a[32], const u8 b[32]){ return neq0(x32(a, b)); } +int crypto_verify64(const u8 a[64], const u8 b[64]){ return neq0(x64(a, b)); } + +void crypto_wipe(void *secret, size_t size) +{ + volatile u8 *v_secret = (u8*)secret; + ZERO(v_secret, size); +} + +///////////////// +/// Chacha 20 /// +///////////////// +#define QUARTERROUND(a, b, c, d) \ + a += b; d = rotl32(d ^ a, 16); \ + c += d; b = rotl32(b ^ c, 12); \ + a += b; d = rotl32(d ^ a, 8); \ + c += d; b = rotl32(b ^ c, 7) + +static void chacha20_rounds(u32 out[16], const u32 in[16]) +{ + // The temporary variables make Chacha20 10% faster. + u32 t0 = in[ 0]; u32 t1 = in[ 1]; u32 t2 = in[ 2]; u32 t3 = in[ 3]; + u32 t4 = in[ 4]; u32 t5 = in[ 5]; u32 t6 = in[ 6]; u32 t7 = in[ 7]; + u32 t8 = in[ 8]; u32 t9 = in[ 9]; u32 t10 = in[10]; u32 t11 = in[11]; + u32 t12 = in[12]; u32 t13 = in[13]; u32 t14 = in[14]; u32 t15 = in[15]; + + FOR (i, 0, 10) { // 20 rounds, 2 rounds per loop. + QUARTERROUND(t0, t4, t8 , t12); // column 0 + QUARTERROUND(t1, t5, t9 , t13); // column 1 + QUARTERROUND(t2, t6, t10, t14); // column 2 + QUARTERROUND(t3, t7, t11, t15); // column 3 + QUARTERROUND(t0, t5, t10, t15); // diagonal 0 + QUARTERROUND(t1, t6, t11, t12); // diagonal 1 + QUARTERROUND(t2, t7, t8 , t13); // diagonal 2 + QUARTERROUND(t3, t4, t9 , t14); // diagonal 3 + } + out[ 0] = t0; out[ 1] = t1; out[ 2] = t2; out[ 3] = t3; + out[ 4] = t4; out[ 5] = t5; out[ 6] = t6; out[ 7] = t7; + out[ 8] = t8; out[ 9] = t9; out[10] = t10; out[11] = t11; + out[12] = t12; out[13] = t13; out[14] = t14; out[15] = t15; +} + +static const u8 *chacha20_constant = (const u8*)"expand 32-byte k"; // 16 bytes + +void crypto_chacha20_h(u8 out[32], const u8 key[32], const u8 in [16]) +{ + u32 block[16]; + load32_le_buf(block , chacha20_constant, 4); + load32_le_buf(block + 4, key , 8); + load32_le_buf(block + 12, in , 4); + + chacha20_rounds(block, block); + + // prevent reversal of the rounds by revealing only half of the buffer. + store32_le_buf(out , block , 4); // constant + store32_le_buf(out+16, block+12, 4); // counter and nonce + WIPE_BUFFER(block); +} + +u64 crypto_chacha20_djb(u8 *cipher_text, const u8 *plain_text, + size_t text_size, const u8 key[32], const u8 nonce[8], + u64 ctr) +{ + u32 input[16]; + load32_le_buf(input , chacha20_constant, 4); + load32_le_buf(input + 4, key , 8); + load32_le_buf(input + 14, nonce , 2); + input[12] = (u32) ctr; + input[13] = (u32)(ctr >> 32); + + // Whole blocks + u32 pool[16]; + size_t nb_blocks = text_size >> 6; + FOR (i, 0, nb_blocks) { + chacha20_rounds(pool, input); + if (plain_text != 0) { + FOR (j, 0, 16) { + u32 p = pool[j] + input[j]; + store32_le(cipher_text, p ^ load32_le(plain_text)); + cipher_text += 4; + plain_text += 4; + } + } else { + FOR (j, 0, 16) { + u32 p = pool[j] + input[j]; + store32_le(cipher_text, p); + cipher_text += 4; + } + } + input[12]++; + if (input[12] == 0) { + input[13]++; + } + } + text_size &= 63; + + // Last (incomplete) block + if (text_size > 0) { + if (plain_text == 0) { + plain_text = zero; + } + chacha20_rounds(pool, input); + u8 tmp[64]; + FOR (i, 0, 16) { + store32_le(tmp + i*4, pool[i] + input[i]); + } + FOR (i, 0, text_size) { + cipher_text[i] = tmp[i] ^ plain_text[i]; + } + WIPE_BUFFER(tmp); + } + ctr = input[12] + ((u64)input[13] << 32) + (text_size > 0); + + WIPE_BUFFER(pool); + WIPE_BUFFER(input); + return ctr; +} + +u32 crypto_chacha20_ietf(u8 *cipher_text, const u8 *plain_text, + size_t text_size, + const u8 key[32], const u8 nonce[12], u32 ctr) +{ + u64 big_ctr = ctr + ((u64)load32_le(nonce) << 32); + return (u32)crypto_chacha20_djb(cipher_text, plain_text, text_size, + key, nonce + 4, big_ctr); +} + +u64 crypto_chacha20_x(u8 *cipher_text, const u8 *plain_text, + size_t text_size, + const u8 key[32], const u8 nonce[24], u64 ctr) +{ + u8 sub_key[32]; + crypto_chacha20_h(sub_key, key, nonce); + ctr = crypto_chacha20_djb(cipher_text, plain_text, text_size, + sub_key, nonce + 16, ctr); + WIPE_BUFFER(sub_key); + return ctr; +} + +///////////////// +/// Poly 1305 /// +///////////////// + +// h = (h + c) * r +// preconditions: +// ctx->h <= 4_ffffffff_ffffffff_ffffffff_ffffffff +// ctx->r <= 0ffffffc_0ffffffc_0ffffffc_0fffffff +// end <= 1 +// Postcondition: +// ctx->h <= 4_ffffffff_ffffffff_ffffffff_ffffffff +static void poly_block(crypto_poly1305_ctx *ctx, const u8 in[16], unsigned end) +{ + u32 s[4]; + load32_le_buf(s, in, 4); + + //- PROOF Poly1305 + //- + //- # Inputs & preconditions + //- ctx->h[0] = u32() + //- ctx->h[1] = u32() + //- ctx->h[2] = u32() + //- ctx->h[3] = u32() + //- ctx->h[4] = u32(limit = 4) + //- + //- ctx->r[0] = u32(limit = 0x0fffffff) + //- ctx->r[1] = u32(limit = 0x0ffffffc) + //- ctx->r[2] = u32(limit = 0x0ffffffc) + //- ctx->r[3] = u32(limit = 0x0ffffffc) + //- + //- s[0] = u32() + //- s[1] = u32() + //- s[2] = u32() + //- s[3] = u32() + //- + //- end = unsigned(limit = 1) + + // s = h + c, without carry propagation + const u64 s0 = ctx->h[0] + (u64)s[0]; // s0 <= 1_fffffffe + const u64 s1 = ctx->h[1] + (u64)s[1]; // s1 <= 1_fffffffe + const u64 s2 = ctx->h[2] + (u64)s[2]; // s2 <= 1_fffffffe + const u64 s3 = ctx->h[3] + (u64)s[3]; // s3 <= 1_fffffffe + const u32 s4 = ctx->h[4] + end; // s4 <= 5 + + // Local all the things! + const u32 r0 = ctx->r[0]; // r0 <= 0fffffff + const u32 r1 = ctx->r[1]; // r1 <= 0ffffffc + const u32 r2 = ctx->r[2]; // r2 <= 0ffffffc + const u32 r3 = ctx->r[3]; // r3 <= 0ffffffc + const u32 rr0 = (r0 >> 2) * 5; // rr0 <= 13fffffb // lose 2 bits... + const u32 rr1 = (r1 >> 2) + r1; // rr1 <= 13fffffb // rr1 == (r1 >> 2) * 5 + const u32 rr2 = (r2 >> 2) + r2; // rr2 <= 13fffffb // rr1 == (r2 >> 2) * 5 + const u32 rr3 = (r3 >> 2) + r3; // rr3 <= 13fffffb // rr1 == (r3 >> 2) * 5 + + // (h + c) * r, without carry propagation + const u64 x0 = s0*r0+ s1*rr3+ s2*rr2+ s3*rr1+ s4*rr0; // <= 97ffffe007fffff8 + const u64 x1 = s0*r1+ s1*r0 + s2*rr3+ s3*rr2+ s4*rr1; // <= 8fffffe20ffffff6 + const u64 x2 = s0*r2+ s1*r1 + s2*r0 + s3*rr3+ s4*rr2; // <= 87ffffe417fffff4 + const u64 x3 = s0*r3+ s1*r2 + s2*r1 + s3*r0 + s4*rr3; // <= 7fffffe61ffffff2 + const u32 x4 = s4 * (r0 & 3); // ...recover 2 bits // <= f + + // partial reduction modulo 2^130 - 5 + const u32 u5 = x4 + (x3 >> 32); // u5 <= 7ffffff5 + const u64 u0 = (u5 >> 2) * 5 + (x0 & 0xffffffff); + const u64 u1 = (u0 >> 32) + (x1 & 0xffffffff) + (x0 >> 32); + const u64 u2 = (u1 >> 32) + (x2 & 0xffffffff) + (x1 >> 32); + const u64 u3 = (u2 >> 32) + (x3 & 0xffffffff) + (x2 >> 32); + const u64 u4 = (u3 >> 32) + (u5 & 3); + + // Update the hash + ctx->h[0] = u0 & 0xffffffff; // u0 <= 1_9ffffff0 + ctx->h[1] = u1 & 0xffffffff; // u1 <= 1_97ffffe0 + ctx->h[2] = u2 & 0xffffffff; // u2 <= 1_8fffffe2 + ctx->h[3] = u3 & 0xffffffff; // u3 <= 1_87ffffe4 + ctx->h[4] = u4 & 0xffffffff; // u4 <= 4 + + //- # postconditions + //- ASSERT(ctx->h[4].limit() <= 4) + //- CQFD Poly1305 +} + +void crypto_poly1305_init(crypto_poly1305_ctx *ctx, const u8 key[32]) +{ + ZERO(ctx->h, 5); // Initial hash is zero + ctx->c_idx = 0; + // load r and pad (r has some of its bits cleared) + load32_le_buf(ctx->r , key , 4); + load32_le_buf(ctx->pad, key+16, 4); + FOR (i, 0, 1) { ctx->r[i] &= 0x0fffffff; } + FOR (i, 1, 4) { ctx->r[i] &= 0x0ffffffc; } +} + +void crypto_poly1305_update(crypto_poly1305_ctx *ctx, + const u8 *message, size_t message_size) +{ + // Align ourselves with block boundaries + size_t aligned = MIN(align(ctx->c_idx, 16), message_size); + FOR (i, 0, aligned) { + ctx->c[ctx->c_idx] = *message; + ctx->c_idx++; + message++; + message_size--; + } + + // If block is complete, process it + if (ctx->c_idx == 16) { + poly_block(ctx, ctx->c, 1); + ctx->c_idx = 0; + } + + // Process the message block by block + size_t nb_blocks = message_size >> 4; + FOR (i, 0, nb_blocks) { + poly_block(ctx, message, 1); + message += 16; + } + message_size &= 15; + + // remaining bytes (we never complete a block here) + FOR (i, 0, message_size) { + ctx->c[ctx->c_idx] = message[i]; + ctx->c_idx++; + } +} + +void crypto_poly1305_final(crypto_poly1305_ctx *ctx, u8 mac[16]) +{ + // Process the last block (if any) + // We move the final 1 according to remaining input length + // (this will add less than 2^130 to the last input block) + if (ctx->c_idx != 0) { + ZERO(ctx->c + ctx->c_idx, 16 - ctx->c_idx); + ctx->c[ctx->c_idx] = 1; + poly_block(ctx, ctx->c, 0); + } + + // check if we should subtract 2^130-5 by performing the + // corresponding carry propagation. + u64 c = 5; + FOR (i, 0, 4) { + c += ctx->h[i]; + c >>= 32; + } + c += ctx->h[4]; + c = (c >> 2) * 5; // shift the carry back to the beginning + // c now indicates how many times we should subtract 2^130-5 (0 or 1) + FOR (i, 0, 4) { + c += (u64)ctx->h[i] + ctx->pad[i]; + store32_le(mac + i*4, (u32)c); + c = c >> 32; + } + WIPE_CTX(ctx); +} + +void crypto_poly1305(u8 mac[16], const u8 *message, + size_t message_size, const u8 key[32]) +{ + crypto_poly1305_ctx ctx; + crypto_poly1305_init (&ctx, key); + crypto_poly1305_update(&ctx, message, message_size); + crypto_poly1305_final (&ctx, mac); +} + +//////////////// +/// BLAKE2 b /// +//////////////// +static const u64 iv[8] = { + 0x6a09e667f3bcc908, 0xbb67ae8584caa73b, + 0x3c6ef372fe94f82b, 0xa54ff53a5f1d36f1, + 0x510e527fade682d1, 0x9b05688c2b3e6c1f, + 0x1f83d9abfb41bd6b, 0x5be0cd19137e2179, +}; + +static void blake2b_compress(crypto_blake2b_ctx *ctx, int is_last_block) +{ + static const u8 sigma[12][16] = { + { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 }, + { 14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3 }, + { 11, 8, 12, 0, 5, 2, 15, 13, 10, 14, 3, 6, 7, 1, 9, 4 }, + { 7, 9, 3, 1, 13, 12, 11, 14, 2, 6, 5, 10, 4, 0, 15, 8 }, + { 9, 0, 5, 7, 2, 4, 10, 15, 14, 1, 11, 12, 6, 8, 3, 13 }, + { 2, 12, 6, 10, 0, 11, 8, 3, 4, 13, 7, 5, 15, 14, 1, 9 }, + { 12, 5, 1, 15, 14, 13, 4, 10, 0, 7, 6, 3, 9, 2, 8, 11 }, + { 13, 11, 7, 14, 12, 1, 3, 9, 5, 0, 15, 4, 8, 6, 2, 10 }, + { 6, 15, 14, 9, 11, 3, 0, 8, 12, 2, 13, 7, 1, 4, 10, 5 }, + { 10, 2, 8, 4, 7, 6, 1, 5, 15, 11, 9, 14, 3, 12, 13, 0 }, + { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 }, + { 14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3 }, + }; + + // increment input offset + u64 *x = ctx->input_offset; + size_t y = ctx->input_idx; + x[0] += y; + if (x[0] < y) { + x[1]++; + } + + // init work vector + u64 v0 = ctx->hash[0]; u64 v8 = iv[0]; + u64 v1 = ctx->hash[1]; u64 v9 = iv[1]; + u64 v2 = ctx->hash[2]; u64 v10 = iv[2]; + u64 v3 = ctx->hash[3]; u64 v11 = iv[3]; + u64 v4 = ctx->hash[4]; u64 v12 = iv[4] ^ ctx->input_offset[0]; + u64 v5 = ctx->hash[5]; u64 v13 = iv[5] ^ ctx->input_offset[1]; + u64 v6 = ctx->hash[6]; u64 v14 = iv[6] ^ (u64)~(is_last_block - 1); + u64 v7 = ctx->hash[7]; u64 v15 = iv[7]; + + // mangle work vector + u64 *input = ctx->input; +#define BLAKE2_G(a, b, c, d, x, y) \ + a += b + x; d = rotr64(d ^ a, 32); \ + c += d; b = rotr64(b ^ c, 24); \ + a += b + y; d = rotr64(d ^ a, 16); \ + c += d; b = rotr64(b ^ c, 63) +#define BLAKE2_ROUND(i) \ + BLAKE2_G(v0, v4, v8 , v12, input[sigma[i][ 0]], input[sigma[i][ 1]]); \ + BLAKE2_G(v1, v5, v9 , v13, input[sigma[i][ 2]], input[sigma[i][ 3]]); \ + BLAKE2_G(v2, v6, v10, v14, input[sigma[i][ 4]], input[sigma[i][ 5]]); \ + BLAKE2_G(v3, v7, v11, v15, input[sigma[i][ 6]], input[sigma[i][ 7]]); \ + BLAKE2_G(v0, v5, v10, v15, input[sigma[i][ 8]], input[sigma[i][ 9]]); \ + BLAKE2_G(v1, v6, v11, v12, input[sigma[i][10]], input[sigma[i][11]]); \ + BLAKE2_G(v2, v7, v8 , v13, input[sigma[i][12]], input[sigma[i][13]]); \ + BLAKE2_G(v3, v4, v9 , v14, input[sigma[i][14]], input[sigma[i][15]]) + +#ifdef BLAKE2_NO_UNROLLING + FOR (i, 0, 12) { + BLAKE2_ROUND(i); + } +#else + BLAKE2_ROUND(0); BLAKE2_ROUND(1); BLAKE2_ROUND(2); BLAKE2_ROUND(3); + BLAKE2_ROUND(4); BLAKE2_ROUND(5); BLAKE2_ROUND(6); BLAKE2_ROUND(7); + BLAKE2_ROUND(8); BLAKE2_ROUND(9); BLAKE2_ROUND(10); BLAKE2_ROUND(11); +#endif + + // update hash + ctx->hash[0] ^= v0 ^ v8; ctx->hash[1] ^= v1 ^ v9; + ctx->hash[2] ^= v2 ^ v10; ctx->hash[3] ^= v3 ^ v11; + ctx->hash[4] ^= v4 ^ v12; ctx->hash[5] ^= v5 ^ v13; + ctx->hash[6] ^= v6 ^ v14; ctx->hash[7] ^= v7 ^ v15; +} + +void crypto_blake2b_keyed_init(crypto_blake2b_ctx *ctx, size_t hash_size, + const u8 *key, size_t key_size) +{ + // initial hash + COPY(ctx->hash, iv, 8); + ctx->hash[0] ^= 0x01010000 ^ (key_size << 8) ^ hash_size; + + ctx->input_offset[0] = 0; // beginning of the input, no offset + ctx->input_offset[1] = 0; // beginning of the input, no offset + ctx->hash_size = hash_size; + ctx->input_idx = 0; + ZERO(ctx->input, 16); + + // if there is a key, the first block is that key (padded with zeroes) + if (key_size > 0) { + u8 key_block[128] = {0}; + COPY(key_block, key, key_size); + // same as calling crypto_blake2b_update(ctx, key_block , 128) + load64_le_buf(ctx->input, key_block, 16); + ctx->input_idx = 128; + } +} + +void crypto_blake2b_init(crypto_blake2b_ctx *ctx, size_t hash_size) +{ + crypto_blake2b_keyed_init(ctx, hash_size, 0, 0); +} + +void crypto_blake2b_update(crypto_blake2b_ctx *ctx, + const u8 *message, size_t message_size) +{ + // Avoid undefined NULL pointer increments with empty messages + if (message_size == 0) { + return; + } + + // Align with word boundaries + if ((ctx->input_idx & 7) != 0) { + size_t nb_bytes = MIN(align(ctx->input_idx, 8), message_size); + size_t word = ctx->input_idx >> 3; + size_t byte = ctx->input_idx & 7; + FOR (i, 0, nb_bytes) { + ctx->input[word] |= (u64)message[i] << ((byte + i) << 3); + } + ctx->input_idx += nb_bytes; + message += nb_bytes; + message_size -= nb_bytes; + } + + // Align with block boundaries (faster than byte by byte) + if ((ctx->input_idx & 127) != 0) { + size_t nb_words = MIN(align(ctx->input_idx, 128), message_size) >> 3; + load64_le_buf(ctx->input + (ctx->input_idx >> 3), message, nb_words); + ctx->input_idx += nb_words << 3; + message += nb_words << 3; + message_size -= nb_words << 3; + } + + // Process block by block + size_t nb_blocks = message_size >> 7; + FOR (i, 0, nb_blocks) { + if (ctx->input_idx == 128) { + blake2b_compress(ctx, 0); + } + load64_le_buf(ctx->input, message, 16); + message += 128; + ctx->input_idx = 128; + } + message_size &= 127; + + if (message_size != 0) { + // Compress block & flush input buffer as needed + if (ctx->input_idx == 128) { + blake2b_compress(ctx, 0); + ctx->input_idx = 0; + } + if (ctx->input_idx == 0) { + ZERO(ctx->input, 16); + } + // Fill remaining words (faster than byte by byte) + size_t nb_words = message_size >> 3; + load64_le_buf(ctx->input, message, nb_words); + ctx->input_idx += nb_words << 3; + message += nb_words << 3; + message_size -= nb_words << 3; + + // Fill remaining bytes + FOR (i, 0, message_size) { + size_t word = ctx->input_idx >> 3; + size_t byte = ctx->input_idx & 7; + ctx->input[word] |= (u64)message[i] << (byte << 3); + ctx->input_idx++; + } + } +} + +void crypto_blake2b_final(crypto_blake2b_ctx *ctx, u8 *hash) +{ + blake2b_compress(ctx, 1); // compress the last block + size_t hash_size = MIN(ctx->hash_size, 64); + size_t nb_words = hash_size >> 3; + store64_le_buf(hash, ctx->hash, nb_words); + FOR (i, nb_words << 3, hash_size) { + hash[i] = (ctx->hash[i >> 3] >> (8 * (i & 7))) & 0xff; + } + WIPE_CTX(ctx); +} + +void crypto_blake2b_keyed(u8 *hash, size_t hash_size, + const u8 *key, size_t key_size, + const u8 *message, size_t message_size) +{ + crypto_blake2b_ctx ctx; + crypto_blake2b_keyed_init(&ctx, hash_size, key, key_size); + crypto_blake2b_update (&ctx, message, message_size); + crypto_blake2b_final (&ctx, hash); +} + +void crypto_blake2b(u8 *hash, size_t hash_size, const u8 *msg, size_t msg_size) +{ + crypto_blake2b_keyed(hash, hash_size, 0, 0, msg, msg_size); +} + +////////////// +/// Argon2 /// +////////////// +// references to R, Z, Q etc. come from the spec + +// Argon2 operates on 1024 byte blocks. +typedef struct { u64 a[128]; } blk; + +// updates a BLAKE2 hash with a 32 bit word, little endian. +static void blake_update_32(crypto_blake2b_ctx *ctx, u32 input) +{ + u8 buf[4]; + store32_le(buf, input); + crypto_blake2b_update(ctx, buf, 4); + WIPE_BUFFER(buf); +} + +static void blake_update_32_buf(crypto_blake2b_ctx *ctx, + const u8 *buf, u32 size) +{ + blake_update_32(ctx, size); + crypto_blake2b_update(ctx, buf, size); +} + + +static void copy_block(blk *o,const blk*in){FOR(i, 0, 128) o->a[i] = in->a[i];} +static void xor_block(blk *o,const blk*in){FOR(i, 0, 128) o->a[i] ^= in->a[i];} + +// Hash with a virtually unlimited digest size. +// Doesn't extract more entropy than the base hash function. +// Mainly used for filling a whole kilobyte block with pseudo-random bytes. +// (One could use a stream cipher with a seed hash as the key, but +// this would introduce another dependency —and point of failure.) +static void extended_hash(u8 *digest, u32 digest_size, + const u8 *input , u32 input_size) +{ + crypto_blake2b_ctx ctx; + crypto_blake2b_init (&ctx, MIN(digest_size, 64)); + blake_update_32 (&ctx, digest_size); + crypto_blake2b_update(&ctx, input, input_size); + crypto_blake2b_final (&ctx, digest); + + if (digest_size > 64) { + // the conversion to u64 avoids integer overflow on + // ludicrously big hash sizes. + u32 r = (u32)(((u64)digest_size + 31) >> 5) - 2; + u32 i = 1; + u32 in = 0; + u32 out = 32; + while (i < r) { + // Input and output overlap. This is intentional + crypto_blake2b(digest + out, 64, digest + in, 64); + i += 1; + in += 32; + out += 32; + } + crypto_blake2b(digest + out, digest_size - (32 * r), digest + in , 64); + } +} + +#define LSB(x) ((x) & 0xffffffff) +#define G(a, b, c, d) \ + a += b + 2 * LSB(a) * LSB(b); d ^= a; d = rotr64(d, 32); \ + c += d + 2 * LSB(c) * LSB(d); b ^= c; b = rotr64(b, 24); \ + a += b + 2 * LSB(a) * LSB(b); d ^= a; d = rotr64(d, 16); \ + c += d + 2 * LSB(c) * LSB(d); b ^= c; b = rotr64(b, 63) +#define ROUND(v0, v1, v2, v3, v4, v5, v6, v7, \ + v8, v9, v10, v11, v12, v13, v14, v15) \ + G(v0, v4, v8, v12); G(v1, v5, v9, v13); \ + G(v2, v6, v10, v14); G(v3, v7, v11, v15); \ + G(v0, v5, v10, v15); G(v1, v6, v11, v12); \ + G(v2, v7, v8, v13); G(v3, v4, v9, v14) + +// Core of the compression function G. Computes Z from R in place. +static void g_rounds(blk *b) +{ + // column rounds (work_block = Q) + for (int i = 0; i < 128; i += 16) { + ROUND(b->a[i ], b->a[i+ 1], b->a[i+ 2], b->a[i+ 3], + b->a[i+ 4], b->a[i+ 5], b->a[i+ 6], b->a[i+ 7], + b->a[i+ 8], b->a[i+ 9], b->a[i+10], b->a[i+11], + b->a[i+12], b->a[i+13], b->a[i+14], b->a[i+15]); + } + // row rounds (b = Z) + for (int i = 0; i < 16; i += 2) { + ROUND(b->a[i ], b->a[i+ 1], b->a[i+ 16], b->a[i+ 17], + b->a[i+32], b->a[i+33], b->a[i+ 48], b->a[i+ 49], + b->a[i+64], b->a[i+65], b->a[i+ 80], b->a[i+ 81], + b->a[i+96], b->a[i+97], b->a[i+112], b->a[i+113]); + } +} + +const crypto_argon2_extras crypto_argon2_no_extras = { 0, 0, 0, 0 }; + +void crypto_argon2(u8 *hash, u32 hash_size, void *work_area, + crypto_argon2_config config, + crypto_argon2_inputs inputs, + crypto_argon2_extras extras) +{ + const u32 segment_size = config.nb_blocks / config.nb_lanes / 4; + const u32 lane_size = segment_size * 4; + const u32 nb_blocks = lane_size * config.nb_lanes; // rounding down + + // work area seen as blocks (must be suitably aligned) + blk *blocks = (blk*)work_area; + { + u8 initial_hash[72]; // 64 bytes plus 2 words for future hashes + crypto_blake2b_ctx ctx; + crypto_blake2b_init (&ctx, 64); + blake_update_32 (&ctx, config.nb_lanes ); // p: number of "threads" + blake_update_32 (&ctx, hash_size); + blake_update_32 (&ctx, config.nb_blocks); + blake_update_32 (&ctx, config.nb_passes); + blake_update_32 (&ctx, 0x13); // v: version number + blake_update_32 (&ctx, config.algorithm); // y: Argon2i, Argon2d... + blake_update_32_buf (&ctx, inputs.pass, inputs.pass_size); + blake_update_32_buf (&ctx, inputs.salt, inputs.salt_size); + blake_update_32_buf (&ctx, extras.key, extras.key_size); + blake_update_32_buf (&ctx, extras.ad, extras.ad_size); + crypto_blake2b_final(&ctx, initial_hash); // fill 64 first bytes only + + // fill first 2 blocks of each lane + u8 hash_area[1024]; + FOR_T(u32, l, 0, config.nb_lanes) { + FOR_T(u32, i, 0, 2) { + store32_le(initial_hash + 64, i); // first additional word + store32_le(initial_hash + 68, l); // second additional word + extended_hash(hash_area, 1024, initial_hash, 72); + load64_le_buf(blocks[l * lane_size + i].a, hash_area, 128); + } + } + + WIPE_BUFFER(initial_hash); + WIPE_BUFFER(hash_area); + } + + // Argon2i and Argon2id start with constant time indexing + int constant_time = config.algorithm != CRYPTO_ARGON2_D; + + // Fill (and re-fill) the rest of the blocks + // + // Note: even though each segment within the same slice can be + // computed in parallel, (one thread per lane), we are computing + // them sequentially, because Monocypher doesn't support threads. + // + // Yet optimal performance (and therefore security) requires one + // thread per lane. The only reason Monocypher supports multiple + // lanes is compatibility. + blk tmp; + FOR_T(u32, pass, 0, config.nb_passes) { + FOR_T(u32, slice, 0, 4) { + // On the first slice of the first pass, + // blocks 0 and 1 are already filled, hence pass_offset. + u32 pass_offset = pass == 0 && slice == 0 ? 2 : 0; + u32 slice_offset = slice * segment_size; + + // Argon2id switches back to non-constant time indexing + // after the first two slices of the first pass + if (slice == 2 && config.algorithm == CRYPTO_ARGON2_ID) { + constant_time = 0; + } + + // Each iteration of the following loop may be performed in + // a separate thread. All segments must be fully completed + // before we start filling the next slice. + FOR_T(u32, segment, 0, config.nb_lanes) { + blk index_block; + u32 index_ctr = 1; + FOR_T (u32, block, pass_offset, segment_size) { + // Current and previous blocks + u32 lane_offset = segment * lane_size; + blk *segment_start = blocks + lane_offset + slice_offset; + blk *current = segment_start + block; + blk *previous = + block == 0 && slice_offset == 0 + ? segment_start + lane_size - 1 + : segment_start + block - 1; + + u64 index_seed; + if (constant_time) { + if (block == pass_offset || (block % 128) == 0) { + // Fill or refresh deterministic indices block + + // seed the beginning of the block... + ZERO(index_block.a, 128); + index_block.a[0] = pass; + index_block.a[1] = segment; + index_block.a[2] = slice; + index_block.a[3] = nb_blocks; + index_block.a[4] = config.nb_passes; + index_block.a[5] = config.algorithm; + index_block.a[6] = index_ctr; + index_ctr++; + + // ... then shuffle it + copy_block(&tmp, &index_block); + g_rounds (&index_block); + xor_block (&index_block, &tmp); + copy_block(&tmp, &index_block); + g_rounds (&index_block); + xor_block (&index_block, &tmp); + } + index_seed = index_block.a[block % 128]; + } else { + index_seed = previous->a[0]; + } + + // Establish the reference set. *Approximately* comprises: + // - The last 3 slices (if they exist yet) + // - The already constructed blocks in the current segment + u32 next_slice = ((slice + 1) % 4) * segment_size; + u32 window_start = pass == 0 ? 0 : next_slice; + u32 nb_segments = pass == 0 ? slice : 3; + u32 window_size = nb_segments * segment_size + block - 1; + + // Find reference block + u64 j1 = index_seed & 0xffffffff; // block selector + u64 j2 = index_seed >> 32; // lane selector + u64 x = (j1 * j1) >> 32; + u64 y = (window_size * x) >> 32; + u64 z = (window_size - 1) - y; + u64 ref = (window_start + z) % lane_size; + u32 index = (j2%config.nb_lanes)*lane_size + (u32)ref; + blk *reference = blocks + index; + + // Shuffle the previous & reference block + // into the current block + copy_block(&tmp, previous); + xor_block (&tmp, reference); + if (pass == 0) { copy_block(current, &tmp); } + else { xor_block (current, &tmp); } + g_rounds (&tmp); + xor_block (current, &tmp); + } + } + } + } + + // Wipe temporary block + volatile u64* p = tmp.a; + ZERO(p, 128); + + // XOR last blocks of each lane + blk *last_block = blocks + lane_size - 1; + FOR_T (u32, lane, 1, config.nb_lanes) { + blk *next_block = last_block + lane_size; + xor_block(next_block, last_block); + last_block = next_block; + } + + // Serialize last block + u8 final_block[1024]; + store64_le_buf(final_block, last_block->a, 128); + + // Wipe work area + p = (u64*)work_area; + ZERO(p, 128 * nb_blocks); + + // Hash the very last block with H' into the output hash + extended_hash(hash, hash_size, final_block, 1024); + WIPE_BUFFER(final_block); +} + +//////////////////////////////////// +/// Arithmetic modulo 2^255 - 19 /// +//////////////////////////////////// +// Originally taken from SUPERCOP's ref10 implementation. +// A bit bigger than TweetNaCl, over 4 times faster. + +// field element +typedef i32 fe[10]; + +// field constants +// +// fe_one : 1 +// sqrtm1 : sqrt(-1) +// d : -121665 / 121666 +// D2 : 2 * -121665 / 121666 +// lop_x, lop_y: low order point in Edwards coordinates +// ufactor : -sqrt(-1) * 2 +// A2 : 486662^2 (A squared) +static const fe fe_one = {1}; +static const fe sqrtm1 = { + -32595792, -7943725, 9377950, 3500415, 12389472, + -272473, -25146209, -2005654, 326686, 11406482, +}; +static const fe d = { + -10913610, 13857413, -15372611, 6949391, 114729, + -8787816, -6275908, -3247719, -18696448, -12055116, +}; +static const fe D2 = { + -21827239, -5839606, -30745221, 13898782, 229458, + 15978800, -12551817, -6495438, 29715968, 9444199, +}; +static const fe lop_x = { + 21352778, 5345713, 4660180, -8347857, 24143090, + 14568123, 30185756, -12247770, -33528939, 8345319, +}; +static const fe lop_y = { + -6952922, -1265500, 6862341, -7057498, -4037696, + -5447722, 31680899, -15325402, -19365852, 1569102, +}; +static const fe ufactor = { + -1917299, 15887451, -18755900, -7000830, -24778944, + 544946, -16816446, 4011309, -653372, 10741468, +}; +static const fe A2 = { + 12721188, 3529, 0, 0, 0, 0, 0, 0, 0, 0, +}; + +static void fe_0(fe h) { ZERO(h , 10); } +static void fe_1(fe h) { h[0] = 1; ZERO(h+1, 9); } + +static void fe_copy(fe h,const fe f ){FOR(i,0,10) h[i] = f[i]; } +static void fe_neg (fe h,const fe f ){FOR(i,0,10) h[i] = -f[i]; } +static void fe_add (fe h,const fe f,const fe g){FOR(i,0,10) h[i] = f[i] + g[i];} +static void fe_sub (fe h,const fe f,const fe g){FOR(i,0,10) h[i] = f[i] - g[i];} + +static void fe_cswap(fe f, fe g, int b) +{ + i32 mask = -b; // -1 = 0xffffffff + FOR (i, 0, 10) { + i32 x = (f[i] ^ g[i]) & mask; + f[i] = f[i] ^ x; + g[i] = g[i] ^ x; + } +} + +static void fe_ccopy(fe f, const fe g, int b) +{ + i32 mask = -b; // -1 = 0xffffffff + FOR (i, 0, 10) { + i32 x = (f[i] ^ g[i]) & mask; + f[i] = f[i] ^ x; + } +} + + +// Signed carry propagation +// ------------------------ +// +// Let t be a number. It can be uniquely decomposed thus: +// +// t = h*2^26 + l +// such that -2^25 <= l < 2^25 +// +// Let c = (t + 2^25) / 2^26 (rounded down) +// c = (h*2^26 + l + 2^25) / 2^26 (rounded down) +// c = h + (l + 2^25) / 2^26 (rounded down) +// c = h (exactly) +// Because 0 <= l + 2^25 < 2^26 +// +// Let u = t - c*2^26 +// u = h*2^26 + l - h*2^26 +// u = l +// Therefore, -2^25 <= u < 2^25 +// +// Additionally, if |t| < x, then |h| < x/2^26 (rounded down) +// +// Notations: +// - In C, 1<<25 means 2^25. +// - In C, x>>25 means floor(x / (2^25)). +// - All of the above applies with 25 & 24 as well as 26 & 25. +// +// +// Note on negative right shifts +// ----------------------------- +// +// In C, x >> n, where x is a negative integer, is implementation +// defined. In practice, all platforms do arithmetic shift, which is +// equivalent to division by 2^26, rounded down. Some compilers, like +// GCC, even guarantee it. +// +// If we ever stumble upon a platform that does not propagate the sign +// bit (we won't), visible failures will show at the slightest test, and +// the signed shifts can be replaced by the following: +// +// typedef struct { i64 x:39; } s25; +// typedef struct { i64 x:38; } s26; +// i64 shift25(i64 x) { s25 s; s.x = ((u64)x)>>25; return s.x; } +// i64 shift26(i64 x) { s26 s; s.x = ((u64)x)>>26; return s.x; } +// +// Current compilers cannot optimise this, causing a 30% drop in +// performance. Fairly expensive for something that never happens. +// +// +// Precondition +// ------------ +// +// |t0| < 2^63 +// |t1|..|t9| < 2^62 +// +// Algorithm +// --------- +// c = t0 + 2^25 / 2^26 -- |c| <= 2^36 +// t0 -= c * 2^26 -- |t0| <= 2^25 +// t1 += c -- |t1| <= 2^63 +// +// c = t4 + 2^25 / 2^26 -- |c| <= 2^36 +// t4 -= c * 2^26 -- |t4| <= 2^25 +// t5 += c -- |t5| <= 2^63 +// +// c = t1 + 2^24 / 2^25 -- |c| <= 2^38 +// t1 -= c * 2^25 -- |t1| <= 2^24 +// t2 += c -- |t2| <= 2^63 +// +// c = t5 + 2^24 / 2^25 -- |c| <= 2^38 +// t5 -= c * 2^25 -- |t5| <= 2^24 +// t6 += c -- |t6| <= 2^63 +// +// c = t2 + 2^25 / 2^26 -- |c| <= 2^37 +// t2 -= c * 2^26 -- |t2| <= 2^25 < 1.1 * 2^25 (final t2) +// t3 += c -- |t3| <= 2^63 +// +// c = t6 + 2^25 / 2^26 -- |c| <= 2^37 +// t6 -= c * 2^26 -- |t6| <= 2^25 < 1.1 * 2^25 (final t6) +// t7 += c -- |t7| <= 2^63 +// +// c = t3 + 2^24 / 2^25 -- |c| <= 2^38 +// t3 -= c * 2^25 -- |t3| <= 2^24 < 1.1 * 2^24 (final t3) +// t4 += c -- |t4| <= 2^25 + 2^38 < 2^39 +// +// c = t7 + 2^24 / 2^25 -- |c| <= 2^38 +// t7 -= c * 2^25 -- |t7| <= 2^24 < 1.1 * 2^24 (final t7) +// t8 += c -- |t8| <= 2^63 +// +// c = t4 + 2^25 / 2^26 -- |c| <= 2^13 +// t4 -= c * 2^26 -- |t4| <= 2^25 < 1.1 * 2^25 (final t4) +// t5 += c -- |t5| <= 2^24 + 2^13 < 1.1 * 2^24 (final t5) +// +// c = t8 + 2^25 / 2^26 -- |c| <= 2^37 +// t8 -= c * 2^26 -- |t8| <= 2^25 < 1.1 * 2^25 (final t8) +// t9 += c -- |t9| <= 2^63 +// +// c = t9 + 2^24 / 2^25 -- |c| <= 2^38 +// t9 -= c * 2^25 -- |t9| <= 2^24 < 1.1 * 2^24 (final t9) +// t0 += c * 19 -- |t0| <= 2^25 + 2^38*19 < 2^44 +// +// c = t0 + 2^25 / 2^26 -- |c| <= 2^18 +// t0 -= c * 2^26 -- |t0| <= 2^25 < 1.1 * 2^25 (final t0) +// t1 += c -- |t1| <= 2^24 + 2^18 < 1.1 * 2^24 (final t1) +// +// Postcondition +// ------------- +// |t0|, |t2|, |t4|, |t6|, |t8| < 1.1 * 2^25 +// |t1|, |t3|, |t5|, |t7|, |t9| < 1.1 * 2^24 +#define FE_CARRY \ + i64 c; \ + c = (t0 + ((i64)1<<25)) >> 26; t0 -= c * ((i64)1 << 26); t1 += c; \ + c = (t4 + ((i64)1<<25)) >> 26; t4 -= c * ((i64)1 << 26); t5 += c; \ + c = (t1 + ((i64)1<<24)) >> 25; t1 -= c * ((i64)1 << 25); t2 += c; \ + c = (t5 + ((i64)1<<24)) >> 25; t5 -= c * ((i64)1 << 25); t6 += c; \ + c = (t2 + ((i64)1<<25)) >> 26; t2 -= c * ((i64)1 << 26); t3 += c; \ + c = (t6 + ((i64)1<<25)) >> 26; t6 -= c * ((i64)1 << 26); t7 += c; \ + c = (t3 + ((i64)1<<24)) >> 25; t3 -= c * ((i64)1 << 25); t4 += c; \ + c = (t7 + ((i64)1<<24)) >> 25; t7 -= c * ((i64)1 << 25); t8 += c; \ + c = (t4 + ((i64)1<<25)) >> 26; t4 -= c * ((i64)1 << 26); t5 += c; \ + c = (t8 + ((i64)1<<25)) >> 26; t8 -= c * ((i64)1 << 26); t9 += c; \ + c = (t9 + ((i64)1<<24)) >> 25; t9 -= c * ((i64)1 << 25); t0 += c * 19; \ + c = (t0 + ((i64)1<<25)) >> 26; t0 -= c * ((i64)1 << 26); t1 += c; \ + h[0]=(i32)t0; h[1]=(i32)t1; h[2]=(i32)t2; h[3]=(i32)t3; h[4]=(i32)t4; \ + h[5]=(i32)t5; h[6]=(i32)t6; h[7]=(i32)t7; h[8]=(i32)t8; h[9]=(i32)t9 + +// Decodes a field element from a byte buffer. +// mask specifies how many bits we ignore. +// Traditionally we ignore 1. It's useful for EdDSA, +// which uses that bit to denote the sign of x. +// Elligator however uses positive representatives, +// which means ignoring 2 bits instead. +static void fe_frombytes_mask(fe h, const u8 s[32], unsigned nb_mask) +{ + u32 mask = 0xffffff >> nb_mask; + i64 t0 = load32_le(s); // t0 < 2^32 + i64 t1 = load24_le(s + 4) << 6; // t1 < 2^30 + i64 t2 = load24_le(s + 7) << 5; // t2 < 2^29 + i64 t3 = load24_le(s + 10) << 3; // t3 < 2^27 + i64 t4 = load24_le(s + 13) << 2; // t4 < 2^26 + i64 t5 = load32_le(s + 16); // t5 < 2^32 + i64 t6 = load24_le(s + 20) << 7; // t6 < 2^31 + i64 t7 = load24_le(s + 23) << 5; // t7 < 2^29 + i64 t8 = load24_le(s + 26) << 4; // t8 < 2^28 + i64 t9 = (load24_le(s + 29) & mask) << 2; // t9 < 2^25 + FE_CARRY; // Carry precondition OK +} + +static void fe_frombytes(fe h, const u8 s[32]) +{ + fe_frombytes_mask(h, s, 1); +} + + +// Precondition +// |h[0]|, |h[2]|, |h[4]|, |h[6]|, |h[8]| < 1.1 * 2^25 +// |h[1]|, |h[3]|, |h[5]|, |h[7]|, |h[9]| < 1.1 * 2^24 +// +// Therefore, |h| < 2^255-19 +// There are two possibilities: +// +// - If h is positive, all we need to do is reduce its individual +// limbs down to their tight positive range. +// - If h is negative, we also need to add 2^255-19 to it. +// Or just remove 19 and chop off any excess bit. +static void fe_tobytes(u8 s[32], const fe h) +{ + i32 t[10]; + COPY(t, h, 10); + i32 q = (19 * t[9] + (((i32) 1) << 24)) >> 25; + // |t9| < 1.1 * 2^24 + // -1.1 * 2^24 < t9 < 1.1 * 2^24 + // -21 * 2^24 < 19 * t9 < 21 * 2^24 + // -2^29 < 19 * t9 + 2^24 < 2^29 + // -2^29 / 2^25 < (19 * t9 + 2^24) / 2^25 < 2^29 / 2^25 + // -16 < (19 * t9 + 2^24) / 2^25 < 16 + FOR (i, 0, 5) { + q += t[2*i ]; q >>= 26; // q = 0 or -1 + q += t[2*i+1]; q >>= 25; // q = 0 or -1 + } + // q = 0 iff h >= 0 + // q = -1 iff h < 0 + // Adding q * 19 to h reduces h to its proper range. + q *= 19; // Shift carry back to the beginning + FOR (i, 0, 5) { + t[i*2 ] += q; q = t[i*2 ] >> 26; t[i*2 ] -= q * ((i32)1 << 26); + t[i*2+1] += q; q = t[i*2+1] >> 25; t[i*2+1] -= q * ((i32)1 << 25); + } + // h is now fully reduced, and q represents the excess bit. + + store32_le(s + 0, ((u32)t[0] >> 0) | ((u32)t[1] << 26)); + store32_le(s + 4, ((u32)t[1] >> 6) | ((u32)t[2] << 19)); + store32_le(s + 8, ((u32)t[2] >> 13) | ((u32)t[3] << 13)); + store32_le(s + 12, ((u32)t[3] >> 19) | ((u32)t[4] << 6)); + store32_le(s + 16, ((u32)t[5] >> 0) | ((u32)t[6] << 25)); + store32_le(s + 20, ((u32)t[6] >> 7) | ((u32)t[7] << 19)); + store32_le(s + 24, ((u32)t[7] >> 13) | ((u32)t[8] << 12)); + store32_le(s + 28, ((u32)t[8] >> 20) | ((u32)t[9] << 6)); + + WIPE_BUFFER(t); +} + +// Precondition +// ------------- +// |f0|, |f2|, |f4|, |f6|, |f8| < 1.65 * 2^26 +// |f1|, |f3|, |f5|, |f7|, |f9| < 1.65 * 2^25 +// +// |g0|, |g2|, |g4|, |g6|, |g8| < 1.65 * 2^26 +// |g1|, |g3|, |g5|, |g7|, |g9| < 1.65 * 2^25 +static void fe_mul_small(fe h, const fe f, i32 g) +{ + i64 t0 = f[0] * (i64) g; i64 t1 = f[1] * (i64) g; + i64 t2 = f[2] * (i64) g; i64 t3 = f[3] * (i64) g; + i64 t4 = f[4] * (i64) g; i64 t5 = f[5] * (i64) g; + i64 t6 = f[6] * (i64) g; i64 t7 = f[7] * (i64) g; + i64 t8 = f[8] * (i64) g; i64 t9 = f[9] * (i64) g; + // |t0|, |t2|, |t4|, |t6|, |t8| < 1.65 * 2^26 * 2^31 < 2^58 + // |t1|, |t3|, |t5|, |t7|, |t9| < 1.65 * 2^25 * 2^31 < 2^57 + + FE_CARRY; // Carry precondition OK +} + +// Precondition +// ------------- +// |f0|, |f2|, |f4|, |f6|, |f8| < 1.65 * 2^26 +// |f1|, |f3|, |f5|, |f7|, |f9| < 1.65 * 2^25 +// +// |g0|, |g2|, |g4|, |g6|, |g8| < 1.65 * 2^26 +// |g1|, |g3|, |g5|, |g7|, |g9| < 1.65 * 2^25 +static void fe_mul(fe h, const fe f, const fe g) +{ + // Everything is unrolled and put in temporary variables. + // We could roll the loop, but that would make curve25519 twice as slow. + i32 f0 = f[0]; i32 f1 = f[1]; i32 f2 = f[2]; i32 f3 = f[3]; i32 f4 = f[4]; + i32 f5 = f[5]; i32 f6 = f[6]; i32 f7 = f[7]; i32 f8 = f[8]; i32 f9 = f[9]; + i32 g0 = g[0]; i32 g1 = g[1]; i32 g2 = g[2]; i32 g3 = g[3]; i32 g4 = g[4]; + i32 g5 = g[5]; i32 g6 = g[6]; i32 g7 = g[7]; i32 g8 = g[8]; i32 g9 = g[9]; + i32 F1 = f1*2; i32 F3 = f3*2; i32 F5 = f5*2; i32 F7 = f7*2; i32 F9 = f9*2; + i32 G1 = g1*19; i32 G2 = g2*19; i32 G3 = g3*19; + i32 G4 = g4*19; i32 G5 = g5*19; i32 G6 = g6*19; + i32 G7 = g7*19; i32 G8 = g8*19; i32 G9 = g9*19; + // |F1|, |F3|, |F5|, |F7|, |F9| < 1.65 * 2^26 + // |G0|, |G2|, |G4|, |G6|, |G8| < 2^31 + // |G1|, |G3|, |G5|, |G7|, |G9| < 2^30 + + i64 t0 = f0*(i64)g0 + F1*(i64)G9 + f2*(i64)G8 + F3*(i64)G7 + f4*(i64)G6 + + F5*(i64)G5 + f6*(i64)G4 + F7*(i64)G3 + f8*(i64)G2 + F9*(i64)G1; + i64 t1 = f0*(i64)g1 + f1*(i64)g0 + f2*(i64)G9 + f3*(i64)G8 + f4*(i64)G7 + + f5*(i64)G6 + f6*(i64)G5 + f7*(i64)G4 + f8*(i64)G3 + f9*(i64)G2; + i64 t2 = f0*(i64)g2 + F1*(i64)g1 + f2*(i64)g0 + F3*(i64)G9 + f4*(i64)G8 + + F5*(i64)G7 + f6*(i64)G6 + F7*(i64)G5 + f8*(i64)G4 + F9*(i64)G3; + i64 t3 = f0*(i64)g3 + f1*(i64)g2 + f2*(i64)g1 + f3*(i64)g0 + f4*(i64)G9 + + f5*(i64)G8 + f6*(i64)G7 + f7*(i64)G6 + f8*(i64)G5 + f9*(i64)G4; + i64 t4 = f0*(i64)g4 + F1*(i64)g3 + f2*(i64)g2 + F3*(i64)g1 + f4*(i64)g0 + + F5*(i64)G9 + f6*(i64)G8 + F7*(i64)G7 + f8*(i64)G6 + F9*(i64)G5; + i64 t5 = f0*(i64)g5 + f1*(i64)g4 + f2*(i64)g3 + f3*(i64)g2 + f4*(i64)g1 + + f5*(i64)g0 + f6*(i64)G9 + f7*(i64)G8 + f8*(i64)G7 + f9*(i64)G6; + i64 t6 = f0*(i64)g6 + F1*(i64)g5 + f2*(i64)g4 + F3*(i64)g3 + f4*(i64)g2 + + F5*(i64)g1 + f6*(i64)g0 + F7*(i64)G9 + f8*(i64)G8 + F9*(i64)G7; + i64 t7 = f0*(i64)g7 + f1*(i64)g6 + f2*(i64)g5 + f3*(i64)g4 + f4*(i64)g3 + + f5*(i64)g2 + f6*(i64)g1 + f7*(i64)g0 + f8*(i64)G9 + f9*(i64)G8; + i64 t8 = f0*(i64)g8 + F1*(i64)g7 + f2*(i64)g6 + F3*(i64)g5 + f4*(i64)g4 + + F5*(i64)g3 + f6*(i64)g2 + F7*(i64)g1 + f8*(i64)g0 + F9*(i64)G9; + i64 t9 = f0*(i64)g9 + f1*(i64)g8 + f2*(i64)g7 + f3*(i64)g6 + f4*(i64)g5 + + f5*(i64)g4 + f6*(i64)g3 + f7*(i64)g2 + f8*(i64)g1 + f9*(i64)g0; + // t0 < 0.67 * 2^61 + // t1 < 0.41 * 2^61 + // t2 < 0.52 * 2^61 + // t3 < 0.32 * 2^61 + // t4 < 0.38 * 2^61 + // t5 < 0.22 * 2^61 + // t6 < 0.23 * 2^61 + // t7 < 0.13 * 2^61 + // t8 < 0.09 * 2^61 + // t9 < 0.03 * 2^61 + + FE_CARRY; // Everything below 2^62, Carry precondition OK +} + +// Precondition +// ------------- +// |f0|, |f2|, |f4|, |f6|, |f8| < 1.65 * 2^26 +// |f1|, |f3|, |f5|, |f7|, |f9| < 1.65 * 2^25 +// +// Note: we could use fe_mul() for this, but this is significantly faster +static void fe_sq(fe h, const fe f) +{ + i32 f0 = f[0]; i32 f1 = f[1]; i32 f2 = f[2]; i32 f3 = f[3]; i32 f4 = f[4]; + i32 f5 = f[5]; i32 f6 = f[6]; i32 f7 = f[7]; i32 f8 = f[8]; i32 f9 = f[9]; + i32 f0_2 = f0*2; i32 f1_2 = f1*2; i32 f2_2 = f2*2; i32 f3_2 = f3*2; + i32 f4_2 = f4*2; i32 f5_2 = f5*2; i32 f6_2 = f6*2; i32 f7_2 = f7*2; + i32 f5_38 = f5*38; i32 f6_19 = f6*19; i32 f7_38 = f7*38; + i32 f8_19 = f8*19; i32 f9_38 = f9*38; + // |f0_2| , |f2_2| , |f4_2| , |f6_2| , |f8_2| < 1.65 * 2^27 + // |f1_2| , |f3_2| , |f5_2| , |f7_2| , |f9_2| < 1.65 * 2^26 + // |f5_38|, |f6_19|, |f7_38|, |f8_19|, |f9_38| < 2^31 + + i64 t0 = f0 *(i64)f0 + f1_2*(i64)f9_38 + f2_2*(i64)f8_19 + + f3_2*(i64)f7_38 + f4_2*(i64)f6_19 + f5 *(i64)f5_38; + i64 t1 = f0_2*(i64)f1 + f2 *(i64)f9_38 + f3_2*(i64)f8_19 + + f4 *(i64)f7_38 + f5_2*(i64)f6_19; + i64 t2 = f0_2*(i64)f2 + f1_2*(i64)f1 + f3_2*(i64)f9_38 + + f4_2*(i64)f8_19 + f5_2*(i64)f7_38 + f6 *(i64)f6_19; + i64 t3 = f0_2*(i64)f3 + f1_2*(i64)f2 + f4 *(i64)f9_38 + + f5_2*(i64)f8_19 + f6 *(i64)f7_38; + i64 t4 = f0_2*(i64)f4 + f1_2*(i64)f3_2 + f2 *(i64)f2 + + f5_2*(i64)f9_38 + f6_2*(i64)f8_19 + f7 *(i64)f7_38; + i64 t5 = f0_2*(i64)f5 + f1_2*(i64)f4 + f2_2*(i64)f3 + + f6 *(i64)f9_38 + f7_2*(i64)f8_19; + i64 t6 = f0_2*(i64)f6 + f1_2*(i64)f5_2 + f2_2*(i64)f4 + + f3_2*(i64)f3 + f7_2*(i64)f9_38 + f8 *(i64)f8_19; + i64 t7 = f0_2*(i64)f7 + f1_2*(i64)f6 + f2_2*(i64)f5 + + f3_2*(i64)f4 + f8 *(i64)f9_38; + i64 t8 = f0_2*(i64)f8 + f1_2*(i64)f7_2 + f2_2*(i64)f6 + + f3_2*(i64)f5_2 + f4 *(i64)f4 + f9 *(i64)f9_38; + i64 t9 = f0_2*(i64)f9 + f1_2*(i64)f8 + f2_2*(i64)f7 + + f3_2*(i64)f6 + f4 *(i64)f5_2; + // t0 < 0.67 * 2^61 + // t1 < 0.41 * 2^61 + // t2 < 0.52 * 2^61 + // t3 < 0.32 * 2^61 + // t4 < 0.38 * 2^61 + // t5 < 0.22 * 2^61 + // t6 < 0.23 * 2^61 + // t7 < 0.13 * 2^61 + // t8 < 0.09 * 2^61 + // t9 < 0.03 * 2^61 + + FE_CARRY; +} + +// Parity check. Returns 0 if even, 1 if odd +static int fe_isodd(const fe f) +{ + u8 s[32]; + fe_tobytes(s, f); + u8 isodd = s[0] & 1; + WIPE_BUFFER(s); + return isodd; +} + +// Returns 1 if equal, 0 if not equal +static int fe_isequal(const fe f, const fe g) +{ + u8 fs[32]; + u8 gs[32]; + fe_tobytes(fs, f); + fe_tobytes(gs, g); + int isdifferent = crypto_verify32(fs, gs); + WIPE_BUFFER(fs); + WIPE_BUFFER(gs); + return 1 + isdifferent; +} + +// Inverse square root. +// Returns true if x is a square, false otherwise. +// After the call: +// isr = sqrt(1/x) if x is a non-zero square. +// isr = sqrt(sqrt(-1)/x) if x is not a square. +// isr = 0 if x is zero. +// We do not guarantee the sign of the square root. +// +// Notes: +// Let quartic = x^((p-1)/4) +// +// x^((p-1)/2) = chi(x) +// quartic^2 = chi(x) +// quartic = sqrt(chi(x)) +// quartic = 1 or -1 or sqrt(-1) or -sqrt(-1) +// +// Note that x is a square if quartic is 1 or -1 +// There are 4 cases to consider: +// +// if quartic = 1 (x is a square) +// then x^((p-1)/4) = 1 +// x^((p-5)/4) * x = 1 +// x^((p-5)/4) = 1/x +// x^((p-5)/8) = sqrt(1/x) or -sqrt(1/x) +// +// if quartic = -1 (x is a square) +// then x^((p-1)/4) = -1 +// x^((p-5)/4) * x = -1 +// x^((p-5)/4) = -1/x +// x^((p-5)/8) = sqrt(-1) / sqrt(x) +// x^((p-5)/8) * sqrt(-1) = sqrt(-1)^2 / sqrt(x) +// x^((p-5)/8) * sqrt(-1) = -1/sqrt(x) +// x^((p-5)/8) * sqrt(-1) = -sqrt(1/x) or sqrt(1/x) +// +// if quartic = sqrt(-1) (x is not a square) +// then x^((p-1)/4) = sqrt(-1) +// x^((p-5)/4) * x = sqrt(-1) +// x^((p-5)/4) = sqrt(-1)/x +// x^((p-5)/8) = sqrt(sqrt(-1)/x) or -sqrt(sqrt(-1)/x) +// +// Note that the product of two non-squares is always a square: +// For any non-squares a and b, chi(a) = -1 and chi(b) = -1. +// Since chi(x) = x^((p-1)/2), chi(a)*chi(b) = chi(a*b) = 1. +// Therefore a*b is a square. +// +// Since sqrt(-1) and x are both non-squares, their product is a +// square, and we can compute their square root. +// +// if quartic = -sqrt(-1) (x is not a square) +// then x^((p-1)/4) = -sqrt(-1) +// x^((p-5)/4) * x = -sqrt(-1) +// x^((p-5)/4) = -sqrt(-1)/x +// x^((p-5)/8) = sqrt(-sqrt(-1)/x) +// x^((p-5)/8) = sqrt( sqrt(-1)/x) * sqrt(-1) +// x^((p-5)/8) * sqrt(-1) = sqrt( sqrt(-1)/x) * sqrt(-1)^2 +// x^((p-5)/8) * sqrt(-1) = sqrt( sqrt(-1)/x) * -1 +// x^((p-5)/8) * sqrt(-1) = -sqrt(sqrt(-1)/x) or sqrt(sqrt(-1)/x) +static int invsqrt(fe isr, const fe x) +{ + fe t0, t1, t2; + + // t0 = x^((p-5)/8) + // Can be achieved with a simple double & add ladder, + // but it would be slower. + fe_sq(t0, x); + fe_sq(t1,t0); fe_sq(t1, t1); fe_mul(t1, x, t1); + fe_mul(t0, t0, t1); + fe_sq(t0, t0); fe_mul(t0, t1, t0); + fe_sq(t1, t0); FOR (i, 1, 5) fe_sq(t1, t1); fe_mul(t0, t1, t0); + fe_sq(t1, t0); FOR (i, 1, 10) fe_sq(t1, t1); fe_mul(t1, t1, t0); + fe_sq(t2, t1); FOR (i, 1, 20) fe_sq(t2, t2); fe_mul(t1, t2, t1); + fe_sq(t1, t1); FOR (i, 1, 10) fe_sq(t1, t1); fe_mul(t0, t1, t0); + fe_sq(t1, t0); FOR (i, 1, 50) fe_sq(t1, t1); fe_mul(t1, t1, t0); + fe_sq(t2, t1); FOR (i, 1, 100) fe_sq(t2, t2); fe_mul(t1, t2, t1); + fe_sq(t1, t1); FOR (i, 1, 50) fe_sq(t1, t1); fe_mul(t0, t1, t0); + fe_sq(t0, t0); FOR (i, 1, 2) fe_sq(t0, t0); fe_mul(t0, t0, x); + + // quartic = x^((p-1)/4) + i32 *quartic = t1; + fe_sq (quartic, t0); + fe_mul(quartic, quartic, x); + + i32 *check = t2; + fe_0 (check); int z0 = fe_isequal(x , check); + fe_1 (check); int p1 = fe_isequal(quartic, check); + fe_neg(check, check ); int m1 = fe_isequal(quartic, check); + fe_neg(check, sqrtm1); int ms = fe_isequal(quartic, check); + + // if quartic == -1 or sqrt(-1) + // then isr = x^((p-1)/4) * sqrt(-1) + // else isr = x^((p-1)/4) + fe_mul(isr, t0, sqrtm1); + fe_ccopy(isr, t0, 1 - (m1 | ms)); + + WIPE_BUFFER(t0); + WIPE_BUFFER(t1); + WIPE_BUFFER(t2); + return p1 | m1 | z0; +} + +// Inverse in terms of inverse square root. +// Requires two additional squarings to get rid of the sign. +// +// 1/x = x * (+invsqrt(x^2))^2 +// = x * (-invsqrt(x^2))^2 +// +// A fully optimised exponentiation by p-1 would save 6 field +// multiplications, but it would require more code. +static void fe_invert(fe out, const fe x) +{ + fe tmp; + fe_sq(tmp, x); + invsqrt(tmp, tmp); + fe_sq(tmp, tmp); + fe_mul(out, tmp, x); + WIPE_BUFFER(tmp); +} + +// trim a scalar for scalar multiplication +void crypto_eddsa_trim_scalar(u8 out[32], const u8 in[32]) +{ + COPY(out, in, 32); + out[ 0] &= 248; + out[31] &= 127; + out[31] |= 64; +} + +// get bit from scalar at position i +static int scalar_bit(const u8 s[32], int i) +{ + if (i < 0) { return 0; } // handle -1 for sliding windows + return (s[i>>3] >> (i&7)) & 1; +} + +/////////////// +/// X-25519 /// Taken from SUPERCOP's ref10 implementation. +/////////////// +static void scalarmult(u8 q[32], const u8 scalar[32], const u8 p[32], + int nb_bits) +{ + // computes the scalar product + fe x1; + fe_frombytes(x1, p); + + // computes the actual scalar product (the result is in x2 and z2) + fe x2, z2, x3, z3, t0, t1; + // Montgomery ladder + // In projective coordinates, to avoid divisions: x = X / Z + // We don't care about the y coordinate, it's only 1 bit of information + fe_1(x2); fe_0(z2); // "zero" point + fe_copy(x3, x1); fe_1(z3); // "one" point + int swap = 0; + for (int pos = nb_bits-1; pos >= 0; --pos) { + // constant time conditional swap before ladder step + int b = scalar_bit(scalar, pos); + swap ^= b; // xor trick avoids swapping at the end of the loop + fe_cswap(x2, x3, swap); + fe_cswap(z2, z3, swap); + swap = b; // anticipates one last swap after the loop + + // Montgomery ladder step: replaces (P2, P3) by (P2*2, P2+P3) + // with differential addition + fe_sub(t0, x3, z3); + fe_sub(t1, x2, z2); + fe_add(x2, x2, z2); + fe_add(z2, x3, z3); + fe_mul(z3, t0, x2); + fe_mul(z2, z2, t1); + fe_sq (t0, t1 ); + fe_sq (t1, x2 ); + fe_add(x3, z3, z2); + fe_sub(z2, z3, z2); + fe_mul(x2, t1, t0); + fe_sub(t1, t1, t0); + fe_sq (z2, z2 ); + fe_mul_small(z3, t1, 121666); + fe_sq (x3, x3 ); + fe_add(t0, t0, z3); + fe_mul(z3, x1, z2); + fe_mul(z2, t1, t0); + } + // last swap is necessary to compensate for the xor trick + // Note: after this swap, P3 == P2 + P1. + fe_cswap(x2, x3, swap); + fe_cswap(z2, z3, swap); + + // normalises the coordinates: x == X / Z + fe_invert(z2, z2); + fe_mul(x2, x2, z2); + fe_tobytes(q, x2); + + WIPE_BUFFER(x1); + WIPE_BUFFER(x2); WIPE_BUFFER(z2); WIPE_BUFFER(t0); + WIPE_BUFFER(x3); WIPE_BUFFER(z3); WIPE_BUFFER(t1); +} + +void crypto_x25519(u8 raw_shared_secret[32], + const u8 your_secret_key [32], + const u8 their_public_key [32]) +{ + // restrict the possible scalar values + u8 e[32]; + crypto_eddsa_trim_scalar(e, your_secret_key); + scalarmult(raw_shared_secret, e, their_public_key, 255); + WIPE_BUFFER(e); +} + +void crypto_x25519_public_key(u8 public_key[32], + const u8 secret_key[32]) +{ + static const u8 base_point[32] = {9}; + crypto_x25519(public_key, secret_key, base_point); +} + +/////////////////////////// +/// Arithmetic modulo L /// +/////////////////////////// +static const u32 L[8] = { + 0x5cf5d3ed, 0x5812631a, 0xa2f79cd6, 0x14def9de, + 0x00000000, 0x00000000, 0x00000000, 0x10000000, +}; + +// p = a*b + p +static void multiply(u32 p[16], const u32 a[8], const u32 b[8]) +{ + FOR (i, 0, 8) { + u64 carry = 0; + FOR (j, 0, 8) { + carry += p[i+j] + (u64)a[i] * b[j]; + p[i+j] = (u32)carry; + carry >>= 32; + } + p[i+8] = (u32)carry; + } +} + +static int is_above_l(const u32 x[8]) +{ + // We work with L directly, in a 2's complement encoding + // (-L == ~L + 1) + u64 carry = 1; + FOR (i, 0, 8) { + carry += (u64)x[i] + (~L[i] & 0xffffffff); + carry >>= 32; + } + return (int)carry; // carry is either 0 or 1 +} + +// Final reduction modulo L, by conditionally removing L. +// if x < l , then r = x +// if l <= x 2*l, then r = x-l +// otherwise the result will be wrong +static void remove_l(u32 r[8], const u32 x[8]) +{ + u64 carry = (u64)is_above_l(x); + u32 mask = ~(u32)carry + 1; // carry == 0 or 1 + FOR (i, 0, 8) { + carry += (u64)x[i] + (~L[i] & mask); + r[i] = (u32)carry; + carry >>= 32; + } +} + +// Full reduction modulo L (Barrett reduction) +static void mod_l(u8 reduced[32], const u32 x[16]) +{ + static const u32 r[9] = { + 0x0a2c131b,0xed9ce5a3,0x086329a7,0x2106215d, + 0xffffffeb,0xffffffff,0xffffffff,0xffffffff,0xf, + }; + // xr = x * r + u32 xr[25] = {0}; + FOR (i, 0, 9) { + u64 carry = 0; + FOR (j, 0, 16) { + carry += xr[i+j] + (u64)r[i] * x[j]; + xr[i+j] = (u32)carry; + carry >>= 32; + } + xr[i+16] = (u32)carry; + } + // xr = floor(xr / 2^512) * L + // Since the result is guaranteed to be below 2*L, + // it is enough to only compute the first 256 bits. + // The division is performed by saying xr[i+16]. (16 * 32 = 512) + ZERO(xr, 8); + FOR (i, 0, 8) { + u64 carry = 0; + FOR (j, 0, 8-i) { + carry += xr[i+j] + (u64)xr[i+16] * L[j]; + xr[i+j] = (u32)carry; + carry >>= 32; + } + } + // xr = x - xr + u64 carry = 1; + FOR (i, 0, 8) { + carry += (u64)x[i] + (~xr[i] & 0xffffffff); + xr[i] = (u32)carry; + carry >>= 32; + } + // Final reduction modulo L (conditional subtraction) + remove_l(xr, xr); + store32_le_buf(reduced, xr, 8); + + WIPE_BUFFER(xr); +} + +void crypto_eddsa_reduce(u8 reduced[32], const u8 expanded[64]) +{ + u32 x[16]; + load32_le_buf(x, expanded, 16); + mod_l(reduced, x); + WIPE_BUFFER(x); +} + +// r = (a * b) + c +void crypto_eddsa_mul_add(u8 r[32], + const u8 a[32], const u8 b[32], const u8 c[32]) +{ + u32 A[8]; load32_le_buf(A, a, 8); + u32 B[8]; load32_le_buf(B, b, 8); + u32 p[16]; load32_le_buf(p, c, 8); ZERO(p + 8, 8); + multiply(p, A, B); + mod_l(r, p); + WIPE_BUFFER(p); + WIPE_BUFFER(A); + WIPE_BUFFER(B); +} + +/////////////// +/// Ed25519 /// +/////////////// + +// Point (group element, ge) in a twisted Edwards curve, +// in extended projective coordinates. +// ge : x = X/Z, y = Y/Z, T = XY/Z +// ge_cached : Yp = X+Y, Ym = X-Y, T2 = T*D2 +// ge_precomp: Z = 1 +typedef struct { fe X; fe Y; fe Z; fe T; } ge; +typedef struct { fe Yp; fe Ym; fe Z; fe T2; } ge_cached; +typedef struct { fe Yp; fe Ym; fe T2; } ge_precomp; + +static void ge_zero(ge *p) +{ + fe_0(p->X); + fe_1(p->Y); + fe_1(p->Z); + fe_0(p->T); +} + +static void ge_tobytes(u8 s[32], const ge *h) +{ + fe recip, x, y; + fe_invert(recip, h->Z); + fe_mul(x, h->X, recip); + fe_mul(y, h->Y, recip); + fe_tobytes(s, y); + s[31] ^= fe_isodd(x) << 7; + + WIPE_BUFFER(recip); + WIPE_BUFFER(x); + WIPE_BUFFER(y); +} + +// h = -s, where s is a point encoded in 32 bytes +// +// Variable time! Inputs must not be secret! +// => Use only to *check* signatures. +// +// From the specifications: +// The encoding of s contains y and the sign of x +// x = sqrt((y^2 - 1) / (d*y^2 + 1)) +// In extended coordinates: +// X = x, Y = y, Z = 1, T = x*y +// +// Note that num * den is a square iff num / den is a square +// If num * den is not a square, the point was not on the curve. +// From the above: +// Let num = y^2 - 1 +// Let den = d*y^2 + 1 +// x = sqrt((y^2 - 1) / (d*y^2 + 1)) +// x = sqrt(num / den) +// x = sqrt(num^2 / (num * den)) +// x = num * sqrt(1 / (num * den)) +// +// Therefore, we can just compute: +// num = y^2 - 1 +// den = d*y^2 + 1 +// isr = invsqrt(num * den) // abort if not square +// x = num * isr +// Finally, negate x if its sign is not as specified. +static int ge_frombytes_neg_vartime(ge *h, const u8 s[32]) +{ + fe_frombytes(h->Y, s); + fe_1(h->Z); + fe_sq (h->T, h->Y); // t = y^2 + fe_mul(h->X, h->T, d ); // x = d*y^2 + fe_sub(h->T, h->T, h->Z); // t = y^2 - 1 + fe_add(h->X, h->X, h->Z); // x = d*y^2 + 1 + fe_mul(h->X, h->T, h->X); // x = (y^2 - 1) * (d*y^2 + 1) + int is_square = invsqrt(h->X, h->X); + if (!is_square) { + return -1; // Not on the curve, abort + } + fe_mul(h->X, h->T, h->X); // x = sqrt((y^2 - 1) / (d*y^2 + 1)) + if (fe_isodd(h->X) == (s[31] >> 7)) { + fe_neg(h->X, h->X); + } + fe_mul(h->T, h->X, h->Y); + return 0; +} + +static void ge_cache(ge_cached *c, const ge *p) +{ + fe_add (c->Yp, p->Y, p->X); + fe_sub (c->Ym, p->Y, p->X); + fe_copy(c->Z , p->Z ); + fe_mul (c->T2, p->T, D2 ); +} + +// Internal buffers are not wiped! Inputs must not be secret! +// => Use only to *check* signatures. +static void ge_add(ge *s, const ge *p, const ge_cached *q) +{ + fe a, b; + fe_add(a , p->Y, p->X ); + fe_sub(b , p->Y, p->X ); + fe_mul(a , a , q->Yp); + fe_mul(b , b , q->Ym); + fe_add(s->Y, a , b ); + fe_sub(s->X, a , b ); + + fe_add(s->Z, p->Z, p->Z ); + fe_mul(s->Z, s->Z, q->Z ); + fe_mul(s->T, p->T, q->T2); + fe_add(a , s->Z, s->T ); + fe_sub(b , s->Z, s->T ); + + fe_mul(s->T, s->X, s->Y); + fe_mul(s->X, s->X, b ); + fe_mul(s->Y, s->Y, a ); + fe_mul(s->Z, a , b ); +} + +// Internal buffers are not wiped! Inputs must not be secret! +// => Use only to *check* signatures. +static void ge_sub(ge *s, const ge *p, const ge_cached *q) +{ + ge_cached neg; + fe_copy(neg.Ym, q->Yp); + fe_copy(neg.Yp, q->Ym); + fe_copy(neg.Z , q->Z ); + fe_neg (neg.T2, q->T2); + ge_add(s, p, &neg); +} + +static void ge_madd(ge *s, const ge *p, const ge_precomp *q, fe a, fe b) +{ + fe_add(a , p->Y, p->X ); + fe_sub(b , p->Y, p->X ); + fe_mul(a , a , q->Yp); + fe_mul(b , b , q->Ym); + fe_add(s->Y, a , b ); + fe_sub(s->X, a , b ); + + fe_add(s->Z, p->Z, p->Z ); + fe_mul(s->T, p->T, q->T2); + fe_add(a , s->Z, s->T ); + fe_sub(b , s->Z, s->T ); + + fe_mul(s->T, s->X, s->Y); + fe_mul(s->X, s->X, b ); + fe_mul(s->Y, s->Y, a ); + fe_mul(s->Z, a , b ); +} + +// Internal buffers are not wiped! Inputs must not be secret! +// => Use only to *check* signatures. +static void ge_msub(ge *s, const ge *p, const ge_precomp *q, fe a, fe b) +{ + ge_precomp neg; + fe_copy(neg.Ym, q->Yp); + fe_copy(neg.Yp, q->Ym); + fe_neg (neg.T2, q->T2); + ge_madd(s, p, &neg, a, b); +} + +static void ge_double(ge *s, const ge *p, ge *q) +{ + fe_sq (q->X, p->X); + fe_sq (q->Y, p->Y); + fe_sq (q->Z, p->Z); // qZ = pZ^2 + fe_mul_small(q->Z, q->Z, 2); // qZ = pZ^2 * 2 + fe_add(q->T, p->X, p->Y); + fe_sq (s->T, q->T); + fe_add(q->T, q->Y, q->X); + fe_sub(q->Y, q->Y, q->X); + fe_sub(q->X, s->T, q->T); + fe_sub(q->Z, q->Z, q->Y); + + fe_mul(s->X, q->X , q->Z); + fe_mul(s->Y, q->T , q->Y); + fe_mul(s->Z, q->Y , q->Z); + fe_mul(s->T, q->X , q->T); +} + +// 5-bit signed window in cached format (Niels coordinates, Z=1) +static const ge_precomp b_window[8] = { + {{25967493,-14356035,29566456,3660896,-12694345, + 4014787,27544626,-11754271,-6079156,2047605,}, + {-12545711,934262,-2722910,3049990,-727428, + 9406986,12720692,5043384,19500929,-15469378,}, + {-8738181,4489570,9688441,-14785194,10184609, + -12363380,29287919,11864899,-24514362,-4438546,},}, + {{15636291,-9688557,24204773,-7912398,616977, + -16685262,27787600,-14772189,28944400,-1550024,}, + {16568933,4717097,-11556148,-1102322,15682896, + -11807043,16354577,-11775962,7689662,11199574,}, + {30464156,-5976125,-11779434,-15670865,23220365, + 15915852,7512774,10017326,-17749093,-9920357,},}, + {{10861363,11473154,27284546,1981175,-30064349, + 12577861,32867885,14515107,-15438304,10819380,}, + {4708026,6336745,20377586,9066809,-11272109, + 6594696,-25653668,12483688,-12668491,5581306,}, + {19563160,16186464,-29386857,4097519,10237984, + -4348115,28542350,13850243,-23678021,-15815942,},}, + {{5153746,9909285,1723747,-2777874,30523605, + 5516873,19480852,5230134,-23952439,-15175766,}, + {-30269007,-3463509,7665486,10083793,28475525, + 1649722,20654025,16520125,30598449,7715701,}, + {28881845,14381568,9657904,3680757,-20181635, + 7843316,-31400660,1370708,29794553,-1409300,},}, + {{-22518993,-6692182,14201702,-8745502,-23510406, + 8844726,18474211,-1361450,-13062696,13821877,}, + {-6455177,-7839871,3374702,-4740862,-27098617, + -10571707,31655028,-7212327,18853322,-14220951,}, + {4566830,-12963868,-28974889,-12240689,-7602672, + -2830569,-8514358,-10431137,2207753,-3209784,},}, + {{-25154831,-4185821,29681144,7868801,-6854661, + -9423865,-12437364,-663000,-31111463,-16132436,}, + {25576264,-2703214,7349804,-11814844,16472782, + 9300885,3844789,15725684,171356,6466918,}, + {23103977,13316479,9739013,-16149481,817875, + -15038942,8965339,-14088058,-30714912,16193877,},}, + {{-33521811,3180713,-2394130,14003687,-16903474, + -16270840,17238398,4729455,-18074513,9256800,}, + {-25182317,-4174131,32336398,5036987,-21236817, + 11360617,22616405,9761698,-19827198,630305,}, + {-13720693,2639453,-24237460,-7406481,9494427, + -5774029,-6554551,-15960994,-2449256,-14291300,},}, + {{-3151181,-5046075,9282714,6866145,-31907062, + -863023,-18940575,15033784,25105118,-7894876,}, + {-24326370,15950226,-31801215,-14592823,-11662737, + -5090925,1573892,-2625887,2198790,-15804619,}, + {-3099351,10324967,-2241613,7453183,-5446979, + -2735503,-13812022,-16236442,-32461234,-12290683,},}, +}; + +// Incremental sliding windows (left to right) +// Based on Roberto Maria Avanzi[2005] +typedef struct { + i16 next_index; // position of the next signed digit + i8 next_digit; // next signed digit (odd number below 2^window_width) + u8 next_check; // point at which we must check for a new window +} slide_ctx; + +static void slide_init(slide_ctx *ctx, const u8 scalar[32]) +{ + // scalar is guaranteed to be below L, either because we checked (s), + // or because we reduced it modulo L (h_ram). L is under 2^253, so + // so bits 253 to 255 are guaranteed to be zero. No need to test them. + // + // Note however that L is very close to 2^252, so bit 252 is almost + // always zero. If we were to start at bit 251, the tests wouldn't + // catch the off-by-one error (constructing one that does would be + // prohibitively expensive). + // + // We should still check bit 252, though. + int i = 252; + while (i > 0 && scalar_bit(scalar, i) == 0) { + i--; + } + ctx->next_check = (u8)(i + 1); + ctx->next_index = -1; + ctx->next_digit = -1; +} + +static int slide_step(slide_ctx *ctx, int width, int i, const u8 scalar[32]) +{ + if (i == ctx->next_check) { + if (scalar_bit(scalar, i) == scalar_bit(scalar, i - 1)) { + ctx->next_check--; + } else { + // compute digit of next window + int w = MIN(width, i + 1); + int v = -(scalar_bit(scalar, i) << (w-1)); + FOR_T (int, j, 0, w-1) { + v += scalar_bit(scalar, i-(w-1)+j) << j; + } + v += scalar_bit(scalar, i-w); + int lsb = v & (~v + 1); // smallest bit of v + int s = // log2(lsb) + (((lsb & 0xAA) != 0) << 0) | + (((lsb & 0xCC) != 0) << 1) | + (((lsb & 0xF0) != 0) << 2); + ctx->next_index = (i16)(i-(w-1)+s); + ctx->next_digit = (i8) (v >> s ); + ctx->next_check -= (u8) w; + } + } + return i == ctx->next_index ? ctx->next_digit: 0; +} + +#define P_W_WIDTH 3 // Affects the size of the stack +#define B_W_WIDTH 5 // Affects the size of the binary +#define P_W_SIZE (1<<(P_W_WIDTH-2)) + +int crypto_eddsa_check_equation(const u8 signature[64], const u8 public_key[32], + const u8 h[32]) +{ + ge minus_A; // -public_key + ge minus_R; // -first_half_of_signature + const u8 *s = signature + 32; + + // Check that A and R are on the curve + // Check that 0 <= S < L (prevents malleability) + // *Allow* non-cannonical encoding for A and R + { + u32 s32[8]; + load32_le_buf(s32, s, 8); + if (ge_frombytes_neg_vartime(&minus_A, public_key) || + ge_frombytes_neg_vartime(&minus_R, signature) || + is_above_l(s32)) { + return -1; + } + } + + // look-up table for minus_A + ge_cached lutA[P_W_SIZE]; + { + ge minus_A2, tmp; + ge_double(&minus_A2, &minus_A, &tmp); + ge_cache(&lutA[0], &minus_A); + FOR (i, 1, P_W_SIZE) { + ge_add(&tmp, &minus_A2, &lutA[i-1]); + ge_cache(&lutA[i], &tmp); + } + } + + // sum = [s]B - [h]A + // Merged double and add ladder, fused with sliding + slide_ctx h_slide; slide_init(&h_slide, h); + slide_ctx s_slide; slide_init(&s_slide, s); + int i = MAX(h_slide.next_check, s_slide.next_check); + ge *sum = &minus_A; // reuse minus_A for the sum + ge_zero(sum); + while (i >= 0) { + ge tmp; + ge_double(sum, sum, &tmp); + int h_digit = slide_step(&h_slide, P_W_WIDTH, i, h); + int s_digit = slide_step(&s_slide, B_W_WIDTH, i, s); + if (h_digit > 0) { ge_add(sum, sum, &lutA[ h_digit / 2]); } + if (h_digit < 0) { ge_sub(sum, sum, &lutA[-h_digit / 2]); } + fe t1, t2; + if (s_digit > 0) { ge_madd(sum, sum, b_window + s_digit/2, t1, t2); } + if (s_digit < 0) { ge_msub(sum, sum, b_window + -s_digit/2, t1, t2); } + i--; + } + + // Compare [8](sum-R) and the zero point + // The multiplication by 8 eliminates any low-order component + // and ensures consistency with batched verification. + ge_cached cached; + u8 check[32]; + static const u8 zero_point[32] = {1}; // Point of order 1 + ge_cache(&cached, &minus_R); + ge_add(sum, sum, &cached); + ge_double(sum, sum, &minus_R); // reuse minus_R as temporary + ge_double(sum, sum, &minus_R); // reuse minus_R as temporary + ge_double(sum, sum, &minus_R); // reuse minus_R as temporary + ge_tobytes(check, sum); + return crypto_verify32(check, zero_point); +} + +// 5-bit signed comb in cached format (Niels coordinates, Z=1) +static const ge_precomp b_comb_low[8] = { + {{-6816601,-2324159,-22559413,124364,18015490, + 8373481,19993724,1979872,-18549925,9085059,}, + {10306321,403248,14839893,9633706,8463310, + -8354981,-14305673,14668847,26301366,2818560,}, + {-22701500,-3210264,-13831292,-2927732,-16326337, + -14016360,12940910,177905,12165515,-2397893,},}, + {{-12282262,-7022066,9920413,-3064358,-32147467, + 2927790,22392436,-14852487,2719975,16402117,}, + {-7236961,-4729776,2685954,-6525055,-24242706, + -15940211,-6238521,14082855,10047669,12228189,}, + {-30495588,-12893761,-11161261,3539405,-11502464, + 16491580,-27286798,-15030530,-7272871,-15934455,},}, + {{17650926,582297,-860412,-187745,-12072900, + -10683391,-20352381,15557840,-31072141,-5019061,}, + {-6283632,-2259834,-4674247,-4598977,-4089240, + 12435688,-31278303,1060251,6256175,10480726,}, + {-13871026,2026300,-21928428,-2741605,-2406664, + -8034988,7355518,15733500,-23379862,7489131,},}, + {{6883359,695140,23196907,9644202,-33430614, + 11354760,-20134606,6388313,-8263585,-8491918,}, + {-7716174,-13605463,-13646110,14757414,-19430591, + -14967316,10359532,-11059670,-21935259,12082603,}, + {-11253345,-15943946,10046784,5414629,24840771, + 8086951,-6694742,9868723,15842692,-16224787,},}, + {{9639399,11810955,-24007778,-9320054,3912937, + -9856959,996125,-8727907,-8919186,-14097242,}, + {7248867,14468564,25228636,-8795035,14346339, + 8224790,6388427,-7181107,6468218,-8720783,}, + {15513115,15439095,7342322,-10157390,18005294, + -7265713,2186239,4884640,10826567,7135781,},}, + {{-14204238,5297536,-5862318,-6004934,28095835, + 4236101,-14203318,1958636,-16816875,3837147,}, + {-5511166,-13176782,-29588215,12339465,15325758, + -15945770,-8813185,11075932,-19608050,-3776283,}, + {11728032,9603156,-4637821,-5304487,-7827751, + 2724948,31236191,-16760175,-7268616,14799772,},}, + {{-28842672,4840636,-12047946,-9101456,-1445464, + 381905,-30977094,-16523389,1290540,12798615,}, + {27246947,-10320914,14792098,-14518944,5302070, + -8746152,-3403974,-4149637,-27061213,10749585,}, + {25572375,-6270368,-15353037,16037944,1146292, + 32198,23487090,9585613,24714571,-1418265,},}, + {{19844825,282124,-17583147,11004019,-32004269, + -2716035,6105106,-1711007,-21010044,14338445,}, + {8027505,8191102,-18504907,-12335737,25173494, + -5923905,15446145,7483684,-30440441,10009108,}, + {-14134701,-4174411,10246585,-14677495,33553567, + -14012935,23366126,15080531,-7969992,7663473,},}, +}; + +static const ge_precomp b_comb_high[8] = { + {{33055887,-4431773,-521787,6654165,951411, + -6266464,-5158124,6995613,-5397442,-6985227,}, + {4014062,6967095,-11977872,3960002,8001989, + 5130302,-2154812,-1899602,-31954493,-16173976,}, + {16271757,-9212948,23792794,731486,-25808309, + -3546396,6964344,-4767590,10976593,10050757,},}, + {{2533007,-4288439,-24467768,-12387405,-13450051, + 14542280,12876301,13893535,15067764,8594792,}, + {20073501,-11623621,3165391,-13119866,13188608, + -11540496,-10751437,-13482671,29588810,2197295,}, + {-1084082,11831693,6031797,14062724,14748428, + -8159962,-20721760,11742548,31368706,13161200,},}, + {{2050412,-6457589,15321215,5273360,25484180, + 124590,-18187548,-7097255,-6691621,-14604792,}, + {9938196,2162889,-6158074,-1711248,4278932, + -2598531,-22865792,-7168500,-24323168,11746309,}, + {-22691768,-14268164,5965485,9383325,20443693, + 5854192,28250679,-1381811,-10837134,13717818,},}, + {{-8495530,16382250,9548884,-4971523,-4491811, + -3902147,6182256,-12832479,26628081,10395408,}, + {27329048,-15853735,7715764,8717446,-9215518, + -14633480,28982250,-5668414,4227628,242148,}, + {-13279943,-7986904,-7100016,8764468,-27276630, + 3096719,29678419,-9141299,3906709,11265498,},}, + {{11918285,15686328,-17757323,-11217300,-27548967, + 4853165,-27168827,6807359,6871949,-1075745,}, + {-29002610,13984323,-27111812,-2713442,28107359, + -13266203,6155126,15104658,3538727,-7513788,}, + {14103158,11233913,-33165269,9279850,31014152, + 4335090,-1827936,4590951,13960841,12787712,},}, + {{1469134,-16738009,33411928,13942824,8092558, + -8778224,-11165065,1437842,22521552,-2792954,}, + {31352705,-4807352,-25327300,3962447,12541566, + -9399651,-27425693,7964818,-23829869,5541287,}, + {-25732021,-6864887,23848984,3039395,-9147354, + 6022816,-27421653,10590137,25309915,-1584678,},}, + {{-22951376,5048948,31139401,-190316,-19542447, + -626310,-17486305,-16511925,-18851313,-12985140,}, + {-9684890,14681754,30487568,7717771,-10829709, + 9630497,30290549,-10531496,-27798994,-13812825,}, + {5827835,16097107,-24501327,12094619,7413972, + 11447087,28057551,-1793987,-14056981,4359312,},}, + {{26323183,2342588,-21887793,-1623758,-6062284, + 2107090,-28724907,9036464,-19618351,-13055189,}, + {-29697200,14829398,-4596333,14220089,-30022969, + 2955645,12094100,-13693652,-5941445,7047569,}, + {-3201977,14413268,-12058324,-16417589,-9035655, + -7224648,9258160,1399236,30397584,-5684634,},}, +}; + +static void lookup_add(ge *p, ge_precomp *tmp_c, fe tmp_a, fe tmp_b, + const ge_precomp comb[8], const u8 scalar[32], int i) +{ + u8 teeth = (u8)((scalar_bit(scalar, i) ) + + (scalar_bit(scalar, i + 32) << 1) + + (scalar_bit(scalar, i + 64) << 2) + + (scalar_bit(scalar, i + 96) << 3)); + u8 high = teeth >> 3; + u8 index = (teeth ^ (high - 1)) & 7; + FOR (j, 0, 8) { + i32 select = 1 & (((j ^ index) - 1) >> 8); + fe_ccopy(tmp_c->Yp, comb[j].Yp, select); + fe_ccopy(tmp_c->Ym, comb[j].Ym, select); + fe_ccopy(tmp_c->T2, comb[j].T2, select); + } + fe_neg(tmp_a, tmp_c->T2); + fe_cswap(tmp_c->T2, tmp_a , high ^ 1); + fe_cswap(tmp_c->Yp, tmp_c->Ym, high ^ 1); + ge_madd(p, p, tmp_c, tmp_a, tmp_b); +} + +// p = [scalar]B, where B is the base point +static void ge_scalarmult_base(ge *p, const u8 scalar[32]) +{ + // twin 4-bits signed combs, from Mike Hamburg's + // Fast and compact elliptic-curve cryptography (2012) + // 1 / 2 modulo L + static const u8 half_mod_L[32] = { + 247,233,122,46,141,49,9,44,107,206,123,81,239,124,111,10, + 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,8, + }; + // (2^256 - 1) / 2 modulo L + static const u8 half_ones[32] = { + 142,74,204,70,186,24,118,107,184,231,190,57,250,173,119,99, + 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,7, + }; + + // All bits set form: 1 means 1, 0 means -1 + u8 s_scalar[32]; + crypto_eddsa_mul_add(s_scalar, scalar, half_mod_L, half_ones); + + // Double and add ladder + fe tmp_a, tmp_b; // temporaries for addition + ge_precomp tmp_c; // temporary for comb lookup + ge tmp_d; // temporary for doubling + fe_1(tmp_c.Yp); + fe_1(tmp_c.Ym); + fe_0(tmp_c.T2); + + // Save a double on the first iteration + ge_zero(p); + lookup_add(p, &tmp_c, tmp_a, tmp_b, b_comb_low , s_scalar, 31); + lookup_add(p, &tmp_c, tmp_a, tmp_b, b_comb_high, s_scalar, 31+128); + // Regular double & add for the rest + for (int i = 30; i >= 0; i--) { + ge_double(p, p, &tmp_d); + lookup_add(p, &tmp_c, tmp_a, tmp_b, b_comb_low , s_scalar, i); + lookup_add(p, &tmp_c, tmp_a, tmp_b, b_comb_high, s_scalar, i+128); + } + // Note: we could save one addition at the end if we assumed the + // scalar fit in 252 bits. Which it does in practice if it is + // selected at random. However, non-random, non-hashed scalars + // *can* overflow 252 bits in practice. Better account for that + // than leaving that kind of subtle corner case. + + WIPE_BUFFER(tmp_a); WIPE_CTX(&tmp_d); + WIPE_BUFFER(tmp_b); WIPE_CTX(&tmp_c); + WIPE_BUFFER(s_scalar); +} + +void crypto_eddsa_scalarbase(u8 point[32], const u8 scalar[32]) +{ + ge P; + ge_scalarmult_base(&P, scalar); + ge_tobytes(point, &P); + WIPE_CTX(&P); +} + +void crypto_eddsa_key_pair(u8 secret_key[64], u8 public_key[32], u8 seed[32]) +{ + // To allow overlaps, observable writes happen in this order: + // 1. seed + // 2. secret_key + // 3. public_key + u8 a[64]; + COPY(a, seed, 32); + crypto_wipe(seed, 32); + COPY(secret_key, a, 32); + crypto_blake2b(a, 64, a, 32); + crypto_eddsa_trim_scalar(a, a); + crypto_eddsa_scalarbase(secret_key + 32, a); + COPY(public_key, secret_key + 32, 32); + WIPE_BUFFER(a); +} + +static void hash_reduce(u8 h[32], + const u8 *a, size_t a_size, + const u8 *b, size_t b_size, + const u8 *c, size_t c_size) +{ + u8 hash[64]; + crypto_blake2b_ctx ctx; + crypto_blake2b_init (&ctx, 64); + crypto_blake2b_update(&ctx, a, a_size); + crypto_blake2b_update(&ctx, b, b_size); + crypto_blake2b_update(&ctx, c, c_size); + crypto_blake2b_final (&ctx, hash); + crypto_eddsa_reduce(h, hash); +} + +// Digital signature of a message with from a secret key. +// +// The secret key comprises two parts: +// - The seed that generates the key (secret_key[ 0..31]) +// - The public key (secret_key[32..63]) +// +// The seed and the public key are bundled together to make sure users +// don't use mismatched seeds and public keys, which would instantly +// leak the secret scalar and allow forgeries (allowing this to happen +// has resulted in critical vulnerabilities in the wild). +// +// The seed is hashed to derive the secret scalar and a secret prefix. +// The sole purpose of the prefix is to generate a secret random nonce. +// The properties of that nonce must be as follows: +// - Unique: we need a different one for each message. +// - Secret: third parties must not be able to predict it. +// - Random: any detectable bias would break all security. +// +// There are two ways to achieve these properties. The obvious one is +// to simply generate a random number. Here that would be a parameter +// (Monocypher doesn't have an RNG). It works, but then users may reuse +// the nonce by accident, which _also_ leaks the secret scalar and +// allows forgeries. This has happened in the wild too. +// +// This is no good, so instead we generate that nonce deterministically +// by reducing modulo L a hash of the secret prefix and the message. +// The secret prefix makes the nonce unpredictable, the message makes it +// unique, and the hash/reduce removes all bias. +// +// The cost of that safety is hashing the message twice. If that cost +// is unacceptable, there are two alternatives: +// +// - Signing a hash of the message instead of the message itself. This +// is fine as long as the hash is collision resistant. It is not +// compatible with existing "pure" signatures, but at least it's safe. +// +// - Using a random nonce. Please exercise **EXTREME CAUTION** if you +// ever do that. It is absolutely **critical** that the nonce is +// really an unbiased random number between 0 and L-1, never reused, +// and wiped immediately. +// +// To lower the likelihood of complete catastrophe if the RNG is +// either flawed or misused, you can hash the RNG output together with +// the secret prefix and the beginning of the message, and use the +// reduction of that hash instead of the RNG output itself. It's not +// foolproof (you'd need to hash the whole message) but it helps. +// +// Signing a message involves the following operations: +// +// scalar, prefix = HASH(secret_key) +// r = HASH(prefix || message) % L +// R = [r]B +// h = HASH(R || public_key || message) % L +// S = ((h * a) + r) % L +// signature = R || S +void crypto_eddsa_sign(u8 signature [64], const u8 secret_key[64], + const u8 *message, size_t message_size) +{ + u8 a[64]; // secret scalar and prefix + u8 r[32]; // secret deterministic "random" nonce + u8 h[32]; // publically verifiable hash of the message (not wiped) + u8 R[32]; // first half of the signature (allows overlapping inputs) + + crypto_blake2b(a, 64, secret_key, 32); + crypto_eddsa_trim_scalar(a, a); + hash_reduce(r, a + 32, 32, message, message_size, 0, 0); + crypto_eddsa_scalarbase(R, r); + hash_reduce(h, R, 32, secret_key + 32, 32, message, message_size); + COPY(signature, R, 32); + crypto_eddsa_mul_add(signature + 32, h, a, r); + + WIPE_BUFFER(a); + WIPE_BUFFER(r); +} + +// To check the signature R, S of the message M with the public key A, +// there are 3 steps: +// +// compute h = HASH(R || A || message) % L +// check that A is on the curve. +// check that R == [s]B - [h]A +// +// The last two steps are done in crypto_eddsa_check_equation() +int crypto_eddsa_check(const u8 signature[64], const u8 public_key[32], + const u8 *message, size_t message_size) +{ + u8 h[32]; + hash_reduce(h, signature, 32, public_key, 32, message, message_size); + return crypto_eddsa_check_equation(signature, public_key, h); +} + +///////////////////////// +/// EdDSA <--> X25519 /// +///////////////////////// +void crypto_eddsa_to_x25519(u8 x25519[32], const u8 eddsa[32]) +{ + // (u, v) = ((1+y)/(1-y), sqrt(-486664)*u/x) + // Only converting y to u, the sign of x is ignored. + fe t1, t2; + fe_frombytes(t2, eddsa); + fe_add(t1, fe_one, t2); + fe_sub(t2, fe_one, t2); + fe_invert(t2, t2); + fe_mul(t1, t1, t2); + fe_tobytes(x25519, t1); + WIPE_BUFFER(t1); + WIPE_BUFFER(t2); +} + +void crypto_x25519_to_eddsa(u8 eddsa[32], const u8 x25519[32]) +{ + // (x, y) = (sqrt(-486664)*u/v, (u-1)/(u+1)) + // Only converting u to y, x is assumed positive. + fe t1, t2; + fe_frombytes(t2, x25519); + fe_sub(t1, t2, fe_one); + fe_add(t2, t2, fe_one); + fe_invert(t2, t2); + fe_mul(t1, t1, t2); + fe_tobytes(eddsa, t1); + WIPE_BUFFER(t1); + WIPE_BUFFER(t2); +} + +///////////////////////////////////////////// +/// Dirty ephemeral public key generation /// +///////////////////////////////////////////// + +// Those functions generates a public key, *without* clearing the +// cofactor. Sending that key over the network leaks 3 bits of the +// private key. Use only to generate ephemeral keys that will be hidden +// with crypto_curve_to_hidden(). +// +// The public key is otherwise compatible with crypto_x25519(), which +// properly clears the cofactor. +// +// Note that the distribution of the resulting public keys is almost +// uniform. Flipping the sign of the v coordinate (not provided by this +// function), covers the entire key space almost perfectly, where +// "almost" means a 2^-128 bias (undetectable). This uniformity is +// needed to ensure the proper randomness of the resulting +// representatives (once we apply crypto_curve_to_hidden()). +// +// Recall that Curve25519 has order C = 2^255 + e, with e < 2^128 (not +// to be confused with the prime order of the main subgroup, L, which is +// 8 times less than that). +// +// Generating all points would require us to multiply a point of order C +// (the base point plus any point of order 8) by all scalars from 0 to +// C-1. Clamping limits us to scalars between 2^254 and 2^255 - 1. But +// by negating the resulting point at random, we also cover scalars from +// -2^255 + 1 to -2^254 (which modulo C is congruent to e+1 to 2^254 + e). +// +// In practice: +// - Scalars from 0 to e + 1 are never generated +// - Scalars from 2^255 to 2^255 + e are never generated +// - Scalars from 2^254 + 1 to 2^254 + e are generated twice +// +// Since e < 2^128, detecting this bias requires observing over 2^100 +// representatives from a given source (this will never happen), *and* +// recovering enough of the private key to determine that they do, or do +// not, belong to the biased set (this practically requires solving +// discrete logarithm, which is conjecturally intractable). +// +// In practice, this means the bias is impossible to detect. + +// s + (x*L) % 8*L +// Guaranteed to fit in 256 bits iff s fits in 255 bits. +// L < 2^253 +// x%8 < 2^3 +// L * (x%8) < 2^255 +// s < 2^255 +// s + L * (x%8) < 2^256 +static void add_xl(u8 s[32], u8 x) +{ + u64 mod8 = x & 7; + u64 carry = 0; + FOR (i , 0, 8) { + carry = carry + load32_le(s + 4*i) + L[i] * mod8; + store32_le(s + 4*i, (u32)carry); + carry >>= 32; + } +} + +// "Small" dirty ephemeral key. +// Use if you need to shrink the size of the binary, and can afford to +// slow down by a factor of two (compared to the fast version) +// +// This version works by decoupling the cofactor from the main factor. +// +// - The trimmed scalar determines the main factor +// - The clamped bits of the scalar determine the cofactor. +// +// Cofactor and main factor are combined into a single scalar, which is +// then multiplied by a point of order 8*L (unlike the base point, which +// has prime order). That "dirty" base point is the addition of the +// regular base point (9), and a point of order 8. +void crypto_x25519_dirty_small(u8 public_key[32], const u8 secret_key[32]) +{ + // Base point of order 8*L + // Raw scalar multiplication with it does not clear the cofactor, + // and the resulting public key will reveal 3 bits of the scalar. + // + // The low order component of this base point has been chosen + // to yield the same results as crypto_x25519_dirty_fast(). + static const u8 dirty_base_point[32] = { + 0xd8, 0x86, 0x1a, 0xa2, 0x78, 0x7a, 0xd9, 0x26, + 0x8b, 0x74, 0x74, 0xb6, 0x82, 0xe3, 0xbe, 0xc3, + 0xce, 0x36, 0x9a, 0x1e, 0x5e, 0x31, 0x47, 0xa2, + 0x6d, 0x37, 0x7c, 0xfd, 0x20, 0xb5, 0xdf, 0x75, + }; + // separate the main factor & the cofactor of the scalar + u8 scalar[32]; + crypto_eddsa_trim_scalar(scalar, secret_key); + + // Separate the main factor and the cofactor + // + // The scalar is trimmed, so its cofactor is cleared. The three + // least significant bits however still have a main factor. We must + // remove it for X25519 compatibility. + // + // cofactor = lsb * L (modulo 8*L) + // combined = scalar + cofactor (modulo 8*L) + add_xl(scalar, secret_key[0]); + scalarmult(public_key, scalar, dirty_base_point, 256); + WIPE_BUFFER(scalar); +} + +// Select low order point +// We're computing the [cofactor]lop scalar multiplication, where: +// +// cofactor = tweak & 7. +// lop = (lop_x, lop_y) +// lop_x = sqrt((sqrt(d + 1) + 1) / d) +// lop_y = -lop_x * sqrtm1 +// +// The low order point has order 8. There are 4 such points. We've +// chosen the one whose both coordinates are positive (below p/2). +// The 8 low order points are as follows: +// +// [0]lop = ( 0 , 1 ) +// [1]lop = ( lop_x , lop_y) +// [2]lop = ( sqrt(-1), -0 ) +// [3]lop = ( lop_x , -lop_y) +// [4]lop = (-0 , -1 ) +// [5]lop = (-lop_x , -lop_y) +// [6]lop = (-sqrt(-1), 0 ) +// [7]lop = (-lop_x , lop_y) +// +// The x coordinate is either 0, sqrt(-1), lop_x, or their opposite. +// The y coordinate is either 0, -1 , lop_y, or their opposite. +// The pattern for both is the same, except for a rotation of 2 (modulo 8) +// +// This helper function captures the pattern, and we can use it thus: +// +// select_lop(x, lop_x, sqrtm1, cofactor); +// select_lop(y, lop_y, fe_one, cofactor + 2); +// +// This is faster than an actual scalar multiplication, +// and requires less code than naive constant time look up. +static void select_lop(fe out, const fe x, const fe k, u8 cofactor) +{ + fe tmp; + fe_0(out); + fe_ccopy(out, k , (cofactor >> 1) & 1); // bit 1 + fe_ccopy(out, x , (cofactor >> 0) & 1); // bit 0 + fe_neg (tmp, out); + fe_ccopy(out, tmp, (cofactor >> 2) & 1); // bit 2 + WIPE_BUFFER(tmp); +} + +// "Fast" dirty ephemeral key +// We use this one by default. +// +// This version works by performing a regular scalar multiplication, +// then add a low order point. The scalar multiplication is done in +// Edwards space for more speed (*2 compared to the "small" version). +// The cost is a bigger binary for programs that don't also sign messages. +void crypto_x25519_dirty_fast(u8 public_key[32], const u8 secret_key[32]) +{ + // Compute clean scalar multiplication + u8 scalar[32]; + ge pk; + crypto_eddsa_trim_scalar(scalar, secret_key); + ge_scalarmult_base(&pk, scalar); + + // Compute low order point + fe t1, t2; + select_lop(t1, lop_x, sqrtm1, secret_key[0]); + select_lop(t2, lop_y, fe_one, secret_key[0] + 2); + ge_precomp low_order_point; + fe_add(low_order_point.Yp, t2, t1); + fe_sub(low_order_point.Ym, t2, t1); + fe_mul(low_order_point.T2, t2, t1); + fe_mul(low_order_point.T2, low_order_point.T2, D2); + + // Add low order point to the public key + ge_madd(&pk, &pk, &low_order_point, t1, t2); + + // Convert to Montgomery u coordinate (we ignore the sign) + fe_add(t1, pk.Z, pk.Y); + fe_sub(t2, pk.Z, pk.Y); + fe_invert(t2, t2); + fe_mul(t1, t1, t2); + + fe_tobytes(public_key, t1); + + WIPE_BUFFER(t1); WIPE_CTX(&pk); + WIPE_BUFFER(t2); WIPE_CTX(&low_order_point); + WIPE_BUFFER(scalar); +} + +/////////////////// +/// Elligator 2 /// +/////////////////// +static const fe A = {486662}; + +// Elligator direct map +// +// Computes the point corresponding to a representative, encoded in 32 +// bytes (little Endian). Since positive representatives fits in 254 +// bits, The two most significant bits are ignored. +// +// From the paper: +// w = -A / (fe(1) + non_square * r^2) +// e = chi(w^3 + A*w^2 + w) +// u = e*w - (fe(1)-e)*(A//2) +// v = -e * sqrt(u^3 + A*u^2 + u) +// +// We ignore v because we don't need it for X25519 (the Montgomery +// ladder only uses u). +// +// Note that e is either 0, 1 or -1 +// if e = 0 u = 0 and v = 0 +// if e = 1 u = w +// if e = -1 u = -w - A = w * non_square * r^2 +// +// Let r1 = non_square * r^2 +// Let r2 = 1 + r1 +// Note that r2 cannot be zero, -1/non_square is not a square. +// We can (tediously) verify that: +// w^3 + A*w^2 + w = (A^2*r1 - r2^2) * A / r2^3 +// Therefore: +// chi(w^3 + A*w^2 + w) = chi((A^2*r1 - r2^2) * (A / r2^3)) +// chi(w^3 + A*w^2 + w) = chi((A^2*r1 - r2^2) * (A / r2^3)) * 1 +// chi(w^3 + A*w^2 + w) = chi((A^2*r1 - r2^2) * (A / r2^3)) * chi(r2^6) +// chi(w^3 + A*w^2 + w) = chi((A^2*r1 - r2^2) * (A / r2^3) * r2^6) +// chi(w^3 + A*w^2 + w) = chi((A^2*r1 - r2^2) * A * r2^3) +// Corollary: +// e = 1 if (A^2*r1 - r2^2) * A * r2^3) is a non-zero square +// e = -1 if (A^2*r1 - r2^2) * A * r2^3) is not a square +// Note that w^3 + A*w^2 + w (and therefore e) can never be zero: +// w^3 + A*w^2 + w = w * (w^2 + A*w + 1) +// w^3 + A*w^2 + w = w * (w^2 + A*w + A^2/4 - A^2/4 + 1) +// w^3 + A*w^2 + w = w * (w + A/2)^2 - A^2/4 + 1) +// which is zero only if: +// w = 0 (impossible) +// (w + A/2)^2 = A^2/4 - 1 (impossible, because A^2/4-1 is not a square) +// +// Let isr = invsqrt((A^2*r1 - r2^2) * A * r2^3) +// isr = sqrt(1 / ((A^2*r1 - r2^2) * A * r2^3)) if e = 1 +// isr = sqrt(sqrt(-1) / ((A^2*r1 - r2^2) * A * r2^3)) if e = -1 +// +// if e = 1 +// let u1 = -A * (A^2*r1 - r2^2) * A * r2^2 * isr^2 +// u1 = w +// u1 = u +// +// if e = -1 +// let ufactor = -non_square * sqrt(-1) * r^2 +// let vfactor = sqrt(ufactor) +// let u2 = -A * (A^2*r1 - r2^2) * A * r2^2 * isr^2 * ufactor +// u2 = w * -1 * -non_square * r^2 +// u2 = w * non_square * r^2 +// u2 = u +void crypto_elligator_map(u8 curve[32], const u8 hidden[32]) +{ + fe r, u, t1, t2, t3; + fe_frombytes_mask(r, hidden, 2); // r is encoded in 254 bits. + fe_sq(r, r); + fe_add(t1, r, r); + fe_add(u, t1, fe_one); + fe_sq (t2, u); + fe_mul(t3, A2, t1); + fe_sub(t3, t3, t2); + fe_mul(t3, t3, A); + fe_mul(t1, t2, u); + fe_mul(t1, t3, t1); + int is_square = invsqrt(t1, t1); + fe_mul(u, r, ufactor); + fe_ccopy(u, fe_one, is_square); + fe_sq (t1, t1); + fe_mul(u, u, A); + fe_mul(u, u, t3); + fe_mul(u, u, t2); + fe_mul(u, u, t1); + fe_neg(u, u); + fe_tobytes(curve, u); + + WIPE_BUFFER(t1); WIPE_BUFFER(r); + WIPE_BUFFER(t2); WIPE_BUFFER(u); + WIPE_BUFFER(t3); +} + +// Elligator inverse map +// +// Computes the representative of a point, if possible. If not, it does +// nothing and returns -1. Note that the success of the operation +// depends only on the point (more precisely its u coordinate). The +// tweak parameter is used only upon success +// +// The tweak should be a random byte. Beyond that, its contents are an +// implementation detail. Currently, the tweak comprises: +// - Bit 1 : sign of the v coordinate (0 if positive, 1 if negative) +// - Bit 2-5: not used +// - Bits 6-7: random padding +// +// From the paper: +// Let sq = -non_square * u * (u+A) +// if sq is not a square, or u = -A, there is no mapping +// Assuming there is a mapping: +// if v is positive: r = sqrt(-u / (non_square * (u+A))) +// if v is negative: r = sqrt(-(u+A) / (non_square * u )) +// +// We compute isr = invsqrt(-non_square * u * (u+A)) +// if it wasn't a square, abort. +// else, isr = sqrt(-1 / (non_square * u * (u+A)) +// +// If v is positive, we return isr * u: +// isr * u = sqrt(-1 / (non_square * u * (u+A)) * u +// isr * u = sqrt(-u / (non_square * (u+A)) +// +// If v is negative, we return isr * (u+A): +// isr * (u+A) = sqrt(-1 / (non_square * u * (u+A)) * (u+A) +// isr * (u+A) = sqrt(-(u+A) / (non_square * u) +int crypto_elligator_rev(u8 hidden[32], const u8 public_key[32], u8 tweak) +{ + fe t1, t2, t3; + fe_frombytes(t1, public_key); // t1 = u + + fe_add(t2, t1, A); // t2 = u + A + fe_mul(t3, t1, t2); + fe_mul_small(t3, t3, -2); + int is_square = invsqrt(t3, t3); // t3 = sqrt(-1 / non_square * u * (u+A)) + if (is_square) { + // The only variable time bit. This ultimately reveals how many + // tries it took us to find a representable key. + // This does not affect security as long as we try keys at random. + + fe_ccopy (t1, t2, tweak & 1); // multiply by u if v is positive, + fe_mul (t3, t1, t3); // multiply by u+A otherwise + fe_mul_small(t1, t3, 2); + fe_neg (t2, t3); + fe_ccopy (t3, t2, fe_isodd(t1)); + fe_tobytes(hidden, t3); + + // Pad with two random bits + hidden[31] |= tweak & 0xc0; + } + + WIPE_BUFFER(t1); + WIPE_BUFFER(t2); + WIPE_BUFFER(t3); + return is_square - 1; +} + +void crypto_elligator_key_pair(u8 hidden[32], u8 secret_key[32], u8 seed[32]) +{ + u8 pk [32]; // public key + u8 buf[64]; // seed + representative + COPY(buf + 32, seed, 32); + do { + crypto_chacha20_djb(buf, 0, 64, buf+32, zero, 0); + crypto_x25519_dirty_fast(pk, buf); // or the "small" version + } while(crypto_elligator_rev(buf+32, pk, buf[32])); + // Note that the return value of crypto_elligator_rev() is + // independent from its tweak parameter. + // Therefore, buf[32] is not actually reused. Either we loop one + // more time and buf[32] is used for the new seed, or we succeeded, + // and buf[32] becomes the tweak parameter. + + crypto_wipe(seed, 32); + COPY(hidden , buf + 32, 32); + COPY(secret_key, buf , 32); + WIPE_BUFFER(buf); + WIPE_BUFFER(pk); +} + +/////////////////////// +/// Scalar division /// +/////////////////////// + +// Montgomery reduction. +// Divides x by (2^256), and reduces the result modulo L +// +// Precondition: +// x < L * 2^256 +// Constants: +// r = 2^256 (makes division by r trivial) +// k = (r * (1/r) - 1) // L (1/r is computed modulo L ) +// Algorithm: +// s = (x * k) % r +// t = x + s*L (t is always a multiple of r) +// u = (t/r) % L (u is always below 2*L, conditional subtraction is enough) +static void redc(u32 u[8], u32 x[16]) +{ + static const u32 k[8] = { + 0x12547e1b, 0xd2b51da3, 0xfdba84ff, 0xb1a206f2, + 0xffa36bea, 0x14e75438, 0x6fe91836, 0x9db6c6f2, + }; + + // s = x * k (modulo 2^256) + // This is cheaper than the full multiplication. + u32 s[8] = {0}; + FOR (i, 0, 8) { + u64 carry = 0; + FOR (j, 0, 8-i) { + carry += s[i+j] + (u64)x[i] * k[j]; + s[i+j] = (u32)carry; + carry >>= 32; + } + } + u32 t[16] = {0}; + multiply(t, s, L); + + // t = t + x + u64 carry = 0; + FOR (i, 0, 16) { + carry += (u64)t[i] + x[i]; + t[i] = (u32)carry; + carry >>= 32; + } + + // u = (t / 2^256) % L + // Note that t / 2^256 is always below 2*L, + // So a constant time conditional subtraction is enough + remove_l(u, t+8); + + WIPE_BUFFER(s); + WIPE_BUFFER(t); +} + +void crypto_x25519_inverse(u8 blind_salt [32], const u8 private_key[32], + const u8 curve_point[32]) +{ + static const u8 Lm2[32] = { // L - 2 + 0xeb, 0xd3, 0xf5, 0x5c, 0x1a, 0x63, 0x12, 0x58, + 0xd6, 0x9c, 0xf7, 0xa2, 0xde, 0xf9, 0xde, 0x14, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x10, + }; + // 1 in Montgomery form + u32 m_inv [8] = { + 0x8d98951d, 0xd6ec3174, 0x737dcf70, 0xc6ef5bf4, + 0xfffffffe, 0xffffffff, 0xffffffff, 0x0fffffff, + }; + + u8 scalar[32]; + crypto_eddsa_trim_scalar(scalar, private_key); + + // Convert the scalar in Montgomery form + // m_scl = scalar * 2^256 (modulo L) + u32 m_scl[8]; + { + u32 tmp[16]; + ZERO(tmp, 8); + load32_le_buf(tmp+8, scalar, 8); + mod_l(scalar, tmp); + load32_le_buf(m_scl, scalar, 8); + WIPE_BUFFER(tmp); // Wipe ASAP to save stack space + } + + // Compute the inverse + u32 product[16]; + for (int i = 252; i >= 0; i--) { + ZERO(product, 16); + multiply(product, m_inv, m_inv); + redc(m_inv, product); + if (scalar_bit(Lm2, i)) { + ZERO(product, 16); + multiply(product, m_inv, m_scl); + redc(m_inv, product); + } + } + // Convert the inverse *out* of Montgomery form + // scalar = m_inv / 2^256 (modulo L) + COPY(product, m_inv, 8); + ZERO(product + 8, 8); + redc(m_inv, product); + store32_le_buf(scalar, m_inv, 8); // the *inverse* of the scalar + + // Clear the cofactor of scalar: + // cleared = scalar * (3*L + 1) (modulo 8*L) + // cleared = scalar + scalar * 3 * L (modulo 8*L) + // Note that (scalar * 3) is reduced modulo 8, so we only need the + // first byte. + add_xl(scalar, scalar[0] * 3); + + // Recall that 8*L < 2^256. However it is also very close to + // 2^255. If we spanned the ladder over 255 bits, random tests + // wouldn't catch the off-by-one error. + scalarmult(blind_salt, scalar, curve_point, 256); + + WIPE_BUFFER(scalar); WIPE_BUFFER(m_scl); + WIPE_BUFFER(product); WIPE_BUFFER(m_inv); +} + +//////////////////////////////// +/// Authenticated encryption /// +//////////////////////////////// +static void lock_auth(u8 mac[16], const u8 auth_key[32], + const u8 *ad , size_t ad_size, + const u8 *cipher_text, size_t text_size) +{ + u8 sizes[16]; // Not secret, not wiped + store64_le(sizes + 0, ad_size); + store64_le(sizes + 8, text_size); + crypto_poly1305_ctx poly_ctx; // auto wiped... + crypto_poly1305_init (&poly_ctx, auth_key); + crypto_poly1305_update(&poly_ctx, ad , ad_size); + crypto_poly1305_update(&poly_ctx, zero , align(ad_size, 16)); + crypto_poly1305_update(&poly_ctx, cipher_text, text_size); + crypto_poly1305_update(&poly_ctx, zero , align(text_size, 16)); + crypto_poly1305_update(&poly_ctx, sizes , 16); + crypto_poly1305_final (&poly_ctx, mac); // ...here +} + +void crypto_aead_init_x(crypto_aead_ctx *ctx, + u8 const key[32], const u8 nonce[24]) +{ + crypto_chacha20_h(ctx->key, key, nonce); + COPY(ctx->nonce, nonce + 16, 8); + ctx->counter = 0; +} + +void crypto_aead_init_djb(crypto_aead_ctx *ctx, + const u8 key[32], const u8 nonce[8]) +{ + COPY(ctx->key , key , 32); + COPY(ctx->nonce, nonce, 8); + ctx->counter = 0; +} + +void crypto_aead_init_ietf(crypto_aead_ctx *ctx, + const u8 key[32], const u8 nonce[12]) +{ + COPY(ctx->key , key , 32); + COPY(ctx->nonce, nonce + 4, 8); + ctx->counter = (u64)load32_le(nonce) << 32; +} + +void crypto_aead_write(crypto_aead_ctx *ctx, u8 *cipher_text, u8 mac[16], + const u8 *ad, size_t ad_size, + const u8 *plain_text, size_t text_size) +{ + u8 auth_key[64]; // the last 32 bytes are used for rekeying. + crypto_chacha20_djb(auth_key, 0, 64, ctx->key, ctx->nonce, ctx->counter); + crypto_chacha20_djb(cipher_text, plain_text, text_size, + ctx->key, ctx->nonce, ctx->counter + 1); + lock_auth(mac, auth_key, ad, ad_size, cipher_text, text_size); + COPY(ctx->key, auth_key + 32, 32); + WIPE_BUFFER(auth_key); +} + +int crypto_aead_read(crypto_aead_ctx *ctx, u8 *plain_text, const u8 mac[16], + const u8 *ad, size_t ad_size, + const u8 *cipher_text, size_t text_size) +{ + u8 auth_key[64]; // the last 32 bytes are used for rekeying. + u8 real_mac[16]; + crypto_chacha20_djb(auth_key, 0, 64, ctx->key, ctx->nonce, ctx->counter); + lock_auth(real_mac, auth_key, ad, ad_size, cipher_text, text_size); + int mismatch = crypto_verify16(mac, real_mac); + if (!mismatch) { + crypto_chacha20_djb(plain_text, cipher_text, text_size, + ctx->key, ctx->nonce, ctx->counter + 1); + COPY(ctx->key, auth_key + 32, 32); + } + WIPE_BUFFER(auth_key); + WIPE_BUFFER(real_mac); + return mismatch; +} + +void crypto_aead_lock(u8 *cipher_text, u8 mac[16], const u8 key[32], + const u8 nonce[24], const u8 *ad, size_t ad_size, + const u8 *plain_text, size_t text_size) +{ + crypto_aead_ctx ctx; + crypto_aead_init_x(&ctx, key, nonce); + crypto_aead_write(&ctx, cipher_text, mac, ad, ad_size, + plain_text, text_size); + crypto_wipe(&ctx, sizeof(ctx)); +} + +int crypto_aead_unlock(u8 *plain_text, const u8 mac[16], const u8 key[32], + const u8 nonce[24], const u8 *ad, size_t ad_size, + const u8 *cipher_text, size_t text_size) +{ + crypto_aead_ctx ctx; + crypto_aead_init_x(&ctx, key, nonce); + int mismatch = crypto_aead_read(&ctx, plain_text, mac, ad, ad_size, + cipher_text, text_size); + crypto_wipe(&ctx, sizeof(ctx)); + return mismatch; +} + +#ifdef MONOCYPHER_CPP_NAMESPACE +} +#endif diff --git a/src/monocypher/monocypher.h b/src/monocypher/monocypher.h new file mode 100644 index 000000000..d0afddc03 --- /dev/null +++ b/src/monocypher/monocypher.h @@ -0,0 +1,321 @@ +// Monocypher version 4.0.0 +// +// This file is dual-licensed. Choose whichever licence you want from +// the two licences listed below. +// +// The first licence is a regular 2-clause BSD licence. The second licence +// is the CC-0 from Creative Commons. It is intended to release Monocypher +// to the public domain. The BSD licence serves as a fallback option. +// +// SPDX-License-Identifier: BSD-2-Clause OR CC0-1.0 +// +// ------------------------------------------------------------------------ +// +// Copyright (c) 2017-2019, Loup Vaillant +// All rights reserved. +// +// +// Redistribution and use in source and binary forms, with or without +// modification, are permitted provided that the following conditions are +// met: +// +// 1. Redistributions of source code must retain the above copyright +// notice, this list of conditions and the following disclaimer. +// +// 2. Redistributions in binary form must reproduce the above copyright +// notice, this list of conditions and the following disclaimer in the +// documentation and/or other materials provided with the +// distribution. +// +// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +// HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +// +// ------------------------------------------------------------------------ +// +// Written in 2017-2019 by Loup Vaillant +// +// To the extent possible under law, the author(s) have dedicated all copyright +// and related neighboring rights to this software to the public domain +// worldwide. This software is distributed without any warranty. +// +// You should have received a copy of the CC0 Public Domain Dedication along +// with this software. If not, see +// + +#ifndef MONOCYPHER_H +#define MONOCYPHER_H + +#include +#include + +#ifdef MONOCYPHER_CPP_NAMESPACE +namespace MONOCYPHER_CPP_NAMESPACE { +#elif defined(__cplusplus) +extern "C" { +#endif + +// Constant time comparisons +// ------------------------- + +// Return 0 if a and b are equal, -1 otherwise +int crypto_verify16(const uint8_t a[16], const uint8_t b[16]); +int crypto_verify32(const uint8_t a[32], const uint8_t b[32]); +int crypto_verify64(const uint8_t a[64], const uint8_t b[64]); + + +// Erase sensitive data +// -------------------- +void crypto_wipe(void *secret, size_t size); + + +// Authenticated encryption +// ------------------------ +void crypto_aead_lock(uint8_t *cipher_text, + uint8_t mac [16], + const uint8_t key [32], + const uint8_t nonce[24], + const uint8_t *ad, size_t ad_size, + const uint8_t *plain_text, size_t text_size); +int crypto_aead_unlock(uint8_t *plain_text, + const uint8_t mac [16], + const uint8_t key [32], + const uint8_t nonce[24], + const uint8_t *ad, size_t ad_size, + const uint8_t *cipher_text, size_t text_size); + +// Authenticated stream +// -------------------- +typedef struct { + uint64_t counter; + uint8_t key[32]; + uint8_t nonce[8]; +} crypto_aead_ctx; + +void crypto_aead_init_x(crypto_aead_ctx *ctx, + const uint8_t key[32], const uint8_t nonce[24]); +void crypto_aead_init_djb(crypto_aead_ctx *ctx, + const uint8_t key[32], const uint8_t nonce[8]); +void crypto_aead_init_ietf(crypto_aead_ctx *ctx, + const uint8_t key[32], const uint8_t nonce[12]); + +void crypto_aead_write(crypto_aead_ctx *ctx, + uint8_t *cipher_text, + uint8_t mac[16], + const uint8_t *ad , size_t ad_size, + const uint8_t *plain_text, size_t text_size); +int crypto_aead_read(crypto_aead_ctx *ctx, + uint8_t *plain_text, + const uint8_t mac[16], + const uint8_t *ad , size_t ad_size, + const uint8_t *cipher_text, size_t text_size); + + +// General purpose hash (BLAKE2b) +// ------------------------------ + +// Direct interface +void crypto_blake2b(uint8_t *hash, size_t hash_size, + const uint8_t *message, size_t message_size); + +void crypto_blake2b_keyed(uint8_t *hash, size_t hash_size, + const uint8_t *key, size_t key_size, + const uint8_t *message, size_t message_size); + +// Incremental interface +typedef struct { + // Do not rely on the size or contents of this type, + // for they may change without notice. + uint64_t hash[8]; + uint64_t input_offset[2]; + uint64_t input[16]; + size_t input_idx; + size_t hash_size; +} crypto_blake2b_ctx; + +void crypto_blake2b_init(crypto_blake2b_ctx *ctx, size_t hash_size); +void crypto_blake2b_keyed_init(crypto_blake2b_ctx *ctx, size_t hash_size, + const uint8_t *key, size_t key_size); +void crypto_blake2b_update(crypto_blake2b_ctx *ctx, + const uint8_t *message, size_t message_size); +void crypto_blake2b_final(crypto_blake2b_ctx *ctx, uint8_t *hash); + + +// Password key derivation (Argon2) +// -------------------------------- +#define CRYPTO_ARGON2_D 0 +#define CRYPTO_ARGON2_I 1 +#define CRYPTO_ARGON2_ID 2 + +typedef struct { + uint32_t algorithm; // Argon2d, Argon2i, Argon2id + uint32_t nb_blocks; // memory hardness, >= 8 * nb_lanes + uint32_t nb_passes; // CPU hardness, >= 1 (>= 3 recommended for Argon2i) + uint32_t nb_lanes; // parallelism level (single threaded anyway) +} crypto_argon2_config; + +typedef struct { + const uint8_t *pass; + const uint8_t *salt; + uint32_t pass_size; + uint32_t salt_size; // 16 bytes recommended +} crypto_argon2_inputs; + +typedef struct { + const uint8_t *key; // may be NULL if no key + const uint8_t *ad; // may be NULL if no additional data + uint32_t key_size; // 0 if no key (32 bytes recommended otherwise) + uint32_t ad_size; // 0 if no additional data +} crypto_argon2_extras; + +extern const crypto_argon2_extras crypto_argon2_no_extras; + +void crypto_argon2(uint8_t *hash, uint32_t hash_size, void *work_area, + crypto_argon2_config config, + crypto_argon2_inputs inputs, + crypto_argon2_extras extras); + + +// Key exchange (X-25519) +// ---------------------- + +// Shared secrets are not quite random. +// Hash them to derive an actual shared key. +void crypto_x25519_public_key(uint8_t public_key[32], + const uint8_t secret_key[32]); +void crypto_x25519(uint8_t raw_shared_secret[32], + const uint8_t your_secret_key [32], + const uint8_t their_public_key [32]); + +// Conversion to EdDSA +void crypto_x25519_to_eddsa(uint8_t eddsa[32], const uint8_t x25519[32]); + +// scalar "division" +// Used for OPRF. Be aware that exponential blinding is less secure +// than Diffie-Hellman key exchange. +void crypto_x25519_inverse(uint8_t blind_salt [32], + const uint8_t private_key[32], + const uint8_t curve_point[32]); + +// "Dirty" versions of x25519_public_key(). +// Use with crypto_elligator_rev(). +// Leaks 3 bits of the private key. +void crypto_x25519_dirty_small(uint8_t pk[32], const uint8_t sk[32]); +void crypto_x25519_dirty_fast (uint8_t pk[32], const uint8_t sk[32]); + + +// Signatures +// ---------- + +// EdDSA with curve25519 + BLAKE2b +void crypto_eddsa_key_pair(uint8_t secret_key[64], + uint8_t public_key[32], + uint8_t seed[32]); +void crypto_eddsa_sign(uint8_t signature [64], + const uint8_t secret_key[64], + const uint8_t *message, size_t message_size); +int crypto_eddsa_check(const uint8_t signature [64], + const uint8_t public_key[32], + const uint8_t *message, size_t message_size); + +// Conversion to X25519 +void crypto_eddsa_to_x25519(uint8_t x25519[32], const uint8_t eddsa[32]); + +// EdDSA building blocks +void crypto_eddsa_trim_scalar(uint8_t out[32], const uint8_t in[32]); +void crypto_eddsa_reduce(uint8_t reduced[32], const uint8_t expanded[64]); +void crypto_eddsa_mul_add(uint8_t r[32], + const uint8_t a[32], + const uint8_t b[32], + const uint8_t c[32]); +void crypto_eddsa_scalarbase(uint8_t point[32], const uint8_t scalar[32]); +int crypto_eddsa_check_equation(const uint8_t signature[64], + const uint8_t public_key[32], + const uint8_t h_ram[32]); + + +// Chacha20 +// -------- + +// Specialised hash. +// Used to hash X25519 shared secrets. +void crypto_chacha20_h(uint8_t out[32], + const uint8_t key[32], + const uint8_t in [16]); + +// Unauthenticated stream cipher. +// Don't forget to add authentication. +uint64_t crypto_chacha20_djb(uint8_t *cipher_text, + const uint8_t *plain_text, + size_t text_size, + const uint8_t key[32], + const uint8_t nonce[8], + uint64_t ctr); +uint32_t crypto_chacha20_ietf(uint8_t *cipher_text, + const uint8_t *plain_text, + size_t text_size, + const uint8_t key[32], + const uint8_t nonce[12], + uint32_t ctr); +uint64_t crypto_chacha20_x(uint8_t *cipher_text, + const uint8_t *plain_text, + size_t text_size, + const uint8_t key[32], + const uint8_t nonce[24], + uint64_t ctr); + + +// Poly 1305 +// --------- + +// This is a *one time* authenticator. +// Disclosing the mac reveals the key. +// See crypto_lock() on how to use it properly. + +// Direct interface +void crypto_poly1305(uint8_t mac[16], + const uint8_t *message, size_t message_size, + const uint8_t key[32]); + +// Incremental interface +typedef struct { + // Do not rely on the size or contents of this type, + // for they may change without notice. + uint8_t c[16]; // chunk of the message + size_t c_idx; // How many bytes are there in the chunk. + uint32_t r [4]; // constant multiplier (from the secret key) + uint32_t pad[4]; // random number added at the end (from the secret key) + uint32_t h [5]; // accumulated hash +} crypto_poly1305_ctx; + +void crypto_poly1305_init (crypto_poly1305_ctx *ctx, const uint8_t key[32]); +void crypto_poly1305_update(crypto_poly1305_ctx *ctx, + const uint8_t *message, size_t message_size); +void crypto_poly1305_final (crypto_poly1305_ctx *ctx, uint8_t mac[16]); + + +// Elligator 2 +// ----------- + +// Elligator mappings proper +void crypto_elligator_map(uint8_t curve [32], const uint8_t hidden[32]); +int crypto_elligator_rev(uint8_t hidden[32], const uint8_t curve [32], + uint8_t tweak); + +// Easy to use key pair generation +void crypto_elligator_key_pair(uint8_t hidden[32], uint8_t secret_key[32], + uint8_t seed[32]); + +#ifdef __cplusplus +} +#endif + +#endif // MONOCYPHER_H diff --git a/src/p_saveg.c b/src/p_saveg.c index fdbcb48de..4770f0602 100644 --- a/src/p_saveg.c +++ b/src/p_saveg.c @@ -407,6 +407,8 @@ static void P_NetArchivePlayers(savebuffer_t *save) WRITEUINT8(save->p, players[i].sliptideZipDelay); WRITEUINT16(save->p, players[i].sliptideZipBoost); + WRITEMEM(save->p, players[i].public_key, PUBKEYLENGTH); + // respawnvars_t WRITEUINT8(save->p, players[i].respawn.state); WRITEUINT32(save->p, K_GetWaypointHeapIndex(players[i].respawn.wp)); @@ -787,6 +789,8 @@ static void P_NetUnArchivePlayers(savebuffer_t *save) players[i].sliptideZipDelay = READUINT8(save->p); players[i].sliptideZipBoost = READUINT16(save->p); + READMEM(save->p, players[i].public_key, PUBKEYLENGTH); + // respawnvars_t players[i].respawn.state = READUINT8(save->p); players[i].respawn.wp = (waypoint_t *)(size_t)READUINT32(save->p); diff --git a/src/stun.c b/src/stun.c index 722d32906..9845900af 100644 --- a/src/stun.c +++ b/src/stun.c @@ -84,7 +84,7 @@ STUN_node (void) return node; } -static void +void csprng ( void * const buffer, diff --git a/src/stun.h b/src/stun.h index 7fe6bf42a..17cfafa31 100644 --- a/src/stun.h +++ b/src/stun.h @@ -18,6 +18,8 @@ extern "C" { typedef void (*stun_callback_t)(UINT32 address); +void csprng (void * const buffer, const size_t size); + void STUN_bind (stun_callback_t); boolean STUN_got_response (const char * const buffer, const size_t size); diff --git a/src/typedef.h b/src/typedef.h index 04298277c..2683bff70 100644 --- a/src/typedef.h +++ b/src/typedef.h @@ -71,6 +71,11 @@ TYPEDEF (filesneededconfig_pak); TYPEDEF (doomdata_t); TYPEDEF (serverelem_t); TYPEDEF (rewind_t); +TYPEDEF (clientkey_pak); +TYPEDEF (serverchallenge_pak); +TYPEDEF (challengeall_pak); +TYPEDEF (responseall_pak); +TYPEDEF (resultsall_pak); // d_event.h TYPEDEF (event_t);