sebastian
/
xmr-stak-generic-opencl11
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

First!

This commit is contained in:
fireice-uk 2017-01-21 11:46:45 +00:00
commit f5b60562e8
86 changed files with 29401 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
.gitignore vendored Normal file
View File

@ -0,0 +1,5 @@
bin/
obj/
xmr-stak-amd.layout
xmr-stak-amd.depend
config-debug.txt

35
CMakeLists.txt Normal file
View File

@ -0,0 +1,35 @@
project(xmr-stak-amd)
cmake_minimum_required(VERSION 2.8.10)
if("${CMAKE_CXX_COMPILER_ID}" STREQUAL "GNU")
if (CMAKE_CXX_COMPILER_VERSION VERSION_LESS 5.1)
message(FATAL_ERROR "GCC version must be at least 5.1!")
endif()
endif()
find_library(MHTD NAMES microhttpd)
if("${MHTD}" STREQUAL "MHTD-NOTFOUND")
message(FATAL_ERROR "libmicrohttpd is required")
endif()
#set(CMAKE_VERBOSE_MAKEFILE ON)
set(CMAKE_CONFIGURATION_TYPES "RELEASE;STATIC")
if("${CMAKE_BUILD_TYPE}" STREQUAL "")
set(CMAKE_BUILD_TYPE RELEASE)
endif()
set(CMAKE_C_FLAGS "-DNDEBUG -march=westmere -O3 -m64 -s")
set(CMAKE_CXX_FLAGS "${CMAKE_C_FLAGS} -std=c++11")
set(CMAKE_EXE_LINKER_FLAGS_RELSEASE "")
set(CMAKE_EXE_LINKER_FLAGS_STATIC "-static-libgcc -static-libstdc++")
set(EXECUTABLE_OUTPUT_PATH "bin")
file(GLOB SOURCES "crypto/*.c" "crypto/*.cpp" "amd_gpu/*.c" "*.cpp")
add_executable(xmr-stak-amd ${SOURCES})
target_link_libraries(xmr-stak-amd pthread microhttpd OpenCL)

620
LICENSE Normal file
View File

@ -0,0 +1,620 @@
GNU GENERAL PUBLIC LICENSE
Version 3, 29 June 2007
Copyright (C) 2007 Free Software Foundation, Inc. <http://fsf.org/>
Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.
Preamble
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.

158
README.md Normal file
View File

@ -0,0 +1,158 @@
### XMR-Stak-AMD - Monero mining software
XMR-Stak is a universal Stratum pool miner. This is the AMD version.
#### Usage on Windows
1) Edit the config.txt file to enter your pool login and password.
2) Double click the exe file.
XMR-Stak should compile on any C++11 compliant compiler. Windows compiler is assumed to be MSVC 2015 CE. MSVC build environment is not vendored.
```
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
Windows binary release checksums
sha1sum xmr-stak-cpu.exe
32f551c891040eda2c25e18e6287665471a5a653 xmr-stak-cpu.exe
sha3sum xmr-stak-cpu.exe
ed12841738c899a3eb61f51787aa670c25b64ce3c5a626717e6a8f6b xmr-stak-cpu.exe
date
Mon 16 Jan 15:13:25 GMT 2017
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQEcBAEBCAAGBQJYfONvAAoJEPsk95p+1Bw02coH/0by+VMK76gnmpNjIxDcphkV
S1GG+f0sIAYUrGpoMCJTXbr7hU3Na+grTbt6xLM2Tb0xJjX4Mc47Cixajzy7+TTx
R2+CvBRl8LG9zob6JNiohvxD1+SK7RWDKWenFyDlr9BewgE/ArqZM+16BQBrLP9H
XIWy1wh/lcSYuS548tnUYdNOmEnR9TqA454M4r8PED85HSpNmvI+eG8fZ8OK471C
3yMupjYlAbiEBT+gE6bZwLeeCH9NO2gGeBAb31w8RBsMRjy+VvhFhTOoJwZbXj9e
sMUwNBu+fLVoilMVvp8SDpQ7Uw/WFT085N2eJiCCuEbHgFAwM3uwD6VHz3eXd0s=
=QJQj
-----END PGP SIGNATURE-----
```
#### Usage on Linux
```
sudo apt-get install ocl-icd-opencl-dev
sudo apt-get install libmicrohttpd-dev
cmake .
make
```
GCC version 5.1 or higher is required for full C++11 support. CMake release compile scripts, as well as CodeBlocks build environment for debug builds is included.
To do a static build for a system without gcc 5.1+
```
cmake -DCMAKE_BUILD_TYPE=STATIC
make
```
Note - cmake caches variables, so if you want to do a dynamic build later you need to specify '-DCMAKE_BUILD_TYPE=RELEASE'
#### CPU mining performance
Performance is nearly identical to the closed source paid miners. Here are some numbers:
* **I7-2600K** - 266 H/s
* **I7-6700** - 276 H/s (with a separate GPU miner)
* **Dual X5650** - 466 H/s (depends on NUMA)
* **Dual E5640** - 365 H/s (same as above)
#### Example reports
```
HASHRATE REPORT
| ID | 2.5s | 60s | 15m | ID | 2.5s | 60s | 15m |
| 0 | 31.7 | 30.7 | 30.5 | 1 | 30.6 | 30.6 | 30.6 |
| 2 | 30.3 | 30.6 | 30.6 | 3 | 30.6 | 30.6 | 30.6 |
| 4 | 35.3 | 35.5 | 35.6 | 5 | 35.7 | 35.7 | 35.7 |
| 6 | 35.4 | 35.6 | 35.6 | 7 | 35.7 | 35.7 | 35.7 |
| 8 | 31.7 | 30.7 | 30.5 | 9 | 30.6 | 30.6 | 30.6 |
| 10 | 30.4 | 30.6 | 30.6 | 11 | 30.6 | 30.6 | 30.6 |
-----------------------------------------------------
Totals: 388.7 388.7 388.7 H/s
Highest: 388.7 H/s
```
```
RESULT REPORT
Difficulty : 8192
Good results : 5825 / 5826 (100.0 %)
Avg result time : 10.3 sec
Pool-side hashes : 22683648
Top 10 best results found:
| 0 | 15407238 | 1 | 12699745 |
| 2 | 12194202 | 3 | 6999845 |
| 4 | 5533935 | 5 | 5315338 |
| 6 | 4700351 | 7 | 4500227 |
| 8 | 4023567 | 9 | 4021473 |
Error details:
| Count | Error text | Last seen |
| 1 | [NETWORK ERROR] | 2017-01-02 21:29:15 |
```
```
CONNECTION REPORT
Connected since : 2017-01-02 21:29:40
Pool ping time : 288 ms
Network error log:
| Date | Error text |
| 2017-01-02 21:29:15 | CALL error: Timeout while waiting for a reply |
| 2017-01-02 21:29:30 | CONNECT error: GetAddrInfo: Name or service not known |
```
#### Default dev donation
By default the miner will donate 1% of the hashpower (1 minute in 100 minutes) to my pool. If you want to change that, edit **donate-level.h** before you build the binaries.
If you want to donate directly to support further development, here is my wallet
* 4581HhZkQHgZrZjKeCfCJxZff9E3xCgHGF25zABZz7oR71TnbbgiS7sK9jveE6Dx6uMs2LwszDuvQJgRZQotdpHt1fTdDhk
#### PGP Key
```
-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: GnuPG v2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=B5I+
-----END PGP PUBLIC KEY BLOCK-----
```
### Common Issues
**msvcp140.dll and vcruntime140.dll not available errors**
Download and install this [runtime package](https://www.microsoft.com/en-us/download/details.aspx?id=48145) from Microsoft. *Warning: Do NOT use "missing dll" sites - dll's are exe files with another name, and it is a fairly safe bet that any dll on a shady site like that will be trojaned. Please download offical runtimes from Microsoft above.*

668
amd_gpu/gpu.c Normal file
View File

@ -0,0 +1,668 @@
/*
* 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
* 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 <http://www.gnu.org/licenses/>.
*/
#include <stdio.h>
#include <string.h>
#ifdef _WIN32
#include <windows.h>
const char* sSourcePath = "opencl\\cryptonight.cl";
static inline void port_sleep(size_t sec)
{
Sleep(sec * 1000);
}
#else
#include <unistd.h>
const char* sSourcePath = "opencl/cryptonight.cl";
static inline void port_sleep(size_t sec)
{
sleep(sec);
}
#endif // _WIN32
#include "gpu.h"
const char* err_to_str(cl_int ret)
{
switch(ret)
{
case CL_SUCCESS:
return "CL_SUCCESS";
case CL_DEVICE_NOT_FOUND:
return "CL_DEVICE_NOT_FOUND";
case CL_DEVICE_NOT_AVAILABLE:
return "CL_DEVICE_NOT_AVAILABLE";
case CL_COMPILER_NOT_AVAILABLE:
return "CL_COMPILER_NOT_AVAILABLE";
case CL_MEM_OBJECT_ALLOCATION_FAILURE:
return "CL_MEM_OBJECT_ALLOCATION_FAILURE";
case CL_OUT_OF_RESOURCES:
return "CL_OUT_OF_RESOURCES";
case CL_OUT_OF_HOST_MEMORY:
return "CL_OUT_OF_HOST_MEMORY";
case CL_PROFILING_INFO_NOT_AVAILABLE:
return "CL_PROFILING_INFO_NOT_AVAILABLE";
case CL_MEM_COPY_OVERLAP:
return "CL_MEM_COPY_OVERLAP";
case CL_IMAGE_FORMAT_MISMATCH:
return "CL_IMAGE_FORMAT_MISMATCH";
case CL_IMAGE_FORMAT_NOT_SUPPORTED:
return "CL_IMAGE_FORMAT_NOT_SUPPORTED";
case CL_BUILD_PROGRAM_FAILURE:
return "CL_BUILD_PROGRAM_FAILURE";
case CL_MAP_FAILURE:
return "CL_MAP_FAILURE";
case CL_MISALIGNED_SUB_BUFFER_OFFSET:
return "CL_MISALIGNED_SUB_BUFFER_OFFSET";
case CL_EXEC_STATUS_ERROR_FOR_EVENTS_IN_WAIT_LIST:
return "CL_EXEC_STATUS_ERROR_FOR_EVENTS_IN_WAIT_LIST";
case CL_COMPILE_PROGRAM_FAILURE:
return "CL_COMPILE_PROGRAM_FAILURE";
case CL_LINKER_NOT_AVAILABLE:
return "CL_LINKER_NOT_AVAILABLE";
case CL_LINK_PROGRAM_FAILURE:
return "CL_LINK_PROGRAM_FAILURE";
case CL_DEVICE_PARTITION_FAILED:
return "CL_DEVICE_PARTITION_FAILED";
case CL_KERNEL_ARG_INFO_NOT_AVAILABLE:
return "CL_KERNEL_ARG_INFO_NOT_AVAILABLE";
case CL_INVALID_VALUE:
return "CL_INVALID_VALUE";
case CL_INVALID_DEVICE_TYPE:
return "CL_INVALID_DEVICE_TYPE";
case CL_INVALID_PLATFORM:
return "CL_INVALID_PLATFORM";
case CL_INVALID_DEVICE:
return "CL_INVALID_DEVICE";
case CL_INVALID_CONTEXT:
return "CL_INVALID_CONTEXT";
case CL_INVALID_QUEUE_PROPERTIES:
return "CL_INVALID_QUEUE_PROPERTIES";
case CL_INVALID_COMMAND_QUEUE:
return "CL_INVALID_COMMAND_QUEUE";
case CL_INVALID_HOST_PTR:
return "CL_INVALID_HOST_PTR";
case CL_INVALID_MEM_OBJECT:
return "CL_INVALID_MEM_OBJECT";
case CL_INVALID_IMAGE_FORMAT_DESCRIPTOR:
return "CL_INVALID_IMAGE_FORMAT_DESCRIPTOR";
case CL_INVALID_IMAGE_SIZE:
return "CL_INVALID_IMAGE_SIZE";
case CL_INVALID_SAMPLER:
return "CL_INVALID_SAMPLER";
case CL_INVALID_BINARY:
return "CL_INVALID_BINARY";
case CL_INVALID_BUILD_OPTIONS:
return "CL_INVALID_BUILD_OPTIONS";
case CL_INVALID_PROGRAM:
return "CL_INVALID_PROGRAM";
case CL_INVALID_PROGRAM_EXECUTABLE:
return "CL_INVALID_PROGRAM_EXECUTABLE";
case CL_INVALID_KERNEL_NAME:
return "CL_INVALID_KERNEL_NAME";
case CL_INVALID_KERNEL_DEFINITION:
return "CL_INVALID_KERNEL_DEFINITION";
case CL_INVALID_KERNEL:
return "CL_INVALID_KERNEL";
case CL_INVALID_ARG_INDEX:
return "CL_INVALID_ARG_INDEX";
case CL_INVALID_ARG_VALUE:
return "CL_INVALID_ARG_VALUE";
case CL_INVALID_ARG_SIZE:
return "CL_INVALID_ARG_SIZE";
case CL_INVALID_KERNEL_ARGS:
return "CL_INVALID_KERNEL_ARGS";
case CL_INVALID_WORK_DIMENSION:
return "CL_INVALID_WORK_DIMENSION";
case CL_INVALID_WORK_GROUP_SIZE:
return "CL_INVALID_WORK_GROUP_SIZE";
case CL_INVALID_WORK_ITEM_SIZE:
return "CL_INVALID_WORK_ITEM_SIZE";
case CL_INVALID_GLOBAL_OFFSET:
return "CL_INVALID_GLOBAL_OFFSET";
case CL_INVALID_EVENT_WAIT_LIST:
return "CL_INVALID_EVENT_WAIT_LIST";
case CL_INVALID_EVENT:
return "CL_INVALID_EVENT";
case CL_INVALID_OPERATION:
return "CL_INVALID_OPERATION";
case CL_INVALID_GL_OBJECT:
return "CL_INVALID_GL_OBJECT";
case CL_INVALID_BUFFER_SIZE:
return "CL_INVALID_BUFFER_SIZE";
case CL_INVALID_MIP_LEVEL:
return "CL_INVALID_MIP_LEVEL";
case CL_INVALID_GLOBAL_WORK_SIZE:
return "CL_INVALID_GLOBAL_WORK_SIZE";
case CL_INVALID_PROPERTY:
return "CL_INVALID_PROPERTY";
case CL_INVALID_IMAGE_DESCRIPTOR:
return "CL_INVALID_IMAGE_DESCRIPTOR";
case CL_INVALID_COMPILER_OPTIONS:
return "CL_INVALID_COMPILER_OPTIONS";
case CL_INVALID_LINKER_OPTIONS:
return "CL_INVALID_LINKER_OPTIONS";
case CL_INVALID_DEVICE_PARTITION_COUNT:
return "CL_INVALID_DEVICE_PARTITION_COUNT";
case CL_INVALID_PIPE_SIZE:
return "CL_INVALID_PIPE_SIZE";
case CL_INVALID_DEVICE_QUEUE:
return "CL_INVALID_DEVICE_QUEUE";
default:
return "UNKNOWN_ERROR";
}
}
void printer_print_msg(const char* fmt, ...);
char* LoadTextFile(const char* filename)
{
size_t flen;
char* out;
FILE* kernel = fopen(filename, "rb");
if(kernel == NULL)
return NULL;
fseek(kernel, 0, SEEK_END);
flen = ftell(kernel);
fseek(kernel, 0, SEEK_SET);
out = (char*)malloc(flen+1);
size_t r = fread(out, flen, 1, kernel);
fclose(kernel);
if(r != 1)
{
free(out);
return NULL;
}
out[flen] = '\0';
return out;
}
size_t InitOpenCLGpu(cl_context opencl_ctx, GpuContext* ctx, char* source_code)
{
size_t MaximumWorkSize;
cl_int ret;
if((clGetDeviceInfo(ctx->DeviceID, CL_DEVICE_MAX_WORK_GROUP_SIZE, sizeof(size_t), &MaximumWorkSize, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when querying a device's max worksize using clGetDeviceInfo.", err_to_str(ret));
return ERR_OCL_API;
}
printer_print_msg("Device %lu work size %lu / %lu.", ctx->deviceIdx, ctx->workSize, MaximumWorkSize);
#ifdef CL_VERSION_2_0
const cl_queue_properties CommandQueueProperties[] = { 0, 0, 0 };
ctx->CommandQueues = clCreateCommandQueueWithProperties(opencl_ctx, ctx->DeviceID, CommandQueueProperties, &ret);
#else
const cl_command_queue_properties CommandQueueProperties = { 0 };
ctx->CommandQueues = clCreateCommandQueue(opencl_ctx, ctx->DeviceID, CommandQueueProperties, &ret);
#endif
if(ret != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clCreateCommandQueueWithProperties.", err_to_str(ret));
return ERR_OCL_API;
}
ctx->InputBuffer = clCreateBuffer(opencl_ctx, CL_MEM_READ_ONLY, 88, NULL, &ret);
if(ret != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clCreateBuffer to create input buffer.", err_to_str(ret));
return ERR_OCL_API;
}
size_t g_thd = ctx->rawIntensity;
ctx->ExtraBuffers[0] = clCreateBuffer(opencl_ctx, CL_MEM_READ_WRITE, (1 << 21) * g_thd, NULL, &ret);
if(ret != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clCreateBuffer to create hash scratchpads buffer.", err_to_str(ret));
return ERR_OCL_API;
}
ctx->ExtraBuffers[1] = clCreateBuffer(opencl_ctx, CL_MEM_READ_WRITE, 200 * g_thd, NULL, &ret);
if(ret != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clCreateBuffer to create hash states buffer.", err_to_str(ret));
return ERR_OCL_API;
}
// Blake-256 branches
ctx->ExtraBuffers[2] = clCreateBuffer(opencl_ctx, CL_MEM_READ_WRITE, sizeof(cl_uint) * (g_thd + 2), NULL, &ret);
if(ret != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clCreateBuffer to create Branch 0 buffer.", err_to_str(ret));
return ERR_OCL_API;
}
// Groestl-256 branches
ctx->ExtraBuffers[3] = clCreateBuffer(opencl_ctx, CL_MEM_READ_WRITE, sizeof(cl_uint) * (g_thd + 2), NULL, &ret);
if(ret != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clCreateBuffer to create Branch 1 buffer.", err_to_str(ret));
return ERR_OCL_API;
}
// JH-256 branches
ctx->ExtraBuffers[4] = clCreateBuffer(opencl_ctx, CL_MEM_READ_WRITE, sizeof(cl_uint) * (g_thd + 2), NULL, &ret);
if(ret != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clCreateBuffer to create Branch 2 buffer.", err_to_str(ret));
return ERR_OCL_API;
}
// Skein-512 branches
ctx->ExtraBuffers[5] = clCreateBuffer(opencl_ctx, CL_MEM_READ_WRITE, sizeof(cl_uint) * (g_thd + 2), NULL, &ret);
if(ret != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clCreateBuffer to create Branch 3 buffer.", err_to_str(ret));
return ERR_OCL_API;
}
// Assume we may find up to 0xFF nonces in one run - it's reasonable
ctx->OutputBuffer = clCreateBuffer(opencl_ctx, CL_MEM_READ_WRITE, sizeof(cl_uint) * 0x100, NULL, &ret);
if(ret != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clCreateBuffer to create output buffer.", err_to_str(ret));
return ERR_OCL_API;
}
ctx->Program = clCreateProgramWithSource(opencl_ctx, 1, (const char**)&source_code, NULL, &ret);
if(ret != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clCreateProgramWithSource on the contents of cryptonight.cl", err_to_str(ret));
return ERR_OCL_API;
}
char options[32];
snprintf(options, sizeof(options), "-I. -DWORKSIZE=%lu", ctx->workSize);
ret = clBuildProgram(ctx->Program, 1, &ctx->DeviceID, options, NULL, NULL);
if(ret != CL_SUCCESS)
{
size_t len;
printer_print_msg("Error %s when calling clBuildProgram.", err_to_str(ret));
if((ret = clGetProgramBuildInfo(ctx->Program, ctx->DeviceID, CL_PROGRAM_BUILD_LOG, 0, NULL, &len)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clGetProgramBuildInfo for length of build log output.", err_to_str(ret));
return ERR_OCL_API;
}
char* BuildLog = (char*)malloc(len + 1);
if((ret = clGetProgramBuildInfo(ctx->Program, ctx->DeviceID, CL_PROGRAM_BUILD_LOG, len, BuildLog, NULL)) != CL_SUCCESS)
{
free(BuildLog);
printer_print_msg("Error %s when calling clGetProgramBuildInfo for build log.", err_to_str(ret));
return ERR_OCL_API;
}
printer_print_msg("Build Log:\n%s", BuildLog);
free(BuildLog);
return ERR_OCL_API;
}
cl_build_status status;
do
{
if((ret = clGetProgramBuildInfo(ctx->Program, ctx->DeviceID, CL_PROGRAM_BUILD_STATUS, sizeof(cl_build_status), &status, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clGetProgramBuildInfo for status of build.", err_to_str(ret));
return ERR_OCL_API;
}
port_sleep(1);
}
while(status == CL_BUILD_IN_PROGRESS);
const char *KernelNames[] = { "cn0", "cn1", "cn2", "Blake", "Groestl", "JH", "Skein" };
for(int i = 0; i < 7; ++i)
{
ctx->Kernels[i] = clCreateKernel(ctx->Program, KernelNames[i], &ret);
if(ret != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clCreateKernel for kernel %s.", err_to_str(ret), KernelNames[i]);
return ERR_OCL_API;
}
}
ctx->Nonce = 0;
return 0;
}
// RequestedDeviceIdxs is a list of OpenCL device indexes
// NumDevicesRequested is number of devices in RequestedDeviceIdxs list
// Returns 0 on success, -1 on stupid params, -2 on OpenCL API error
size_t InitOpenCL(GpuContext* ctx, size_t num_gpus, size_t platform_idx)
{
cl_context opencl_ctx;
cl_int ret;
cl_uint entries;
if((ret = clGetPlatformIDs(0, NULL, &entries)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clGetPlatformIDs for number of platforms.", err_to_str(ret));
return ERR_OCL_API;
}
// The number of platforms naturally is the index of the last platform plus one.
if(entries <= platform_idx)
{
printer_print_msg("Selected OpenCL platform index %d doesn't exist.", platform_idx);
return ERR_STUPID_PARAMS;
}
cl_platform_id PlatformIDList[entries];
if((ret = clGetPlatformIDs(entries, PlatformIDList, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clGetPlatformIDs for platform ID information.", err_to_str(ret));
return ERR_OCL_API;
}
if((ret = clGetDeviceIDs(PlatformIDList[platform_idx], CL_DEVICE_TYPE_GPU, 0, NULL, &entries)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clGetDeviceIDs for number of devices.", ret);
return ERR_OCL_API;
}
// Same as the platform index sanity check, except we must check all requested device indexes
for(int i = 0; i < num_gpus; ++i)
{
if(entries <= ctx[i].deviceIdx)
{
printer_print_msg("Selected OpenCL device index %lu doesn't exist.\n", ctx[i].deviceIdx);
return ERR_STUPID_PARAMS;
}
}
cl_device_id DeviceIDList[entries];
if((ret = clGetDeviceIDs(PlatformIDList[platform_idx], CL_DEVICE_TYPE_GPU, entries, DeviceIDList, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clGetDeviceIDs for device ID information.", err_to_str(ret));
return ERR_OCL_API;
}
// Indexes sanity checked above
cl_device_id TempDeviceList[num_gpus];
for(int i = 0; i < num_gpus; ++i)
{
ctx[i].DeviceID = DeviceIDList[ctx[i].deviceIdx];
TempDeviceList[i] = DeviceIDList[ctx[i].deviceIdx];
}
opencl_ctx = clCreateContext(NULL, num_gpus, TempDeviceList, NULL, NULL, &ret);
if(ret != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clCreateContext.", err_to_str(ret));
return ERR_OCL_API;
}
char* source_code = LoadTextFile(sSourcePath);
for(int i = 0; i < num_gpus; ++i)
{
if((ret = InitOpenCLGpu(opencl_ctx, &ctx[i], source_code)) != ERR_SUCCESS)
{
free(source_code);
return ret;
}
}
free(source_code);
return ERR_SUCCESS;
}
size_t XMRSetJob(GpuContext* ctx, uint8_t* input, size_t input_len, uint32_t target)
{
cl_int ret;
if(input_len > 84)
return ERR_STUPID_PARAMS;
input[input_len] = 0x01;
memset(input + input_len + 1, 0, 88 - input_len - 1);
if((ret = clEnqueueWriteBuffer(ctx->CommandQueues, ctx->InputBuffer, CL_TRUE, 0, 88, input, 0, NULL, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clEnqueueWriteBuffer to fill input buffer.", err_to_str(ret));
return ERR_OCL_API;
}
if((ret = clSetKernelArg(ctx->Kernels[0], 0, sizeof(cl_mem), &ctx->InputBuffer)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel 0, argument 0.", err_to_str(ret));
return ERR_OCL_API;
}
// Scratchpads
if((ret = clSetKernelArg(ctx->Kernels[0], 1, sizeof(cl_mem), ctx->ExtraBuffers + 0)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel 0, argument 1.", err_to_str(ret));
return ERR_OCL_API;
}
// States
if((ret = clSetKernelArg(ctx->Kernels[0], 2, sizeof(cl_mem), ctx->ExtraBuffers + 1)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel 0, argument 2.", err_to_str(ret));
return ERR_OCL_API;
}
// CN2 Kernel
// Scratchpads
if((ret = clSetKernelArg(ctx->Kernels[1], 0, sizeof(cl_mem), ctx->ExtraBuffers + 0)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel 1, argument 0.", err_to_str(ret));
return ERR_OCL_API;
}
// States
if((ret = clSetKernelArg(ctx->Kernels[1], 1, sizeof(cl_mem), ctx->ExtraBuffers + 1)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel 1, argument 1.", err_to_str(ret));
return ERR_OCL_API;
}
// CN3 Kernel
// Scratchpads
if((ret = clSetKernelArg(ctx->Kernels[2], 0, sizeof(cl_mem), ctx->ExtraBuffers + 0)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel 2, argument 0.", err_to_str(ret));
return ERR_OCL_API;
}
// States
if((ret = clSetKernelArg(ctx->Kernels[2], 1, sizeof(cl_mem), ctx->ExtraBuffers + 1)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel 2, argument 1.", err_to_str(ret));
return ERR_OCL_API;
}
// Branch 0
if((ret = clSetKernelArg(ctx->Kernels[2], 2, sizeof(cl_mem), ctx->ExtraBuffers + 2)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel 2, argument 2.", err_to_str(ret));
return ERR_OCL_API;
}
// Branch 1
if((ret = clSetKernelArg(ctx->Kernels[2], 3, sizeof(cl_mem), ctx->ExtraBuffers + 3)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel 2, argument 3.", err_to_str(ret));
return ERR_OCL_API;
}
// Branch 2
if((ret = clSetKernelArg(ctx->Kernels[2], 4, sizeof(cl_mem), ctx->ExtraBuffers + 4)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel 2, argument 4.", err_to_str(ret));
return ERR_OCL_API;
}
// Branch 3
if((ret = clSetKernelArg(ctx->Kernels[2], 5, sizeof(cl_mem), ctx->ExtraBuffers + 5)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel 2, argument 5.", err_to_str(ret));
return ERR_OCL_API;
}
for(int i = 0; i < 4; ++i)
{
// States
if((ret = clSetKernelArg(ctx->Kernels[i + 3], 0, sizeof(cl_mem), ctx->ExtraBuffers + 1)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel %d, argument %d.", err_to_str(ret), i + 3, 0);
return ERR_OCL_API;
}
// Nonce buffer
if((ret = clSetKernelArg(ctx->Kernels[i + 3], 1, sizeof(cl_mem), ctx->ExtraBuffers + (i + 2))) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel %d, argument %d.", err_to_str(ret), i + 3, 1);
return ERR_OCL_API;
}
// Output
if((ret = clSetKernelArg(ctx->Kernels[i + 3], 2, sizeof(cl_mem), &ctx->OutputBuffer)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel %d, argument %d.", err_to_str(ret), i + 3, 2);
return ERR_OCL_API;
}
// Target
if((ret = clSetKernelArg(ctx->Kernels[i + 3], 3, sizeof(cl_uint), &target)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel %d, argument %d.", err_to_str(ret), i + 3, 3);
return ERR_OCL_API;
}
}
return ERR_SUCCESS;
}
size_t XMRRunJob(GpuContext* ctx, cl_uint* HashOutput)
{
cl_int ret;
cl_uint zero = 0;
size_t BranchNonces[4] = {0};
size_t g_thd = ctx->rawIntensity;
size_t w_size = ctx->workSize;
for(int i = 2; i < 6; ++i)
{
if((ret = clEnqueueWriteBuffer(ctx->CommandQueues, ctx->ExtraBuffers[i], CL_FALSE, sizeof(cl_uint) * g_thd, sizeof(cl_uint), &zero, 0, NULL, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clEnqueueWriteBuffer to zero branch buffer counter %d.", err_to_str(ret), i - 2);
return ERR_OCL_API;
}
}
if((ret = clEnqueueWriteBuffer(ctx->CommandQueues, ctx->OutputBuffer, CL_FALSE, sizeof(cl_uint) * 0xFF, sizeof(cl_uint), &zero, 0, NULL, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clEnqueueReadBuffer to fetch results.", err_to_str(ret));
return ERR_OCL_API;
}
clFinish(ctx->CommandQueues);
size_t Nonce[2] = {ctx->Nonce, 1}, gthreads[2] = { g_thd, 8 }, lthreads[2] = { w_size, 8 };
if((ret = clEnqueueNDRangeKernel(ctx->CommandQueues, ctx->Kernels[0], 2, Nonce, gthreads, lthreads, 0, NULL, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clEnqueueNDRangeKernel for kernel %d.", err_to_str(ret), 0);
return ERR_OCL_API;
}
/*for(int i = 1; i < 3; ++i)
{
if((ret = clEnqueueNDRangeKernel(*ctx->CommandQueues, ctx->Kernels[i], 1, &ctx->Nonce, &g_thd, &w_size, 0, NULL, NULL)) != CL_SUCCESS)
{
Log(LOG_CRITICAL, "Error %s when calling clEnqueueNDRangeKernel for kernel %d.", err_to_str(ret), i);
return(ERR_OCL_API);
}
}*/
if((ret = clEnqueueNDRangeKernel(ctx->CommandQueues, ctx->Kernels[1], 1, &ctx->Nonce, &g_thd, &w_size, 0, NULL, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clEnqueueNDRangeKernel for kernel %d.", err_to_str(ret), 1);
return ERR_OCL_API;
}
if((ret = clEnqueueNDRangeKernel(ctx->CommandQueues, ctx->Kernels[2], 2, Nonce, gthreads, lthreads, 0, NULL, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clEnqueueNDRangeKernel for kernel %d.", err_to_str(ret), 2);
return ERR_OCL_API;
}
if((ret = clEnqueueReadBuffer(ctx->CommandQueues, ctx->ExtraBuffers[2], CL_FALSE, sizeof(cl_uint) * g_thd, sizeof(cl_uint), BranchNonces, 0, NULL, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clEnqueueReadBuffer to fetch results.", err_to_str(ret));
return ERR_OCL_API;
}
if((ret = clEnqueueReadBuffer(ctx->CommandQueues, ctx->ExtraBuffers[3], CL_FALSE, sizeof(cl_uint) * g_thd, sizeof(cl_uint), BranchNonces + 1, 0, NULL, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clEnqueueReadBuffer to fetch results.", err_to_str(ret));
return ERR_OCL_API;
}
if((ret = clEnqueueReadBuffer(ctx->CommandQueues, ctx->ExtraBuffers[4], CL_FALSE, sizeof(cl_uint) * g_thd, sizeof(cl_uint), BranchNonces + 2, 0, NULL, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clEnqueueReadBuffer to fetch results.", err_to_str(ret));
return ERR_OCL_API;
}
if((ret = clEnqueueReadBuffer(ctx->CommandQueues, ctx->ExtraBuffers[5], CL_FALSE, sizeof(cl_uint) * g_thd, sizeof(cl_uint), BranchNonces + 3, 0, NULL, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clEnqueueReadBuffer to fetch results.", err_to_str(ret));
return ERR_OCL_API;
}
clFinish(ctx->CommandQueues);
for(int i = 0; i < 4; ++i)
{
if(BranchNonces[i])
{
// Threads
if((clSetKernelArg(ctx->Kernels[i + 3], 4, sizeof(cl_ulong), BranchNonces + i)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clSetKernelArg for kernel %d, argument %d.", err_to_str(ret), i + 3, 4);
return(ERR_OCL_API);
}
BranchNonces[i] += BranchNonces[i] + (w_size - (BranchNonces[i] & (w_size - 1)));
if((ret = clEnqueueNDRangeKernel(ctx->CommandQueues, ctx->Kernels[i + 3], 1, &ctx->Nonce, BranchNonces + i, &w_size, 0, NULL, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clEnqueueNDRangeKernel for kernel %d.", err_to_str(ret), i + 3);
return ERR_OCL_API;
}
}
}
if((ret = clEnqueueReadBuffer(ctx->CommandQueues, ctx->OutputBuffer, CL_TRUE, 0, sizeof(cl_uint) * 0x100, HashOutput, 0, NULL, NULL)) != CL_SUCCESS)
{
printer_print_msg("Error %s when calling clEnqueueReadBuffer to fetch results.", err_to_str(ret));
return ERR_OCL_API;
}
clFinish(ctx->CommandQueues);
ctx->Nonce += g_thd;
return ERR_SUCCESS;
}

39
amd_gpu/gpu.h Normal file
View File

@ -0,0 +1,39 @@
#pragma once
#include <CL/cl.h>
#include <stdint.h>
#define ERR_SUCCESS (0)
#define ERR_OCL_API (2)
#define ERR_STUPID_PARAMS (1)
#ifdef __cplusplus
extern "C" {
#endif // __cplusplus
typedef struct _GpuContext
{
/*Input vars*/
size_t deviceIdx;
size_t rawIntensity;
size_t workSize;
/*Output vars*/
cl_device_id DeviceID;
cl_command_queue CommandQueues;
cl_mem InputBuffer;
cl_mem OutputBuffer;
cl_mem ExtraBuffers[6];
cl_program Program;
cl_kernel Kernels[7];
size_t Nonce;
} GpuContext;
size_t InitOpenCL(GpuContext* ctx, size_t num_gpus, size_t platform_idx);
size_t XMRSetJob(GpuContext* ctx, uint8_t* input, size_t input_len, uint32_t target);
size_t XMRRunJob(GpuContext* ctx, cl_uint* HashOutput);
#ifdef __cplusplus
}
#endif // __cplusplus

119
cli-miner.cpp Normal file
View File

@ -0,0 +1,119 @@
/*
* 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
* 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 <http://www.gnu.org/licenses/>.
*/
#include "executor.h"
#include "minethd.h"
#include "jconf.h"
#include "console.h"
#include "donate-level.h"
#include "httpd.h"
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
//Do a press any key for the windows folk. *insert any key joke here*
#ifdef _WIN32
void win_exit()
{
printer::inst()->print_str("Press any key to exit.");
get_key();
return;
}
#define strcasecmp _stricmp
#else
void win_exit() { return; }
#endif // _WIN32
int main(int argc, char *argv[])
{
const char* sFilename = "config.txt";
if(argc >= 2)
{
if(strcmp(argv[1], "-h") == 0)
{
printer::inst()->print_msg(L0, "Usage %s [CONFIG FILE]", argv[0]);
win_exit();
return 0;
}
if(argc >= 3 && strcasecmp(argv[1], "-c") == 0)
sFilename = argv[2];
else
sFilename = argv[1];
}
if(!jconf::inst()->parse_config(sFilename))
{
win_exit();
return 0;
}
if(!minethd::init_gpus())
{
win_exit();
return 0;
}
if(jconf::inst()->GetHttpdPort() != 0)
{
if (!httpd::inst()->start_daemon())
{
win_exit();
return 0;
}
}
printer::inst()->print_str("-------------------------------------------------------------------\n");
printer::inst()->print_str("XMR-Stak-CPU mining software, CPU Version.\n");
printer::inst()->print_str("Based on CPU mining code by wolf9466 (heavily optimized by myself).\n");
printer::inst()->print_str("Brought to you by fireice_uk under GPLv3.\n\n");
char buffer[64];
snprintf(buffer, sizeof(buffer), "Configurable dev donation level is set to %.1f %%\n\n", fDevDonationLevel * 100.0);
printer::inst()->print_str(buffer);
printer::inst()->print_str("You can use following keys to display reports:\n");
printer::inst()->print_str("'h' - hashrate\n");
printer::inst()->print_str("'r' - results\n");
printer::inst()->print_str("'c' - connection\n");
printer::inst()->print_str("-------------------------------------------------------------------\n");
executor::inst()->ex_start();
int key;
while(true)
{
key = get_key();
switch(key)
{
case 'h':
executor::inst()->push_event(ex_event(EV_USR_HASHRATE));
break;
case 'r':
executor::inst()->push_event(ex_event(EV_USR_RESULTS));
break;
case 'c':
executor::inst()->push_event(ex_event(EV_USR_CONNSTAT));
break;
default:
break;
}
}
return 0;
}

88
config.txt Normal file
View File

@ -0,0 +1,88 @@
/*
* Number of GPUs that you have in your system. Each GPU will get its own CPU thread.
*/
"gpu_thread_num" : 6,
/*
* GPU configuration. You should play around with intensity and worksize as the fastest settings will vary.
* index - GPU index number usually starts from 0
* intensity - Number of parallel GPU threads (nothing to do with CPU threads)
* worksize - Number of local GPU threads (nothing to do with CPU threads)
* affine_to_cpu - This will affine the thread to a CPU. This can make a GPU miner play along nicer with a CPU miner.
*/
"gpu_threads_conf" : [
{ "index" : 0, "intensity" : 1000, "worksize" : 8, "affine_to_cpu" : false },
{ "index" : 1, "intensity" : 1000, "worksize" : 8, "affine_to_cpu" : false },
{ "index" : 2, "intensity" : 1000, "worksize" : 8, "affine_to_cpu" : false },
{ "index" : 3, "intensity" : 1000, "worksize" : 8, "affine_to_cpu" : false },
{ "index" : 4, "intensity" : 1000, "worksize" : 8, "affine_to_cpu" : false },
{ "index" : 5, "intensity" : 1000, "worksize" : 8, "affine_to_cpu" : false },
],
/*
* Platform index. This will be 0 unless you have different OpenCL platform - eg. AMD and Intel.
*/
"platform_index" : 0,
/*
* pool_address - Pool address should be in the form "pool.supportxmr.com:3333". Only stratum pools are supported.
* wallet_address - Your wallet, or pool login.
* pool_password - Can be empty in most cases or "x".
*/
"pool_address" : "pool.supportxmr.com:3333",
"wallet_address" : "",
"pool_password" : "",
/*
* Network timeouts.
* Because of the way this client is written it doesn't need to constantly talk (keep-alive) to the server to make
* sure it is there. We detect a buggy / overloaded server by the call timeout. The default values will be ok for
* nearly all cases. If they aren't the pool has most likely overload issues. Low call timeout values are preferable -
* long timeouts mean that we waste hashes on potentially stale jobs. Connection report will tell you how long the
* server usually takes to process our calls.
*
* call_timeout - How long should we wait for a response from the server before we assume it is dead and drop the connection.
* retry_time - How long should we wait before another connection attempt.
* Both values are in seconds.
*/
"call_timeout" : 10,
"retry_time" : 10,
/*
* Output control.
* Since most people are used to miners printing all the time, that's what we do by default too. This is suboptimal
* really, since you cannot see errors under pages and pages of text and performance stats. Given that we have internal
* performance monitors, there is very little reason to spew out pages of text instead of concise reports.
* Press 'h' (hashrate), 'r' (results) or 'c' (connection) to print reports.
*
* verbose_level - 0 - Don't print anything.
* 1 - Print intro, connection event, disconnect event
* 2 - All of level 1, and new job (block) event if the difficulty is different from the last job
* 3 - All of level 1, and new job (block) event in all cases, result submission event.
* 4 - All of level 3, and automatic hashrate report printing
*/
"verbose_level" : 3,
/*
* Automatic hashrate report
*
* h_print_time - How often, in seconds, should we print a hashrate report if verbose_level is set to 4.
* This option has no effect if verbose_level is not 4.
*/
"h_print_time" : 60,
/*
* Built-in web server
* I like checking my hashrate on my phone. Don't you?
* Keep in mind that you will need to set up port forwarding on your router if you want to access it from
* outside of your home network. Ports lower than 1024 on Linux systems will require root.
*
* httpd_port - Port we should listen on. Default, 0, will switch off the server.
*/
"httpd_port" : 0,
/*
* prefer_ipv4 - IPv6 preference. If the host is available on both IPv4 and IPv6 net, which one should be choose?
* This setting will only be needed in 2020's. No need to worry about it now.
*/
"prefer_ipv4" : true,

212
console.cpp Normal file
View File

@ -0,0 +1,212 @@
/*
* 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
* 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 <http://www.gnu.org/licenses/>.
*/
#include "console.h"
#include <time.h>
#include <stdio.h>
#include <string.h>
#include <stdarg.h>
#ifdef _WIN32
#include <windows.h>
int get_key()
{
DWORD mode, rd;
HANDLE h;
if ((h = GetStdHandle(STD_INPUT_HANDLE)) == NULL)
return -1;
GetConsoleMode( h, &mode );
SetConsoleMode( h, mode & ~(ENABLE_LINE_INPUT | ENABLE_ECHO_INPUT) );
int c = 0;
ReadConsole( h, &c, 1, &rd, NULL );
SetConsoleMode( h, mode );
return c;
}
void set_colour(out_colours cl)
{
WORD attr = 0;
switch(cl)
{
case K_RED:
attr = FOREGROUND_RED | FOREGROUND_INTENSITY;
break;
case K_GREEN:
attr = FOREGROUND_GREEN | FOREGROUND_INTENSITY;
break;
case K_BLUE:
attr = FOREGROUND_BLUE | FOREGROUND_INTENSITY;
break;
case K_YELLOW:
attr = FOREGROUND_RED | FOREGROUND_GREEN | FOREGROUND_INTENSITY;
break;
case K_CYAN:
attr = FOREGROUND_BLUE | FOREGROUND_GREEN | FOREGROUND_INTENSITY;
break;
case K_MAGENTA:
attr = FOREGROUND_BLUE | FOREGROUND_RED | FOREGROUND_INTENSITY;
break;
case K_WHITE:
attr = FOREGROUND_BLUE | FOREGROUND_RED | FOREGROUND_GREEN | FOREGROUND_INTENSITY;
break;
default:
break;
}
SetConsoleTextAttribute(GetStdHandle(STD_OUTPUT_HANDLE), attr);
}
void reset_colour()
{
SetConsoleTextAttribute(GetStdHandle(STD_OUTPUT_HANDLE), FOREGROUND_RED | FOREGROUND_GREEN | FOREGROUND_BLUE);
}
#else
#include <termios.h>
#include <unistd.h>
#include <stdio.h>
int get_key()
{
struct termios oldattr, newattr;
int ch;
tcgetattr( STDIN_FILENO, &oldattr );
newattr = oldattr;
newattr.c_lflag &= ~( ICANON | ECHO );
tcsetattr( STDIN_FILENO, TCSANOW, &newattr );
ch = getchar();
tcsetattr( STDIN_FILENO, TCSANOW, &oldattr );
return ch;
}
void set_colour(out_colours cl)
{
switch(cl)
{
case K_RED:
fputs("\x1B[1;31m", stdout);
break;
case K_GREEN:
fputs("\x1B[1;32m", stdout);
break;
case K_BLUE:
fputs("\x1B[1;34m", stdout);
break;
case K_YELLOW:
fputs("\x1B[1;33m", stdout);
break;
case K_CYAN:
fputs("\x1B[1;36m", stdout);
break;
case K_MAGENTA:
fputs("\x1B[1;35m", stdout);
break;
case K_WHITE:
fputs("\x1B[1;37m", stdout);
break;
default:
break;
}
}
void reset_colour()
{
fputs("\x1B[0m", stdout);
}
#endif // _WIN32
inline void comp_localtime(const time_t* ctime, tm* stime)
{
#ifdef _WIN32
localtime_s(stime, ctime);
#else
localtime_r(ctime, stime);
#endif // __WIN32
}
printer* printer::oInst = nullptr;
printer::printer()
{
verbose_level = LINF;
}
void printer::print_msg(verbosity verbose, const char* fmt, ...)
{
if(verbose > verbose_level)
return;
char buf[1024];
size_t bpos;
tm stime;
time_t now = time(nullptr);
comp_localtime(&now, &stime);
strftime(buf, sizeof(buf), "[%F %T] : ", &stime);
bpos = strlen(buf);
va_list args;
va_start(args, fmt);
vsnprintf(buf+bpos, sizeof(buf)-bpos, fmt, args);
va_end(args);
bpos = strlen(buf);
if(bpos+2 >= sizeof(buf))
return;
buf[bpos] = '\n';
buf[bpos+1] = '\0';
std::unique_lock<std::mutex> lck(print_mutex);
fputs(buf, stdout);
}
void printer::print_str(const char* str)
{
std::unique_lock<std::mutex> lck(print_mutex);
fputs(str, stdout);
}
extern "C" void printer_print_msg(const char* fmt, ...)
{
char buf[1024];
size_t bpos;
tm stime;
time_t now = time(nullptr);
comp_localtime(&now, &stime);
strftime(buf, sizeof(buf), "[%F %T] : ", &stime);
bpos = strlen(buf);
va_list args;
va_start(args, fmt);
vsnprintf(buf+bpos, sizeof(buf)-bpos, fmt, args);
va_end(args);
bpos = strlen(buf);
if(bpos+2 >= sizeof(buf))
return;
buf[bpos] = '\n';
buf[bpos+1] = '\0';
printer::inst()->print_str(buf);
}

41
console.h Normal file
View File

@ -0,0 +1,41 @@
#pragma once
#include <mutex>
enum out_colours { K_RED, K_GREEN, K_BLUE, K_YELLOW, K_CYAN, K_MAGENTA, K_WHITE, K_NONE };
// Warning - on Linux get_key will detect control keys, but not on Windows.
// We will only use it for alphanum keys anyway.
int get_key();
void set_colour(out_colours cl);
void reset_colour();
// on MSVC sizeof(long int) = 4, gcc sizeof(long int) = 8, this is the workaround
// now we can use %llu on both compilers
inline long long unsigned int int_port(size_t i)
{
return i;
}
enum verbosity : size_t { L0 = 0, L1 = 1, L2 = 2, L3 = 3, L4 = 4, LINF = 100};
class printer
{
public:
static inline printer* inst()
{
if (oInst == nullptr) oInst = new printer;
return oInst;
};
inline void set_verbose_level(size_t level) { verbose_level = (verbosity)level; }
void print_msg(verbosity verbose, const char* fmt, ...);
void print_str(const char* str);
private:
printer();
static printer* oInst;
std::mutex print_mutex;
verbosity verbose_level;
};

326
crypto/c_blake256.c Normal file
View File

@ -0,0 +1,326 @@
/*
* The blake256_* and blake224_* functions are largely copied from
* blake256_light.c and blake224_light.c from the BLAKE website:
*
* http://131002.net/blake/
*
* The hmac_* functions implement HMAC-BLAKE-256 and HMAC-BLAKE-224.
* HMAC is specified by RFC 2104.
*/
#include <string.h>
#include <stdio.h>
#include <stdint.h>
#include "c_blake256.h"
#define U8TO32(p) \
(((uint32_t)((p)[0]) << 24) | ((uint32_t)((p)[1]) << 16) | \
((uint32_t)((p)[2]) << 8) | ((uint32_t)((p)[3]) ))
#define U32TO8(p, v) \
(p)[0] = (uint8_t)((v) >> 24); (p)[1] = (uint8_t)((v) >> 16); \
(p)[2] = (uint8_t)((v) >> 8); (p)[3] = (uint8_t)((v) );
const uint8_t sigma[][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},
{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}
};
const uint32_t cst[16] = {
0x243F6A88, 0x85A308D3, 0x13198A2E, 0x03707344,
0xA4093822, 0x299F31D0, 0x082EFA98, 0xEC4E6C89,
0x452821E6, 0x38D01377, 0xBE5466CF, 0x34E90C6C,
0xC0AC29B7, 0xC97C50DD, 0x3F84D5B5, 0xB5470917
};
static const uint8_t padding[] = {
0x80,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0
};
void blake256_compress(state *S, const uint8_t *block) {
uint32_t v[16], m[16], i;
#define ROT(x,n) (((x)<<(32-n))|((x)>>(n)))
#define G(a,b,c,d,e) \
v[a] += (m[sigma[i][e]] ^ cst[sigma[i][e+1]]) + v[b]; \
v[d] = ROT(v[d] ^ v[a],16); \
v[c] += v[d]; \
v[b] = ROT(v[b] ^ v[c],12); \
v[a] += (m[sigma[i][e+1]] ^ cst[sigma[i][e]])+v[b]; \
v[d] = ROT(v[d] ^ v[a], 8); \
v[c] += v[d]; \
v[b] = ROT(v[b] ^ v[c], 7);
for (i = 0; i < 16; ++i) m[i] = U8TO32(block + i * 4);
for (i = 0; i < 8; ++i) v[i] = S->h[i];
v[ 8] = S->s[0] ^ 0x243F6A88;
v[ 9] = S->s[1] ^ 0x85A308D3;
v[10] = S->s[2] ^ 0x13198A2E;
v[11] = S->s[3] ^ 0x03707344;
v[12] = 0xA4093822;
v[13] = 0x299F31D0;
v[14] = 0x082EFA98;
v[15] = 0xEC4E6C89;
if (S->nullt == 0) {
v[12] ^= S->t[0];
v[13] ^= S->t[0];
v[14] ^= S->t[1];
v[15] ^= S->t[1];
}
for (i = 0; i < 14; ++i) {
G(0, 4, 8, 12, 0);
G(1, 5, 9, 13, 2);
G(2, 6, 10, 14, 4);
G(3, 7, 11, 15, 6);
G(3, 4, 9, 14, 14);
G(2, 7, 8, 13, 12);
G(0, 5, 10, 15, 8);
G(1, 6, 11, 12, 10);
}
for (i = 0; i < 16; ++i) S->h[i % 8] ^= v[i];
for (i = 0; i < 8; ++i) S->h[i] ^= S->s[i % 4];
}
void blake256_init(state *S) {
S->h[0] = 0x6A09E667;
S->h[1] = 0xBB67AE85;
S->h[2] = 0x3C6EF372;
S->h[3] = 0xA54FF53A;
S->h[4] = 0x510E527F;
S->h[5] = 0x9B05688C;
S->h[6] = 0x1F83D9AB;
S->h[7] = 0x5BE0CD19;
S->t[0] = S->t[1] = S->buflen = S->nullt = 0;
S->s[0] = S->s[1] = S->s[2] = S->s[3] = 0;
}
void blake224_init(state *S) {
S->h[0] = 0xC1059ED8;
S->h[1] = 0x367CD507;
S->h[2] = 0x3070DD17;
S->h[3] = 0xF70E5939;
S->h[4] = 0xFFC00B31;
S->h[5] = 0x68581511;
S->h[6] = 0x64F98FA7;
S->h[7] = 0xBEFA4FA4;
S->t[0] = S->t[1] = S->buflen = S->nullt = 0;
S->s[0] = S->s[1] = S->s[2] = S->s[3] = 0;
}
// datalen = number of bits
void blake256_update(state *S, const uint8_t *data, uint64_t datalen) {
int left = S->buflen >> 3;
int fill = 64 - left;
if (left && (((datalen >> 3) & 0x3F) >= (unsigned) fill)) {
memcpy((void *) (S->buf + left), (void *) data, fill);
S->t[0] += 512;
if (S->t[0] == 0) S->t[1]++;
blake256_compress(S, S->buf);
data += fill;
datalen -= (fill << 3);
left = 0;
}
while (datalen >= 512) {
S->t[0] += 512;
if (S->t[0] == 0) S->t[1]++;
blake256_compress(S, data);
data += 64;
datalen -= 512;
}
if (datalen > 0) {
memcpy((void *) (S->buf + left), (void *) data, datalen >> 3);
S->buflen = (left << 3) + datalen;
} else {
S->buflen = 0;
}
}
// datalen = number of bits
void blake224_update(state *S, const uint8_t *data, uint64_t datalen) {
blake256_update(S, data, datalen);
}
void blake256_final_h(state *S, uint8_t *digest, uint8_t pa, uint8_t pb) {
uint8_t msglen[8];
uint32_t lo = S->t[0] + S->buflen, hi = S->t[1];
if (lo < (unsigned) S->buflen) hi++;
U32TO8(msglen + 0, hi);
U32TO8(msglen + 4, lo);
if (S->buflen == 440) { /* one padding byte */
S->t[0] -= 8;
blake256_update(S, &pa, 8);
} else {
if (S->buflen < 440) { /* enough space to fill the block */
if (S->buflen == 0) S->nullt = 1;
S->t[0] -= 440 - S->buflen;
blake256_update(S, padding, 440 - S->buflen);
} else { /* need 2 compressions */
S->t[0] -= 512 - S->buflen;
blake256_update(S, padding, 512 - S->buflen);
S->t[0] -= 440;
blake256_update(S, padding + 1, 440);
S->nullt = 1;
}
blake256_update(S, &pb, 8);
S->t[0] -= 8;
}
S->t[0] -= 64;
blake256_update(S, msglen, 64);
U32TO8(digest + 0, S->h[0]);
U32TO8(digest + 4, S->h[1]);
U32TO8(digest + 8, S->h[2]);
U32TO8(digest + 12, S->h[3]);
U32TO8(digest + 16, S->h[4]);
U32TO8(digest + 20, S->h[5]);
U32TO8(digest + 24, S->h[6]);
U32TO8(digest + 28, S->h[7]);
}
void blake256_final(state *S, uint8_t *digest) {
blake256_final_h(S, digest, 0x81, 0x01);
}
void blake224_final(state *S, uint8_t *digest) {
blake256_final_h(S, digest, 0x80, 0x00);
}
// inlen = number of bytes
void blake256_hash(uint8_t *out, const uint8_t *in, uint64_t inlen) {
state S;
blake256_init(&S);
blake256_update(&S, in, inlen * 8);
blake256_final(&S, out);
}
// inlen = number of bytes
void blake224_hash(uint8_t *out, const uint8_t *in, uint64_t inlen) {
state S;
blake224_init(&S);
blake224_update(&S, in, inlen * 8);
blake224_final(&S, out);
}
// keylen = number of bytes
void hmac_blake256_init(hmac_state *S, const uint8_t *_key, uint64_t keylen) {
const uint8_t *key = _key;
uint8_t keyhash[32];
uint8_t pad[64];
uint64_t i;
if (keylen > 64) {
blake256_hash(keyhash, key, keylen);
key = keyhash;
keylen = 32;
}
blake256_init(&S->inner);
memset(pad, 0x36, 64);
for (i = 0; i < keylen; ++i) {
pad[i] ^= key[i];
}
blake256_update(&S->inner, pad, 512);
blake256_init(&S->outer);
memset(pad, 0x5c, 64);
for (i = 0; i < keylen; ++i) {
pad[i] ^= key[i];
}
blake256_update(&S->outer, pad, 512);
memset(keyhash, 0, 32);
}
// keylen = number of bytes
void hmac_blake224_init(hmac_state *S, const uint8_t *_key, uint64_t keylen) {
const uint8_t *key = _key;
uint8_t keyhash[32];
uint8_t pad[64];
uint64_t i;
if (keylen > 64) {
blake256_hash(keyhash, key, keylen);
key = keyhash;
keylen = 28;
}
blake224_init(&S->inner);
memset(pad, 0x36, 64);
for (i = 0; i < keylen; ++i) {
pad[i] ^= key[i];
}
blake224_update(&S->inner, pad, 512);
blake224_init(&S->outer);
memset(pad, 0x5c, 64);
for (i = 0; i < keylen; ++i) {
pad[i] ^= key[i];
}
blake224_update(&S->outer, pad, 512);
memset(keyhash, 0, 32);
}
// datalen = number of bits
void hmac_blake256_update(hmac_state *S, const uint8_t *data, uint64_t datalen) {
// update the inner state
blake256_update(&S->inner, data, datalen);
}
// datalen = number of bits
void hmac_blake224_update(hmac_state *S, const uint8_t *data, uint64_t datalen) {
// update the inner state
blake224_update(&S->inner, data, datalen);
}
void hmac_blake256_final(hmac_state *S, uint8_t *digest) {
uint8_t ihash[32];
blake256_final(&S->inner, ihash);
blake256_update(&S->outer, ihash, 256);
blake256_final(&S->outer, digest);
memset(ihash, 0, 32);
}
void hmac_blake224_final(hmac_state *S, uint8_t *digest) {
uint8_t ihash[32];
blake224_final(&S->inner, ihash);
blake224_update(&S->outer, ihash, 224);
blake224_final(&S->outer, digest);
memset(ihash, 0, 32);
}
// keylen = number of bytes; inlen = number of bytes
void hmac_blake256_hash(uint8_t *out, const uint8_t *key, uint64_t keylen, const uint8_t *in, uint64_t inlen) {
hmac_state S;
hmac_blake256_init(&S, key, keylen);
hmac_blake256_update(&S, in, inlen * 8);
hmac_blake256_final(&S, out);
}
// keylen = number of bytes; inlen = number of bytes
void hmac_blake224_hash(uint8_t *out, const uint8_t *key, uint64_t keylen, const uint8_t *in, uint64_t inlen) {
hmac_state S;
hmac_blake224_init(&S, key, keylen);
hmac_blake224_update(&S, in, inlen * 8);
hmac_blake224_final(&S, out);
}

43
crypto/c_blake256.h Normal file
View File

@ -0,0 +1,43 @@
#ifndef _BLAKE256_H_
#define _BLAKE256_H_
#include <stdint.h>
typedef struct {
uint32_t h[8], s[4], t[2];
int buflen, nullt;
uint8_t buf[64];
} state;
typedef struct {
state inner;
state outer;
} hmac_state;
void blake256_init(state *);
void blake224_init(state *);
void blake256_update(state *, const uint8_t *, uint64_t);
void blake224_update(state *, const uint8_t *, uint64_t);
void blake256_final(state *, uint8_t *);
void blake224_final(state *, uint8_t *);
void blake256_hash(uint8_t *, const uint8_t *, uint64_t);
void blake224_hash(uint8_t *, const uint8_t *, uint64_t);
/* HMAC functions: */
void hmac_blake256_init(hmac_state *, const uint8_t *, uint64_t);
void hmac_blake224_init(hmac_state *, const uint8_t *, uint64_t);
void hmac_blake256_update(hmac_state *, const uint8_t *, uint64_t);
void hmac_blake224_update(hmac_state *, const uint8_t *, uint64_t);
void hmac_blake256_final(hmac_state *, uint8_t *);
void hmac_blake224_final(hmac_state *, uint8_t *);
void hmac_blake256_hash(uint8_t *, const uint8_t *, uint64_t, const uint8_t *, uint64_t);
void hmac_blake224_hash(uint8_t *, const uint8_t *, uint64_t, const uint8_t *, uint64_t);
#endif /* _BLAKE256_H_ */

360
crypto/c_groestl.c Normal file
View File

@ -0,0 +1,360 @@
/* hash.c April 2012
* Groestl ANSI C code optimised for 32-bit machines
* Author: Thomas Krinninger
*
* This work is based on the implementation of
* Soeren S. Thomsen and Krystian Matusiewicz
*
*
*/
#include "c_groestl.h"
#include "groestl_tables.h"
#define P_TYPE 0
#define Q_TYPE 1
const uint8_t shift_Values[2][8] = {{0,1,2,3,4,5,6,7},{1,3,5,7,0,2,4,6}};
const uint8_t indices_cyclic[15] = {0,1,2,3,4,5,6,7,0,1,2,3,4,5,6};
#define ROTATE_COLUMN_DOWN(v1, v2, amount_bytes, temp_var) {temp_var = (v1<<(8*amount_bytes))|(v2>>(8*(4-amount_bytes))); \
v2 = (v2<<(8*amount_bytes))|(v1>>(8*(4-amount_bytes))); \
v1 = temp_var;}
#define COLUMN(x,y,i,c0,c1,c2,c3,c4,c5,c6,c7,tv1,tv2,tu,tl,t) \
tu = T[2*(uint32_t)x[4*c0+0]]; \
tl = T[2*(uint32_t)x[4*c0+0]+1]; \
tv1 = T[2*(uint32_t)x[4*c1+1]]; \
tv2 = T[2*(uint32_t)x[4*c1+1]+1]; \
ROTATE_COLUMN_DOWN(tv1,tv2,1,t) \
tu ^= tv1; \
tl ^= tv2; \
tv1 = T[2*(uint32_t)x[4*c2+2]]; \
tv2 = T[2*(uint32_t)x[4*c2+2]+1]; \
ROTATE_COLUMN_DOWN(tv1,tv2,2,t) \
tu ^= tv1; \
tl ^= tv2; \
tv1 = T[2*(uint32_t)x[4*c3+3]]; \
tv2 = T[2*(uint32_t)x[4*c3+3]+1]; \
ROTATE_COLUMN_DOWN(tv1,tv2,3,t) \
tu ^= tv1; \
tl ^= tv2; \
tl ^= T[2*(uint32_t)x[4*c4+0]]; \
tu ^= T[2*(uint32_t)x[4*c4+0]+1]; \
tv1 = T[2*(uint32_t)x[4*c5+1]]; \
tv2 = T[2*(uint32_t)x[4*c5+1]+1]; \
ROTATE_COLUMN_DOWN(tv1,tv2,1,t) \
tl ^= tv1; \
tu ^= tv2; \
tv1 = T[2*(uint32_t)x[4*c6+2]]; \
tv2 = T[2*(uint32_t)x[4*c6+2]+1]; \
ROTATE_COLUMN_DOWN(tv1,tv2,2,t) \
tl ^= tv1; \
tu ^= tv2; \
tv1 = T[2*(uint32_t)x[4*c7+3]]; \
tv2 = T[2*(uint32_t)x[4*c7+3]+1]; \
ROTATE_COLUMN_DOWN(tv1,tv2,3,t) \
tl ^= tv1; \
tu ^= tv2; \
y[i] = tu; \
y[i+1] = tl;
/* compute one round of P (short variants) */
static void RND512P(uint8_t *x, uint32_t *y, uint32_t r) {
uint32_t temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp;
uint32_t* x32 = (uint32_t*)x;
x32[ 0] ^= 0x00000000^r;
x32[ 2] ^= 0x00000010^r;
x32[ 4] ^= 0x00000020^r;
x32[ 6] ^= 0x00000030^r;
x32[ 8] ^= 0x00000040^r;
x32[10] ^= 0x00000050^r;
x32[12] ^= 0x00000060^r;
x32[14] ^= 0x00000070^r;
COLUMN(x,y, 0, 0, 2, 4, 6, 9, 11, 13, 15, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 2, 2, 4, 6, 8, 11, 13, 15, 1, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 4, 4, 6, 8, 10, 13, 15, 1, 3, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 6, 6, 8, 10, 12, 15, 1, 3, 5, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 8, 8, 10, 12, 14, 1, 3, 5, 7, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y,10, 10, 12, 14, 0, 3, 5, 7, 9, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y,12, 12, 14, 0, 2, 5, 7, 9, 11, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y,14, 14, 0, 2, 4, 7, 9, 11, 13, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
}
/* compute one round of Q (short variants) */
static void RND512Q(uint8_t *x, uint32_t *y, uint32_t r) {
uint32_t temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp;
uint32_t* x32 = (uint32_t*)x;
x32[ 0] = ~x32[ 0];
x32[ 1] ^= 0xffffffff^r;
x32[ 2] = ~x32[ 2];
x32[ 3] ^= 0xefffffff^r;
x32[ 4] = ~x32[ 4];
x32[ 5] ^= 0xdfffffff^r;
x32[ 6] = ~x32[ 6];
x32[ 7] ^= 0xcfffffff^r;
x32[ 8] = ~x32[ 8];
x32[ 9] ^= 0xbfffffff^r;
x32[10] = ~x32[10];
x32[11] ^= 0xafffffff^r;
x32[12] = ~x32[12];
x32[13] ^= 0x9fffffff^r;
x32[14] = ~x32[14];
x32[15] ^= 0x8fffffff^r;
COLUMN(x,y, 0, 2, 6, 10, 14, 1, 5, 9, 13, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 2, 4, 8, 12, 0, 3, 7, 11, 15, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 4, 6, 10, 14, 2, 5, 9, 13, 1, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 6, 8, 12, 0, 4, 7, 11, 15, 3, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 8, 10, 14, 2, 6, 9, 13, 1, 5, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y,10, 12, 0, 4, 8, 11, 15, 3, 7, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y,12, 14, 2, 6, 10, 13, 1, 5, 9, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y,14, 0, 4, 8, 12, 15, 3, 7, 11, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
}
/* compute compression function (short variants) */
static void F512(uint32_t *h, const uint32_t *m) {
int i;
uint32_t Ptmp[2*COLS512];
uint32_t Qtmp[2*COLS512];
uint32_t y[2*COLS512];
uint32_t z[2*COLS512];
for (i = 0; i < 2*COLS512; i++) {
z[i] = m[i];
Ptmp[i] = h[i]^m[i];
}
/* compute Q(m) */
RND512Q((uint8_t*)z, y, 0x00000000);
RND512Q((uint8_t*)y, z, 0x01000000);
RND512Q((uint8_t*)z, y, 0x02000000);
RND512Q((uint8_t*)y, z, 0x03000000);
RND512Q((uint8_t*)z, y, 0x04000000);
RND512Q((uint8_t*)y, z, 0x05000000);
RND512Q((uint8_t*)z, y, 0x06000000);
RND512Q((uint8_t*)y, z, 0x07000000);
RND512Q((uint8_t*)z, y, 0x08000000);
RND512Q((uint8_t*)y, Qtmp, 0x09000000);
/* compute P(h+m) */
RND512P((uint8_t*)Ptmp, y, 0x00000000);
RND512P((uint8_t*)y, z, 0x00000001);
RND512P((uint8_t*)z, y, 0x00000002);
RND512P((uint8_t*)y, z, 0x00000003);
RND512P((uint8_t*)z, y, 0x00000004);
RND512P((uint8_t*)y, z, 0x00000005);
RND512P((uint8_t*)z, y, 0x00000006);
RND512P((uint8_t*)y, z, 0x00000007);
RND512P((uint8_t*)z, y, 0x00000008);
RND512P((uint8_t*)y, Ptmp, 0x00000009);
/* compute P(h+m) + Q(m) + h */
for (i = 0; i < 2*COLS512; i++) {
h[i] ^= Ptmp[i]^Qtmp[i];
}
}
/* digest up to msglen bytes of input (full blocks only) */
static void Transform(groestlHashState *ctx,
const uint8_t *input,
int msglen) {
/* digest message, one block at a time */
for (; msglen >= SIZE512;
msglen -= SIZE512, input += SIZE512) {
F512(ctx->chaining,(uint32_t*)input);
/* increment block counter */
ctx->block_counter1++;
if (ctx->block_counter1 == 0) ctx->block_counter2++;
}
}
/* given state h, do h <- P(h)+h */
static void OutputTransformation(groestlHashState *ctx) {
int j;
uint32_t temp[2*COLS512];
uint32_t y[2*COLS512];
uint32_t z[2*COLS512];
for (j = 0; j < 2*COLS512; j++) {
temp[j] = ctx->chaining[j];
}
RND512P((uint8_t*)temp, y, 0x00000000);
RND512P((uint8_t*)y, z, 0x00000001);
RND512P((uint8_t*)z, y, 0x00000002);
RND512P((uint8_t*)y, z, 0x00000003);
RND512P((uint8_t*)z, y, 0x00000004);
RND512P((uint8_t*)y, z, 0x00000005);
RND512P((uint8_t*)z, y, 0x00000006);
RND512P((uint8_t*)y, z, 0x00000007);
RND512P((uint8_t*)z, y, 0x00000008);
RND512P((uint8_t*)y, temp, 0x00000009);
for (j = 0; j < 2*COLS512; j++) {
ctx->chaining[j] ^= temp[j];
}
}
/* initialise context */
static void Init(groestlHashState* ctx) {
int i = 0;
/* allocate memory for state and data buffer */
for(;i<(SIZE512/sizeof(uint32_t));i++)
{
ctx->chaining[i] = 0;
}
/* set initial value */
ctx->chaining[2*COLS512-1] = u32BIG((uint32_t)HASH_BIT_LEN);
/* set other variables */
ctx->buf_ptr = 0;
ctx->block_counter1 = 0;
ctx->block_counter2 = 0;
ctx->bits_in_last_byte = 0;
}
/* update state with databitlen bits of input */
static void Update(groestlHashState* ctx,
const BitSequence* input,
DataLength databitlen) {
int index = 0;
int msglen = (int)(databitlen/8);
int rem = (int)(databitlen%8);
/* if the buffer contains data that has not yet been digested, first
add data to buffer until full */
if (ctx->buf_ptr) {
while (ctx->buf_ptr < SIZE512 && index < msglen) {
ctx->buffer[(int)ctx->buf_ptr++] = input[index++];
}
if (ctx->buf_ptr < SIZE512) {
/* buffer still not full, return */
if (rem) {
ctx->bits_in_last_byte = rem;
ctx->buffer[(int)ctx->buf_ptr++] = input[index];
}
return;
}
/* digest buffer */
ctx->buf_ptr = 0;
Transform(ctx, ctx->buffer, SIZE512);
}
/* digest bulk of message */
Transform(ctx, input+index, msglen-index);
index += ((msglen-index)/SIZE512)*SIZE512;
/* store remaining data in buffer */
while (index < msglen) {
ctx->buffer[(int)ctx->buf_ptr++] = input[index++];
}
/* if non-integral number of bytes have been supplied, store
remaining bits in last byte, together with information about
number of bits */
if (rem) {
ctx->bits_in_last_byte = rem;
ctx->buffer[(int)ctx->buf_ptr++] = input[index];
}
}
#define BILB ctx->bits_in_last_byte
/* finalise: process remaining data (including padding), perform
output transformation, and write hash result to 'output' */
static void Final(groestlHashState* ctx,
BitSequence* output) {
int i, j = 0, hashbytelen = HASH_BIT_LEN/8;
uint8_t *s = (BitSequence*)ctx->chaining;
/* pad with '1'-bit and first few '0'-bits */
if (BILB) {
ctx->buffer[(int)ctx->buf_ptr-1] &= ((1<<BILB)-1)<<(8-BILB);
ctx->buffer[(int)ctx->buf_ptr-1] ^= 0x1<<(7-BILB);
BILB = 0;
}
else ctx->buffer[(int)ctx->buf_ptr++] = 0x80;
/* pad with '0'-bits */
if (ctx->buf_ptr > SIZE512-LENGTHFIELDLEN) {
/* padding requires two blocks */
while (ctx->buf_ptr < SIZE512) {
ctx->buffer[(int)ctx->buf_ptr++] = 0;
}
/* digest first padding block */
Transform(ctx, ctx->buffer, SIZE512);
ctx->buf_ptr = 0;
}
while (ctx->buf_ptr < SIZE512-LENGTHFIELDLEN) {
ctx->buffer[(int)ctx->buf_ptr++] = 0;
}
/* length padding */
ctx->block_counter1++;
if (ctx->block_counter1 == 0) ctx->block_counter2++;
ctx->buf_ptr = SIZE512;
while (ctx->buf_ptr > SIZE512-(int)sizeof(uint32_t)) {
ctx->buffer[(int)--ctx->buf_ptr] = (uint8_t)ctx->block_counter1;
ctx->block_counter1 >>= 8;
}
while (ctx->buf_ptr > SIZE512-LENGTHFIELDLEN) {
ctx->buffer[(int)--ctx->buf_ptr] = (uint8_t)ctx->block_counter2;
ctx->block_counter2 >>= 8;
}
/* digest final padding block */
Transform(ctx, ctx->buffer, SIZE512);
/* perform output transformation */
OutputTransformation(ctx);
/* store hash result in output */
for (i = SIZE512-hashbytelen; i < SIZE512; i++,j++) {
output[j] = s[i];
}
/* zeroise relevant variables and deallocate memory */
for (i = 0; i < COLS512; i++) {
ctx->chaining[i] = 0;
}
for (i = 0; i < SIZE512; i++) {
ctx->buffer[i] = 0;
}
}
/* hash bit sequence */
void groestl(const BitSequence* data,
DataLength databitlen,
BitSequence* hashval) {
groestlHashState context;
/* initialise */
Init(&context);
/* process message */
Update(&context, data, databitlen);
/* finalise */
Final(&context, hashval);
}
/*
static int crypto_hash(unsigned char *out,
const unsigned char *in,
unsigned long long len)
{
groestl(in, 8*len, out);
return 0;
}
*/

60
crypto/c_groestl.h Normal file
View File

@ -0,0 +1,60 @@
#ifndef __hash_h
#define __hash_h
/*
#include "crypto_uint8.h"
#include "crypto_uint32.h"
#include "crypto_uint64.h"
#include "crypto_hash.h"
typedef crypto_uint8 uint8_t;
typedef crypto_uint32 uint32_t;
typedef crypto_uint64 uint64_t;
*/
#include <stdint.h>
#include "hash.h"
/* some sizes (number of bytes) */
#define ROWS 8
#define LENGTHFIELDLEN ROWS
#define COLS512 8
#define SIZE512 (ROWS*COLS512)
#define ROUNDS512 10
#define HASH_BIT_LEN 256
#define ROTL32(v, n) ((((v)<<(n))|((v)>>(32-(n))))&li_32(ffffffff))
#define li_32(h) 0x##h##u
#define EXT_BYTE(var,n) ((uint8_t)((uint32_t)(var) >> (8*n)))
#define u32BIG(a) \
((ROTL32(a,8) & li_32(00FF00FF)) | \
(ROTL32(a,24) & li_32(FF00FF00)))
/* NIST API begin */
typedef struct {
uint32_t chaining[SIZE512/sizeof(uint32_t)]; /* actual state */
uint32_t block_counter1,
block_counter2; /* message block counter(s) */
BitSequence buffer[SIZE512]; /* data buffer */
int buf_ptr; /* data buffer pointer */
int bits_in_last_byte; /* no. of message bits in last byte of
data buffer */
} groestlHashState;
/*void Init(hashState*);
void Update(hashState*, const BitSequence*, DataLength);
void Final(hashState*, BitSequence*); */
void groestl(const BitSequence*, DataLength, BitSequence*);
/* NIST API end */
/*
int crypto_hash(unsigned char *out,
const unsigned char *in,
unsigned long long len);
*/
#endif /* __hash_h */

367
crypto/c_jh.c Normal file
View File

@ -0,0 +1,367 @@
/*This program gives the 64-bit optimized bitslice implementation of JH using ANSI C
--------------------------------
Performance
Microprocessor: Intel CORE 2 processor (Core 2 Duo Mobile T6600 2.2GHz)
Operating System: 64-bit Ubuntu 10.04 (Linux kernel 2.6.32-22-generic)
Speed for long message:
1) 45.8 cycles/byte compiler: Intel C++ Compiler 11.1 compilation option: icc -O2
2) 56.8 cycles/byte compiler: gcc 4.4.3 compilation option: gcc -O3
--------------------------------
Last Modified: January 16, 2011
*/
#include "c_jh.h"
#include <stdint.h>
#include <string.h>
/*typedef unsigned long long uint64;*/
typedef uint64_t uint64;
/*define data alignment for different C compilers*/
#if defined(__GNUC__)
#define DATA_ALIGN16(x) x __attribute__ ((aligned(16)))
#else
#define DATA_ALIGN16(x) __declspec(align(16)) x
#endif
typedef struct {
int hashbitlen; /*the message digest size*/
unsigned long long databitlen; /*the message size in bits*/
unsigned long long datasize_in_buffer; /*the size of the message remained in buffer; assumed to be multiple of 8bits except for the last partial block at the end of the message*/
DATA_ALIGN16(uint64 x[8][2]); /*the 1024-bit state, ( x[i][0] || x[i][1] ) is the ith row of the state in the pseudocode*/
unsigned char buffer[64]; /*the 512-bit message block to be hashed;*/
} hashState;
/*The initial hash value H(0)*/
const unsigned char JH224_H0[128]={0x2d,0xfe,0xdd,0x62,0xf9,0x9a,0x98,0xac,0xae,0x7c,0xac,0xd6,0x19,0xd6,0x34,0xe7,0xa4,0x83,0x10,0x5,0xbc,0x30,0x12,0x16,0xb8,0x60,0x38,0xc6,0xc9,0x66,0x14,0x94,0x66,0xd9,0x89,0x9f,0x25,0x80,0x70,0x6f,0xce,0x9e,0xa3,0x1b,0x1d,0x9b,0x1a,0xdc,0x11,0xe8,0x32,0x5f,0x7b,0x36,0x6e,0x10,0xf9,0x94,0x85,0x7f,0x2,0xfa,0x6,0xc1,0x1b,0x4f,0x1b,0x5c,0xd8,0xc8,0x40,0xb3,0x97,0xf6,0xa1,0x7f,0x6e,0x73,0x80,0x99,0xdc,0xdf,0x93,0xa5,0xad,0xea,0xa3,0xd3,0xa4,0x31,0xe8,0xde,0xc9,0x53,0x9a,0x68,0x22,0xb4,0xa9,0x8a,0xec,0x86,0xa1,0xe4,0xd5,0x74,0xac,0x95,0x9c,0xe5,0x6c,0xf0,0x15,0x96,0xd,0xea,0xb5,0xab,0x2b,0xbf,0x96,0x11,0xdc,0xf0,0xdd,0x64,0xea,0x6e};
const unsigned char JH256_H0[128]={0xeb,0x98,0xa3,0x41,0x2c,0x20,0xd3,0xeb,0x92,0xcd,0xbe,0x7b,0x9c,0xb2,0x45,0xc1,0x1c,0x93,0x51,0x91,0x60,0xd4,0xc7,0xfa,0x26,0x0,0x82,0xd6,0x7e,0x50,0x8a,0x3,0xa4,0x23,0x9e,0x26,0x77,0x26,0xb9,0x45,0xe0,0xfb,0x1a,0x48,0xd4,0x1a,0x94,0x77,0xcd,0xb5,0xab,0x26,0x2,0x6b,0x17,0x7a,0x56,0xf0,0x24,0x42,0xf,0xff,0x2f,0xa8,0x71,0xa3,0x96,0x89,0x7f,0x2e,0x4d,0x75,0x1d,0x14,0x49,0x8,0xf7,0x7d,0xe2,0x62,0x27,0x76,0x95,0xf7,0x76,0x24,0x8f,0x94,0x87,0xd5,0xb6,0x57,0x47,0x80,0x29,0x6c,0x5c,0x5e,0x27,0x2d,0xac,0x8e,0xd,0x6c,0x51,0x84,0x50,0xc6,0x57,0x5,0x7a,0xf,0x7b,0xe4,0xd3,0x67,0x70,0x24,0x12,0xea,0x89,0xe3,0xab,0x13,0xd3,0x1c,0xd7,0x69};
const unsigned char JH384_H0[128]={0x48,0x1e,0x3b,0xc6,0xd8,0x13,0x39,0x8a,0x6d,0x3b,0x5e,0x89,0x4a,0xde,0x87,0x9b,0x63,0xfa,0xea,0x68,0xd4,0x80,0xad,0x2e,0x33,0x2c,0xcb,0x21,0x48,0xf,0x82,0x67,0x98,0xae,0xc8,0x4d,0x90,0x82,0xb9,0x28,0xd4,0x55,0xea,0x30,0x41,0x11,0x42,0x49,0x36,0xf5,0x55,0xb2,0x92,0x48,0x47,0xec,0xc7,0x25,0xa,0x93,0xba,0xf4,0x3c,0xe1,0x56,0x9b,0x7f,0x8a,0x27,0xdb,0x45,0x4c,0x9e,0xfc,0xbd,0x49,0x63,0x97,0xaf,0xe,0x58,0x9f,0xc2,0x7d,0x26,0xaa,0x80,0xcd,0x80,0xc0,0x8b,0x8c,0x9d,0xeb,0x2e,0xda,0x8a,0x79,0x81,0xe8,0xf8,0xd5,0x37,0x3a,0xf4,0x39,0x67,0xad,0xdd,0xd1,0x7a,0x71,0xa9,0xb4,0xd3,0xbd,0xa4,0x75,0xd3,0x94,0x97,0x6c,0x3f,0xba,0x98,0x42,0x73,0x7f};
const unsigned char JH512_H0[128]={0x6f,0xd1,0x4b,0x96,0x3e,0x0,0xaa,0x17,0x63,0x6a,0x2e,0x5,0x7a,0x15,0xd5,0x43,0x8a,0x22,0x5e,0x8d,0xc,0x97,0xef,0xb,0xe9,0x34,0x12,0x59,0xf2,0xb3,0xc3,0x61,0x89,0x1d,0xa0,0xc1,0x53,0x6f,0x80,0x1e,0x2a,0xa9,0x5,0x6b,0xea,0x2b,0x6d,0x80,0x58,0x8e,0xcc,0xdb,0x20,0x75,0xba,0xa6,0xa9,0xf,0x3a,0x76,0xba,0xf8,0x3b,0xf7,0x1,0x69,0xe6,0x5,0x41,0xe3,0x4a,0x69,0x46,0xb5,0x8a,0x8e,0x2e,0x6f,0xe6,0x5a,0x10,0x47,0xa7,0xd0,0xc1,0x84,0x3c,0x24,0x3b,0x6e,0x71,0xb1,0x2d,0x5a,0xc1,0x99,0xcf,0x57,0xf6,0xec,0x9d,0xb1,0xf8,0x56,0xa7,0x6,0x88,0x7c,0x57,0x16,0xb1,0x56,0xe3,0xc2,0xfc,0xdf,0xe6,0x85,0x17,0xfb,0x54,0x5a,0x46,0x78,0xcc,0x8c,0xdd,0x4b};
/*42 round constants, each round constant is 32-byte (256-bit)*/
const unsigned char E8_bitslice_roundconstant[42][32]={
{0x72,0xd5,0xde,0xa2,0xdf,0x15,0xf8,0x67,0x7b,0x84,0x15,0xa,0xb7,0x23,0x15,0x57,0x81,0xab,0xd6,0x90,0x4d,0x5a,0x87,0xf6,0x4e,0x9f,0x4f,0xc5,0xc3,0xd1,0x2b,0x40},
{0xea,0x98,0x3a,0xe0,0x5c,0x45,0xfa,0x9c,0x3,0xc5,0xd2,0x99,0x66,0xb2,0x99,0x9a,0x66,0x2,0x96,0xb4,0xf2,0xbb,0x53,0x8a,0xb5,0x56,0x14,0x1a,0x88,0xdb,0xa2,0x31},
{0x3,0xa3,0x5a,0x5c,0x9a,0x19,0xe,0xdb,0x40,0x3f,0xb2,0xa,0x87,0xc1,0x44,0x10,0x1c,0x5,0x19,0x80,0x84,0x9e,0x95,0x1d,0x6f,0x33,0xeb,0xad,0x5e,0xe7,0xcd,0xdc},
{0x10,0xba,0x13,0x92,0x2,0xbf,0x6b,0x41,0xdc,0x78,0x65,0x15,0xf7,0xbb,0x27,0xd0,0xa,0x2c,0x81,0x39,0x37,0xaa,0x78,0x50,0x3f,0x1a,0xbf,0xd2,0x41,0x0,0x91,0xd3},
{0x42,0x2d,0x5a,0xd,0xf6,0xcc,0x7e,0x90,0xdd,0x62,0x9f,0x9c,0x92,0xc0,0x97,0xce,0x18,0x5c,0xa7,0xb,0xc7,0x2b,0x44,0xac,0xd1,0xdf,0x65,0xd6,0x63,0xc6,0xfc,0x23},
{0x97,0x6e,0x6c,0x3,0x9e,0xe0,0xb8,0x1a,0x21,0x5,0x45,0x7e,0x44,0x6c,0xec,0xa8,0xee,0xf1,0x3,0xbb,0x5d,0x8e,0x61,0xfa,0xfd,0x96,0x97,0xb2,0x94,0x83,0x81,0x97},
{0x4a,0x8e,0x85,0x37,0xdb,0x3,0x30,0x2f,0x2a,0x67,0x8d,0x2d,0xfb,0x9f,0x6a,0x95,0x8a,0xfe,0x73,0x81,0xf8,0xb8,0x69,0x6c,0x8a,0xc7,0x72,0x46,0xc0,0x7f,0x42,0x14},
{0xc5,0xf4,0x15,0x8f,0xbd,0xc7,0x5e,0xc4,0x75,0x44,0x6f,0xa7,0x8f,0x11,0xbb,0x80,0x52,0xde,0x75,0xb7,0xae,0xe4,0x88,0xbc,0x82,0xb8,0x0,0x1e,0x98,0xa6,0xa3,0xf4},
{0x8e,0xf4,0x8f,0x33,0xa9,0xa3,0x63,0x15,0xaa,0x5f,0x56,0x24,0xd5,0xb7,0xf9,0x89,0xb6,0xf1,0xed,0x20,0x7c,0x5a,0xe0,0xfd,0x36,0xca,0xe9,0x5a,0x6,0x42,0x2c,0x36},
{0xce,0x29,0x35,0x43,0x4e,0xfe,0x98,0x3d,0x53,0x3a,0xf9,0x74,0x73,0x9a,0x4b,0xa7,0xd0,0xf5,0x1f,0x59,0x6f,0x4e,0x81,0x86,0xe,0x9d,0xad,0x81,0xaf,0xd8,0x5a,0x9f},
{0xa7,0x5,0x6,0x67,0xee,0x34,0x62,0x6a,0x8b,0xb,0x28,0xbe,0x6e,0xb9,0x17,0x27,0x47,0x74,0x7,0x26,0xc6,0x80,0x10,0x3f,0xe0,0xa0,0x7e,0x6f,0xc6,0x7e,0x48,0x7b},
{0xd,0x55,0xa,0xa5,0x4a,0xf8,0xa4,0xc0,0x91,0xe3,0xe7,0x9f,0x97,0x8e,0xf1,0x9e,0x86,0x76,0x72,0x81,0x50,0x60,0x8d,0xd4,0x7e,0x9e,0x5a,0x41,0xf3,0xe5,0xb0,0x62},
{0xfc,0x9f,0x1f,0xec,0x40,0x54,0x20,0x7a,0xe3,0xe4,0x1a,0x0,0xce,0xf4,0xc9,0x84,0x4f,0xd7,0x94,0xf5,0x9d,0xfa,0x95,0xd8,0x55,0x2e,0x7e,0x11,0x24,0xc3,0x54,0xa5},
{0x5b,0xdf,0x72,0x28,0xbd,0xfe,0x6e,0x28,0x78,0xf5,0x7f,0xe2,0xf,0xa5,0xc4,0xb2,0x5,0x89,0x7c,0xef,0xee,0x49,0xd3,0x2e,0x44,0x7e,0x93,0x85,0xeb,0x28,0x59,0x7f},
{0x70,0x5f,0x69,0x37,0xb3,0x24,0x31,0x4a,0x5e,0x86,0x28,0xf1,0x1d,0xd6,0xe4,0x65,0xc7,0x1b,0x77,0x4,0x51,0xb9,0x20,0xe7,0x74,0xfe,0x43,0xe8,0x23,0xd4,0x87,0x8a},
{0x7d,0x29,0xe8,0xa3,0x92,0x76,0x94,0xf2,0xdd,0xcb,0x7a,0x9,0x9b,0x30,0xd9,0xc1,0x1d,0x1b,0x30,0xfb,0x5b,0xdc,0x1b,0xe0,0xda,0x24,0x49,0x4f,0xf2,0x9c,0x82,0xbf},
{0xa4,0xe7,0xba,0x31,0xb4,0x70,0xbf,0xff,0xd,0x32,0x44,0x5,0xde,0xf8,0xbc,0x48,0x3b,0xae,0xfc,0x32,0x53,0xbb,0xd3,0x39,0x45,0x9f,0xc3,0xc1,0xe0,0x29,0x8b,0xa0},
{0xe5,0xc9,0x5,0xfd,0xf7,0xae,0x9,0xf,0x94,0x70,0x34,0x12,0x42,0x90,0xf1,0x34,0xa2,0x71,0xb7,0x1,0xe3,0x44,0xed,0x95,0xe9,0x3b,0x8e,0x36,0x4f,0x2f,0x98,0x4a},
{0x88,0x40,0x1d,0x63,0xa0,0x6c,0xf6,0x15,0x47,0xc1,0x44,0x4b,0x87,0x52,0xaf,0xff,0x7e,0xbb,0x4a,0xf1,0xe2,0xa,0xc6,0x30,0x46,0x70,0xb6,0xc5,0xcc,0x6e,0x8c,0xe6},
{0xa4,0xd5,0xa4,0x56,0xbd,0x4f,0xca,0x0,0xda,0x9d,0x84,0x4b,0xc8,0x3e,0x18,0xae,0x73,0x57,0xce,0x45,0x30,0x64,0xd1,0xad,0xe8,0xa6,0xce,0x68,0x14,0x5c,0x25,0x67},
{0xa3,0xda,0x8c,0xf2,0xcb,0xe,0xe1,0x16,0x33,0xe9,0x6,0x58,0x9a,0x94,0x99,0x9a,0x1f,0x60,0xb2,0x20,0xc2,0x6f,0x84,0x7b,0xd1,0xce,0xac,0x7f,0xa0,0xd1,0x85,0x18},
{0x32,0x59,0x5b,0xa1,0x8d,0xdd,0x19,0xd3,0x50,0x9a,0x1c,0xc0,0xaa,0xa5,0xb4,0x46,0x9f,0x3d,0x63,0x67,0xe4,0x4,0x6b,0xba,0xf6,0xca,0x19,0xab,0xb,0x56,0xee,0x7e},
{0x1f,0xb1,0x79,0xea,0xa9,0x28,0x21,0x74,0xe9,0xbd,0xf7,0x35,0x3b,0x36,0x51,0xee,0x1d,0x57,0xac,0x5a,0x75,0x50,0xd3,0x76,0x3a,0x46,0xc2,0xfe,0xa3,0x7d,0x70,0x1},
{0xf7,0x35,0xc1,0xaf,0x98,0xa4,0xd8,0x42,0x78,0xed,0xec,0x20,0x9e,0x6b,0x67,0x79,0x41,0x83,0x63,0x15,0xea,0x3a,0xdb,0xa8,0xfa,0xc3,0x3b,0x4d,0x32,0x83,0x2c,0x83},
{0xa7,0x40,0x3b,0x1f,0x1c,0x27,0x47,0xf3,0x59,0x40,0xf0,0x34,0xb7,0x2d,0x76,0x9a,0xe7,0x3e,0x4e,0x6c,0xd2,0x21,0x4f,0xfd,0xb8,0xfd,0x8d,0x39,0xdc,0x57,0x59,0xef},
{0x8d,0x9b,0xc,0x49,0x2b,0x49,0xeb,0xda,0x5b,0xa2,0xd7,0x49,0x68,0xf3,0x70,0xd,0x7d,0x3b,0xae,0xd0,0x7a,0x8d,0x55,0x84,0xf5,0xa5,0xe9,0xf0,0xe4,0xf8,0x8e,0x65},
{0xa0,0xb8,0xa2,0xf4,0x36,0x10,0x3b,0x53,0xc,0xa8,0x7,0x9e,0x75,0x3e,0xec,0x5a,0x91,0x68,0x94,0x92,0x56,0xe8,0x88,0x4f,0x5b,0xb0,0x5c,0x55,0xf8,0xba,0xbc,0x4c},
{0xe3,0xbb,0x3b,0x99,0xf3,0x87,0x94,0x7b,0x75,0xda,0xf4,0xd6,0x72,0x6b,0x1c,0x5d,0x64,0xae,0xac,0x28,0xdc,0x34,0xb3,0x6d,0x6c,0x34,0xa5,0x50,0xb8,0x28,0xdb,0x71},
{0xf8,0x61,0xe2,0xf2,0x10,0x8d,0x51,0x2a,0xe3,0xdb,0x64,0x33,0x59,0xdd,0x75,0xfc,0x1c,0xac,0xbc,0xf1,0x43,0xce,0x3f,0xa2,0x67,0xbb,0xd1,0x3c,0x2,0xe8,0x43,0xb0},
{0x33,0xa,0x5b,0xca,0x88,0x29,0xa1,0x75,0x7f,0x34,0x19,0x4d,0xb4,0x16,0x53,0x5c,0x92,0x3b,0x94,0xc3,0xe,0x79,0x4d,0x1e,0x79,0x74,0x75,0xd7,0xb6,0xee,0xaf,0x3f},
{0xea,0xa8,0xd4,0xf7,0xbe,0x1a,0x39,0x21,0x5c,0xf4,0x7e,0x9,0x4c,0x23,0x27,0x51,0x26,0xa3,0x24,0x53,0xba,0x32,0x3c,0xd2,0x44,0xa3,0x17,0x4a,0x6d,0xa6,0xd5,0xad},
{0xb5,0x1d,0x3e,0xa6,0xaf,0xf2,0xc9,0x8,0x83,0x59,0x3d,0x98,0x91,0x6b,0x3c,0x56,0x4c,0xf8,0x7c,0xa1,0x72,0x86,0x60,0x4d,0x46,0xe2,0x3e,0xcc,0x8,0x6e,0xc7,0xf6},
{0x2f,0x98,0x33,0xb3,0xb1,0xbc,0x76,0x5e,0x2b,0xd6,0x66,0xa5,0xef,0xc4,0xe6,0x2a,0x6,0xf4,0xb6,0xe8,0xbe,0xc1,0xd4,0x36,0x74,0xee,0x82,0x15,0xbc,0xef,0x21,0x63},
{0xfd,0xc1,0x4e,0xd,0xf4,0x53,0xc9,0x69,0xa7,0x7d,0x5a,0xc4,0x6,0x58,0x58,0x26,0x7e,0xc1,0x14,0x16,0x6,0xe0,0xfa,0x16,0x7e,0x90,0xaf,0x3d,0x28,0x63,0x9d,0x3f},
{0xd2,0xc9,0xf2,0xe3,0x0,0x9b,0xd2,0xc,0x5f,0xaa,0xce,0x30,0xb7,0xd4,0xc,0x30,0x74,0x2a,0x51,0x16,0xf2,0xe0,0x32,0x98,0xd,0xeb,0x30,0xd8,0xe3,0xce,0xf8,0x9a},
{0x4b,0xc5,0x9e,0x7b,0xb5,0xf1,0x79,0x92,0xff,0x51,0xe6,0x6e,0x4,0x86,0x68,0xd3,0x9b,0x23,0x4d,0x57,0xe6,0x96,0x67,0x31,0xcc,0xe6,0xa6,0xf3,0x17,0xa,0x75,0x5},
{0xb1,0x76,0x81,0xd9,0x13,0x32,0x6c,0xce,0x3c,0x17,0x52,0x84,0xf8,0x5,0xa2,0x62,0xf4,0x2b,0xcb,0xb3,0x78,0x47,0x15,0x47,0xff,0x46,0x54,0x82,0x23,0x93,0x6a,0x48},
{0x38,0xdf,0x58,0x7,0x4e,0x5e,0x65,0x65,0xf2,0xfc,0x7c,0x89,0xfc,0x86,0x50,0x8e,0x31,0x70,0x2e,0x44,0xd0,0xb,0xca,0x86,0xf0,0x40,0x9,0xa2,0x30,0x78,0x47,0x4e},
{0x65,0xa0,0xee,0x39,0xd1,0xf7,0x38,0x83,0xf7,0x5e,0xe9,0x37,0xe4,0x2c,0x3a,0xbd,0x21,0x97,0xb2,0x26,0x1,0x13,0xf8,0x6f,0xa3,0x44,0xed,0xd1,0xef,0x9f,0xde,0xe7},
{0x8b,0xa0,0xdf,0x15,0x76,0x25,0x92,0xd9,0x3c,0x85,0xf7,0xf6,0x12,0xdc,0x42,0xbe,0xd8,0xa7,0xec,0x7c,0xab,0x27,0xb0,0x7e,0x53,0x8d,0x7d,0xda,0xaa,0x3e,0xa8,0xde},
{0xaa,0x25,0xce,0x93,0xbd,0x2,0x69,0xd8,0x5a,0xf6,0x43,0xfd,0x1a,0x73,0x8,0xf9,0xc0,0x5f,0xef,0xda,0x17,0x4a,0x19,0xa5,0x97,0x4d,0x66,0x33,0x4c,0xfd,0x21,0x6a},
{0x35,0xb4,0x98,0x31,0xdb,0x41,0x15,0x70,0xea,0x1e,0xf,0xbb,0xed,0xcd,0x54,0x9b,0x9a,0xd0,0x63,0xa1,0x51,0x97,0x40,0x72,0xf6,0x75,0x9d,0xbf,0x91,0x47,0x6f,0xe2}};
static void E8(hashState *state); /*The bijective function E8, in bitslice form*/
static void F8(hashState *state); /*The compression function F8 */
/*The API functions*/
static HashReturn Init(hashState *state, int hashbitlen);
static HashReturn Update(hashState *state, const BitSequence *data, DataLength databitlen);
static HashReturn Final(hashState *state, BitSequence *hashval);
HashReturn jh_hash(int hashbitlen, const BitSequence *data,DataLength databitlen, BitSequence *hashval);
/*swapping bit 2i with bit 2i+1 of 64-bit x*/
#define SWAP1(x) (x) = ((((x) & 0x5555555555555555ULL) << 1) | (((x) & 0xaaaaaaaaaaaaaaaaULL) >> 1));
/*swapping bits 4i||4i+1 with bits 4i+2||4i+3 of 64-bit x*/
#define SWAP2(x) (x) = ((((x) & 0x3333333333333333ULL) << 2) | (((x) & 0xccccccccccccccccULL) >> 2));
/*swapping bits 8i||8i+1||8i+2||8i+3 with bits 8i+4||8i+5||8i+6||8i+7 of 64-bit x*/
#define SWAP4(x) (x) = ((((x) & 0x0f0f0f0f0f0f0f0fULL) << 4) | (((x) & 0xf0f0f0f0f0f0f0f0ULL) >> 4));
/*swapping bits 16i||16i+1||......||16i+7 with bits 16i+8||16i+9||......||16i+15 of 64-bit x*/
#define SWAP8(x) (x) = ((((x) & 0x00ff00ff00ff00ffULL) << 8) | (((x) & 0xff00ff00ff00ff00ULL) >> 8));
/*swapping bits 32i||32i+1||......||32i+15 with bits 32i+16||32i+17||......||32i+31 of 64-bit x*/
#define SWAP16(x) (x) = ((((x) & 0x0000ffff0000ffffULL) << 16) | (((x) & 0xffff0000ffff0000ULL) >> 16));
/*swapping bits 64i||64i+1||......||64i+31 with bits 64i+32||64i+33||......||64i+63 of 64-bit x*/
#define SWAP32(x) (x) = (((x) << 32) | ((x) >> 32));
/*The MDS transform*/
#define L(m0,m1,m2,m3,m4,m5,m6,m7) \
(m4) ^= (m1); \
(m5) ^= (m2); \
(m6) ^= (m0) ^ (m3); \
(m7) ^= (m0); \
(m0) ^= (m5); \
(m1) ^= (m6); \
(m2) ^= (m4) ^ (m7); \
(m3) ^= (m4);
/*Two Sboxes are computed in parallel, each Sbox implements S0 and S1, selected by a constant bit*/
/*The reason to compute two Sboxes in parallel is to try to fully utilize the parallel processing power*/
#define SS(m0,m1,m2,m3,m4,m5,m6,m7,cc0,cc1) \
m3 = ~(m3); \
m7 = ~(m7); \
m0 ^= ((~(m2)) & (cc0)); \
m4 ^= ((~(m6)) & (cc1)); \
temp0 = (cc0) ^ ((m0) & (m1));\
temp1 = (cc1) ^ ((m4) & (m5));\
m0 ^= ((m2) & (m3)); \
m4 ^= ((m6) & (m7)); \
m3 ^= ((~(m1)) & (m2)); \
m7 ^= ((~(m5)) & (m6)); \
m1 ^= ((m0) & (m2)); \
m5 ^= ((m4) & (m6)); \
m2 ^= ((m0) & (~(m3))); \
m6 ^= ((m4) & (~(m7))); \
m0 ^= ((m1) | (m3)); \
m4 ^= ((m5) | (m7)); \
m3 ^= ((m1) & (m2)); \
m7 ^= ((m5) & (m6)); \
m1 ^= (temp0 & (m0)); \
m5 ^= (temp1 & (m4)); \
m2 ^= temp0; \
m6 ^= temp1;
/*The bijective function E8, in bitslice form*/
static void E8(hashState *state)
{
uint64 i,roundnumber,temp0,temp1;
for (roundnumber = 0; roundnumber < 42; roundnumber = roundnumber+7) {
/*round 7*roundnumber+0: Sbox, MDS and Swapping layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+0])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+0])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
SWAP1(state->x[1][i]); SWAP1(state->x[3][i]); SWAP1(state->x[5][i]); SWAP1(state->x[7][i]);
}
/*round 7*roundnumber+1: Sbox, MDS and Swapping layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+1])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+1])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
SWAP2(state->x[1][i]); SWAP2(state->x[3][i]); SWAP2(state->x[5][i]); SWAP2(state->x[7][i]);
}
/*round 7*roundnumber+2: Sbox, MDS and Swapping layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+2])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+2])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
SWAP4(state->x[1][i]); SWAP4(state->x[3][i]); SWAP4(state->x[5][i]); SWAP4(state->x[7][i]);
}
/*round 7*roundnumber+3: Sbox, MDS and Swapping layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+3])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+3])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
SWAP8(state->x[1][i]); SWAP8(state->x[3][i]); SWAP8(state->x[5][i]); SWAP8(state->x[7][i]);
}
/*round 7*roundnumber+4: Sbox, MDS and Swapping layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+4])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+4])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
SWAP16(state->x[1][i]); SWAP16(state->x[3][i]); SWAP16(state->x[5][i]); SWAP16(state->x[7][i]);
}
/*round 7*roundnumber+5: Sbox, MDS and Swapping layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+5])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+5])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
SWAP32(state->x[1][i]); SWAP32(state->x[3][i]); SWAP32(state->x[5][i]); SWAP32(state->x[7][i]);
}
/*round 7*roundnumber+6: Sbox and MDS layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+6])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+6])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
}
/*round 7*roundnumber+6: swapping layer*/
for (i = 1; i < 8; i = i+2) {
temp0 = state->x[i][0]; state->x[i][0] = state->x[i][1]; state->x[i][1] = temp0;
}
}
}
/*The compression function F8 */
static void F8(hashState *state)
{
uint64 i;
/*xor the 512-bit message with the fist half of the 1024-bit hash state*/
for (i = 0; i < 8; i++) state->x[i >> 1][i & 1] ^= ((uint64*)state->buffer)[i];
/*the bijective function E8 */
E8(state);
/*xor the 512-bit message with the second half of the 1024-bit hash state*/
for (i = 0; i < 8; i++) state->x[(8+i) >> 1][(8+i) & 1] ^= ((uint64*)state->buffer)[i];
}
/*before hashing a message, initialize the hash state as H0 */
static HashReturn Init(hashState *state, int hashbitlen)
{
state->databitlen = 0;
state->datasize_in_buffer = 0;
/*initialize the initial hash value of JH*/
state->hashbitlen = hashbitlen;
/*load the intital hash value into state*/
switch (hashbitlen)
{
case 224: memcpy(state->x,JH224_H0,128); break;
case 256: memcpy(state->x,JH256_H0,128); break;
case 384: memcpy(state->x,JH384_H0,128); break;
case 512: memcpy(state->x,JH512_H0,128); break;
}
return(SUCCESS);
}
/*hash each 512-bit message block, except the last partial block*/
static HashReturn Update(hashState *state, const BitSequence *data, DataLength databitlen)
{
DataLength index; /*the starting address of the data to be compressed*/
state->databitlen += databitlen;
index = 0;
/*if there is remaining data in the buffer, fill it to a full message block first*/
/*we assume that the size of the data in the buffer is the multiple of 8 bits if it is not at the end of a message*/
/*There is data in the buffer, but the incoming data is insufficient for a full block*/
if ( (state->datasize_in_buffer > 0 ) && (( state->datasize_in_buffer + databitlen) < 512) ) {
if ( (databitlen & 7) == 0 ) {
memcpy(state->buffer + (state->datasize_in_buffer >> 3), data, 64-(state->datasize_in_buffer >> 3)) ;
}
else memcpy(state->buffer + (state->datasize_in_buffer >> 3), data, 64-(state->datasize_in_buffer >> 3)+1) ;
state->datasize_in_buffer += databitlen;
databitlen = 0;
}
/*There is data in the buffer, and the incoming data is sufficient for a full block*/
if ( (state->datasize_in_buffer > 0 ) && (( state->datasize_in_buffer + databitlen) >= 512) ) {
memcpy( state->buffer + (state->datasize_in_buffer >> 3), data, 64-(state->datasize_in_buffer >> 3) ) ;
index = 64-(state->datasize_in_buffer >> 3);
databitlen = databitlen - (512 - state->datasize_in_buffer);
F8(state);
state->datasize_in_buffer = 0;
}
/*hash the remaining full message blocks*/
for ( ; databitlen >= 512; index = index+64, databitlen = databitlen - 512) {
memcpy(state->buffer, data+index, 64);
F8(state);
}
/*store the partial block into buffer, assume that -- if part of the last byte is not part of the message, then that part consists of 0 bits*/
if ( databitlen > 0) {
if ((databitlen & 7) == 0)
memcpy(state->buffer, data+index, (databitlen & 0x1ff) >> 3);
else
memcpy(state->buffer, data+index, ((databitlen & 0x1ff) >> 3)+1);
state->datasize_in_buffer = databitlen;
}
return(SUCCESS);
}
/*pad the message, process the padded block(s), truncate the hash value H to obtain the message digest*/
static HashReturn Final(hashState *state, BitSequence *hashval)
{
unsigned int i;
if ( (state->databitlen & 0x1ff) == 0 ) {
/*pad the message when databitlen is multiple of 512 bits, then process the padded block*/
memset(state->buffer, 0, 64);
state->buffer[0] = 0x80;
state->buffer[63] = state->databitlen & 0xff;
state->buffer[62] = (state->databitlen >> 8) & 0xff;
state->buffer[61] = (state->databitlen >> 16) & 0xff;
state->buffer[60] = (state->databitlen >> 24) & 0xff;
state->buffer[59] = (state->databitlen >> 32) & 0xff;
state->buffer[58] = (state->databitlen >> 40) & 0xff;
state->buffer[57] = (state->databitlen >> 48) & 0xff;
state->buffer[56] = (state->databitlen >> 56) & 0xff;
F8(state);
}
else {
/*set the rest of the bytes in the buffer to 0*/
if ( (state->datasize_in_buffer & 7) == 0)
for (i = (state->databitlen & 0x1ff) >> 3; i < 64; i++) state->buffer[i] = 0;
else
for (i = ((state->databitlen & 0x1ff) >> 3)+1; i < 64; i++) state->buffer[i] = 0;
/*pad and process the partial block when databitlen is not multiple of 512 bits, then hash the padded blocks*/
state->buffer[((state->databitlen & 0x1ff) >> 3)] |= 1 << (7- (state->databitlen & 7));
F8(state);
memset(state->buffer, 0, 64);
state->buffer[63] = state->databitlen & 0xff;
state->buffer[62] = (state->databitlen >> 8) & 0xff;
state->buffer[61] = (state->databitlen >> 16) & 0xff;
state->buffer[60] = (state->databitlen >> 24) & 0xff;
state->buffer[59] = (state->databitlen >> 32) & 0xff;
state->buffer[58] = (state->databitlen >> 40) & 0xff;
state->buffer[57] = (state->databitlen >> 48) & 0xff;
state->buffer[56] = (state->databitlen >> 56) & 0xff;
F8(state);
}
/*truncating the final hash value to generate the message digest*/
switch(state->hashbitlen) {
case 224: memcpy(hashval,(unsigned char*)state->x+64+36,28); break;
case 256: memcpy(hashval,(unsigned char*)state->x+64+32,32); break;
case 384: memcpy(hashval,(unsigned char*)state->x+64+16,48); break;
case 512: memcpy(hashval,(unsigned char*)state->x+64,64); break;
}
return(SUCCESS);
}
/* hash a message,
three inputs: message digest size in bits (hashbitlen); message (data); message length in bits (databitlen)
one output: message digest (hashval)
*/
HashReturn jh_hash(int hashbitlen, const BitSequence *data,DataLength databitlen, BitSequence *hashval)
{
hashState state;
if ( hashbitlen == 224 || hashbitlen == 256 || hashbitlen == 384 || hashbitlen == 512 ) {
Init(&state, hashbitlen);
Update(&state, data, databitlen);
Final(&state, hashval);
return SUCCESS;
}
else
return(BAD_HASHLEN);
}

19
crypto/c_jh.h Normal file
View File

@ -0,0 +1,19 @@
/*This program gives the 64-bit optimized bitslice implementation of JH using ANSI C
--------------------------------
Performance
Microprocessor: Intel CORE 2 processor (Core 2 Duo Mobile T6600 2.2GHz)
Operating System: 64-bit Ubuntu 10.04 (Linux kernel 2.6.32-22-generic)
Speed for long message:
1) 45.8 cycles/byte compiler: Intel C++ Compiler 11.1 compilation option: icc -O2
2) 56.8 cycles/byte compiler: gcc 4.4.3 compilation option: gcc -O3
--------------------------------
Last Modified: January 16, 2011
*/
#pragma once
#include "hash.h"
HashReturn jh_hash(int hashbitlen, const BitSequence *data, DataLength databitlen, BitSequence *hashval);

128
crypto/c_keccak.c Normal file
View File

@ -0,0 +1,128 @@
// keccak.c
// 19-Nov-11 Markku-Juhani O. Saarinen <mjos@iki.fi>
// A baseline Keccak (3rd round) implementation.
#include <stdint.h>
#include <memory.h>
#define HASH_DATA_AREA 136
#define KECCAK_ROUNDS 24
#ifndef ROTL64
#define ROTL64(x, y) (((x) << (y)) | ((x) >> (64 - (y))))
#endif
const uint64_t keccakf_rndc[24] =
{
0x0000000000000001, 0x0000000000008082, 0x800000000000808a,
0x8000000080008000, 0x000000000000808b, 0x0000000080000001,
0x8000000080008081, 0x8000000000008009, 0x000000000000008a,
0x0000000000000088, 0x0000000080008009, 0x000000008000000a,
0x000000008000808b, 0x800000000000008b, 0x8000000000008089,
0x8000000000008003, 0x8000000000008002, 0x8000000000000080,
0x000000000000800a, 0x800000008000000a, 0x8000000080008081,
0x8000000000008080, 0x0000000080000001, 0x8000000080008008
};
const int keccakf_rotc[24] =
{
1, 3, 6, 10, 15, 21, 28, 36, 45, 55, 2, 14,
27, 41, 56, 8, 25, 43, 62, 18, 39, 61, 20, 44
};
const int keccakf_piln[24] =
{
10, 7, 11, 17, 18, 3, 5, 16, 8, 21, 24, 4,
15, 23, 19, 13, 12, 2, 20, 14, 22, 9, 6, 1
};
// update the state with given number of rounds
void keccakf(uint64_t st[25], int rounds)
{
int i, j, round;
uint64_t t, bc[5];
for (round = 0; round < rounds; ++round) {
// Theta
bc[0] = st[0] ^ st[5] ^ st[10] ^ st[15] ^ st[20];
bc[1] = st[1] ^ st[6] ^ st[11] ^ st[16] ^ st[21];
bc[2] = st[2] ^ st[7] ^ st[12] ^ st[17] ^ st[22];
bc[3] = st[3] ^ st[8] ^ st[13] ^ st[18] ^ st[23];
bc[4] = st[4] ^ st[9] ^ st[14] ^ st[19] ^ st[24];
for (i = 0; i < 5; ++i) {
t = bc[(i + 4) % 5] ^ ROTL64(bc[(i + 1) % 5], 1);
st[i ] ^= t;
st[i + 5] ^= t;
st[i + 10] ^= t;
st[i + 15] ^= t;
st[i + 20] ^= t;
}
// Rho Pi
t = st[1];
for (i = 0; i < 24; ++i) {
bc[0] = st[keccakf_piln[i]];
st[keccakf_piln[i]] = ROTL64(t, keccakf_rotc[i]);
t = bc[0];
}
// Chi
for (j = 0; j < 25; j += 5) {
bc[0] = st[j ];
bc[1] = st[j + 1];
bc[2] = st[j + 2];
bc[3] = st[j + 3];
bc[4] = st[j + 4];
st[j ] ^= (~bc[1]) & bc[2];
st[j + 1] ^= (~bc[2]) & bc[3];
st[j + 2] ^= (~bc[3]) & bc[4];
st[j + 3] ^= (~bc[4]) & bc[0];
st[j + 4] ^= (~bc[0]) & bc[1];
}
// Iota
st[0] ^= keccakf_rndc[round];
}
}
// compute a keccak hash (md) of given byte length from "in"
typedef uint64_t state_t[25];
void keccak(const uint8_t *in, int inlen, uint8_t *md, int mdlen)
{
state_t st;
uint8_t temp[144];
int i, rsiz, rsizw;
rsiz = sizeof(state_t) == mdlen ? HASH_DATA_AREA : 200 - 2 * mdlen;
rsizw = rsiz / 8;
memset(st, 0, sizeof(st));
for ( ; inlen >= rsiz; inlen -= rsiz, in += rsiz) {
for (i = 0; i < rsizw; i++)
st[i] ^= ((uint64_t *) in)[i];
keccakf(st, KECCAK_ROUNDS);
}
// last block and padding
memcpy(temp, in, inlen);
temp[inlen++] = 1;
memset(temp + inlen, 0, rsiz - inlen);
temp[rsiz - 1] |= 0x80;
for (i = 0; i < rsizw; i++)
st[i] ^= ((uint64_t *) temp)[i];
keccakf(st, KECCAK_ROUNDS);
memcpy(md, st, mdlen);
}
void keccak1600(const uint8_t *in, int inlen, uint8_t *md)
{
keccak(in, inlen, md, sizeof(state_t));
}

26
crypto/c_keccak.h Normal file
View File

@ -0,0 +1,26 @@
// keccak.h
// 19-Nov-11 Markku-Juhani O. Saarinen <mjos@iki.fi>
#ifndef KECCAK_H
#define KECCAK_H
#include <stdint.h>
#include <string.h>
#ifndef KECCAK_ROUNDS
#define KECCAK_ROUNDS 24
#endif
#ifndef ROTL64
#define ROTL64(x, y) (((x) << (y)) | ((x) >> (64 - (y))))
#endif
// compute a keccak hash (md) of given byte length from "in"
int keccak(const uint8_t *in, int inlen, uint8_t *md, int mdlen);
// update the state
void keccakf(uint64_t st[25], int norounds);
void keccak1600(const uint8_t *in, int inlen, uint8_t *md);
#endif

2036
crypto/c_skein.c Normal file
View File

File diff suppressed because it is too large Load Diff

47
crypto/c_skein.h Normal file
View File

@ -0,0 +1,47 @@
#ifndef _SKEIN_H_
#define _SKEIN_H_ 1
/**************************************************************************
**
** Interface declarations and internal definitions for Skein hashing.
**
** Source code author: Doug Whiting, 2008.
**
** This algorithm and source code is released to the public domain.
**
***************************************************************************
**
** The following compile-time switches may be defined to control some
** tradeoffs between speed, code size, error checking, and security.
**
** The "default" note explains what happens when the switch is not defined.
**
** SKEIN_DEBUG -- make callouts from inside Skein code
** to examine/display intermediate values.
** [default: no callouts (no overhead)]
**
** SKEIN_ERR_CHECK -- how error checking is handled inside Skein
** code. If not defined, most error checking
** is disabled (for performance). Otherwise,
** the switch value is interpreted as:
** 0: use assert() to flag errors
** 1: return SKEIN_FAIL to flag errors
**
***************************************************************************/
#include "skein_port.h" /* get platform-specific definitions */
typedef enum
{
SKEIN_SUCCESS = 0, /* return codes from Skein calls */
SKEIN_FAIL = 1,
SKEIN_BAD_HASHLEN = 2
}
SkeinHashReturn;
typedef size_t SkeinDataLength; /* bit count type */
typedef u08b_t SkeinBitSequence; /* bit stream type */
/* "all-in-one" call */
SkeinHashReturn skein_hash(int hashbitlen, const SkeinBitSequence *data,
SkeinDataLength databitlen, SkeinBitSequence *hashval);
#endif /* ifndef _SKEIN_H_ */

25
crypto/cryptonight.h Normal file
View File

@ -0,0 +1,25 @@
#ifndef __CRYPTONIGHT_H_INCLUDED
#define __CRYPTONIGHT_H_INCLUDED
#ifdef __cplusplus
extern "C" {
#endif
#include <stddef.h>
#include <inttypes.h>
#define MEMORY 2097152
typedef struct {
uint8_t hash_state[224]; // Need only 200, explicit align
uint8_t long_state[MEMORY];
} cryptonight_ctx;
void cryptonight_hash_ctx(const void* input, size_t len, void* output, cryptonight_ctx* ctx);
void cryptonight_hash_ctx_soft(const void* input, size_t len, void* output, cryptonight_ctx* ctx);
#ifdef __cplusplus
}
#endif
#endif

420
crypto/cryptonight_aesni.h Normal file
View File

@ -0,0 +1,420 @@
/*
* 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
* 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 <http://www.gnu.org/licenses/>.
*
*/
#pragma once
#include "cryptonight.h"
#include <memory.h>
#include <stdio.h>
#ifdef __GNUC__
#include <x86intrin.h>
static inline uint64_t _umul128(uint64_t a, uint64_t b, uint64_t* hi)
{
unsigned __int128 r = (unsigned __int128)a * (unsigned __int128)b;
*hi = r >> 64;
return (uint64_t)r;
}
#define _mm256_set_m128i(v0, v1) _mm256_insertf128_si256(_mm256_castsi128_si256(v1), (v0), 1)
#else
#include <intrin.h>
#endif // __GNUC__
#if !defined(_LP64) && !defined(_WIN64)
#error You are trying to do a 32-bit build. This will all end in tears. I know it.
#endif
extern "C"
{
void keccak(const uint8_t *in, int inlen, uint8_t *md, int mdlen);
void keccakf(uint64_t st[25], int rounds);
extern void(*const extra_hashes[4])(const void *, size_t, char *);
__m128i soft_aesenc(__m128i in, __m128i key);
__m128i soft_aeskeygenassist(__m128i key, uint8_t rcon);
}
// This will shift and xor tmp1 into itself as 4 32-bit vals such as
// sl_xor(a1 a2 a3 a4) = a1 (a2^a1) (a3^a2^a1) (a4^a3^a2^a1)
static inline __m128i sl_xor(__m128i tmp1)
{
__m128i tmp4;
tmp4 = _mm_slli_si128(tmp1, 0x04);
tmp1 = _mm_xor_si128(tmp1, tmp4);
tmp4 = _mm_slli_si128(tmp4, 0x04);
tmp1 = _mm_xor_si128(tmp1, tmp4);
tmp4 = _mm_slli_si128(tmp4, 0x04);
tmp1 = _mm_xor_si128(tmp1, tmp4);
return tmp1;
}
template<uint8_t rcon>
static inline void aes_genkey_sub(__m128i* xout0, __m128i* xout2)
{
__m128i xout1 = _mm_aeskeygenassist_si128(*xout2, rcon);
xout1 = _mm_shuffle_epi32(xout1, 0xFF); // see PSHUFD, set all elems to 4th elem
*xout0 = sl_xor(*xout0);
*xout0 = _mm_xor_si128(*xout0, xout1);
xout1 = _mm_aeskeygenassist_si128(*xout0, 0x00);
xout1 = _mm_shuffle_epi32(xout1, 0xAA); // see PSHUFD, set all elems to 3rd elem
*xout2 = sl_xor(*xout2);
*xout2 = _mm_xor_si128(*xout2, xout1);
}
static inline void soft_aes_genkey_sub(__m128i* xout0, __m128i* xout2, uint8_t rcon)
{
__m128i xout1 = soft_aeskeygenassist(*xout2, rcon);
xout1 = _mm_shuffle_epi32(xout1, 0xFF); // see PSHUFD, set all elems to 4th elem
*xout0 = sl_xor(*xout0);
*xout0 = _mm_xor_si128(*xout0, xout1);
xout1 = soft_aeskeygenassist(*xout0, 0x00);
xout1 = _mm_shuffle_epi32(xout1, 0xAA); // see PSHUFD, set all elems to 3rd elem
*xout2 = sl_xor(*xout2);
*xout2 = _mm_xor_si128(*xout2, xout1);
}
template<bool SOFT_AES>
static inline void aes_genkey(const __m128i* memory, __m128i* k0, __m128i* k1, __m128i* k2, __m128i* k3,
__m128i* k4, __m128i* k5, __m128i* k6, __m128i* k7, __m128i* k8, __m128i* k9)
{
__m128i xout0, xout2;
xout0 = _mm_load_si128(memory);
xout2 = _mm_load_si128(memory+1);
*k0 = xout0;
*k1 = xout2;
if(SOFT_AES)
soft_aes_genkey_sub(&xout0, &xout2, 0x01);
else
aes_genkey_sub<0x01>(&xout0, &xout2);
*k2 = xout0;
*k3 = xout2;
if(SOFT_AES)
soft_aes_genkey_sub(&xout0, &xout2, 0x02);
else
aes_genkey_sub<0x02>(&xout0, &xout2);
*k4 = xout0;
*k5 = xout2;
if(SOFT_AES)
soft_aes_genkey_sub(&xout0, &xout2, 0x04);
else
aes_genkey_sub<0x04>(&xout0, &xout2);
*k6 = xout0;
*k7 = xout2;
if(SOFT_AES)
soft_aes_genkey_sub(&xout0, &xout2, 0x08);
else
aes_genkey_sub<0x08>(&xout0, &xout2);
*k8 = xout0;
*k9 = xout2;
}
static inline void aes_round(__m128i key, __m128i* x0, __m128i* x1, __m128i* x2, __m128i* x3, __m128i* x4, __m128i* x5, __m128i* x6, __m128i* x7)
{
*x0 = _mm_aesenc_si128(*x0, key);
*x1 = _mm_aesenc_si128(*x1, key);
*x2 = _mm_aesenc_si128(*x2, key);
*x3 = _mm_aesenc_si128(*x3, key);
*x4 = _mm_aesenc_si128(*x4, key);
*x5 = _mm_aesenc_si128(*x5, key);
*x6 = _mm_aesenc_si128(*x6, key);
*x7 = _mm_aesenc_si128(*x7, key);
}
static inline void soft_aes_round(__m128i key, __m128i* x0, __m128i* x1, __m128i* x2, __m128i* x3, __m128i* x4, __m128i* x5, __m128i* x6, __m128i* x7)
{
*x0 = soft_aesenc(*x0, key);
*x1 = soft_aesenc(*x1, key);
*x2 = soft_aesenc(*x2, key);
*x3 = soft_aesenc(*x3, key);
*x4 = soft_aesenc(*x4, key);
*x5 = soft_aesenc(*x5, key);
*x6 = soft_aesenc(*x6, key);
*x7 = soft_aesenc(*x7, key);
}
template<size_t MEM, bool SOFT_AES>
void cn_explode_scratchpad(const __m128i* input, __m128i* output)
{
// This is more than we have registers, compiler will assign 2 keys on the stack
__m128i xin0, xin1, xin2, xin3, xin4, xin5, xin6, xin7;
__m128i k0, k1, k2, k3, k4, k5, k6, k7, k8, k9;
aes_genkey<SOFT_AES>(input, &k0, &k1, &k2, &k3, &k4, &k5, &k6, &k7, &k8, &k9);
xin0 = _mm_load_si128(input + 4);
xin1 = _mm_load_si128(input + 5);
xin2 = _mm_load_si128(input + 6);
xin3 = _mm_load_si128(input + 7);
xin4 = _mm_load_si128(input + 8);
xin5 = _mm_load_si128(input + 9);
xin6 = _mm_load_si128(input + 10);
xin7 = _mm_load_si128(input + 11);
for (size_t i = 0; i < MEM / sizeof(__m128i); i += 8)
{
if(SOFT_AES)
{
soft_aes_round(k0, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
soft_aes_round(k1, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
soft_aes_round(k2, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
soft_aes_round(k3, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
soft_aes_round(k4, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
soft_aes_round(k5, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
soft_aes_round(k6, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
soft_aes_round(k7, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
soft_aes_round(k8, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
soft_aes_round(k9, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
}
else
{
aes_round(k0, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k1, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k2, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k3, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k4, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k5, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k6, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k7, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k8, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k9, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
}
_mm_store_si128(output + i + 0, xin0);
_mm_store_si128(output + i + 1, xin1);
_mm_store_si128(output + i + 2, xin2);
_mm_store_si128(output + i + 3, xin3);
_mm_prefetch((const char*)output + i + 0, _MM_HINT_T2);
_mm_store_si128(output + i + 4, xin4);
_mm_store_si128(output + i + 5, xin5);
_mm_store_si128(output + i + 6, xin6);
_mm_store_si128(output + i + 7, xin7);
_mm_prefetch((const char*)output + i + 4, _MM_HINT_T2);
}
}
template<size_t MEM, bool SOFT_AES>
void cn_implode_scratchpad(const __m128i* input, __m128i* output)
{
// This is more than we have registers, compiler will assign 2 keys on the stack
__m128i xout0, xout1, xout2, xout3, xout4, xout5, xout6, xout7;
__m128i k0, k1, k2, k3, k4, k5, k6, k7, k8, k9;
aes_genkey<SOFT_AES>(output + 2, &k0, &k1, &k2, &k3, &k4, &k5, &k6, &k7, &k8, &k9);
xout0 = _mm_load_si128(output + 4);
xout1 = _mm_load_si128(output + 5);
xout2 = _mm_load_si128(output + 6);
xout3 = _mm_load_si128(output + 7);
xout4 = _mm_load_si128(output + 8);
xout5 = _mm_load_si128(output + 9);
xout6 = _mm_load_si128(output + 10);
xout7 = _mm_load_si128(output + 11);
for (size_t i = 0; i < MEM / sizeof(__m128i); i += 8)
{
_mm_prefetch((const char*)input + i + 0, _MM_HINT_NTA);
xout0 = _mm_xor_si128(_mm_load_si128(input + i + 0), xout0);
xout1 = _mm_xor_si128(_mm_load_si128(input + i + 1), xout1);
xout2 = _mm_xor_si128(_mm_load_si128(input + i + 2), xout2);
xout3 = _mm_xor_si128(_mm_load_si128(input + i + 3), xout3);
_mm_prefetch((const char*)input + i + 4, _MM_HINT_NTA);
xout4 = _mm_xor_si128(_mm_load_si128(input + i + 4), xout4);
xout5 = _mm_xor_si128(_mm_load_si128(input + i + 5), xout5);
xout6 = _mm_xor_si128(_mm_load_si128(input + i + 6), xout6);
xout7 = _mm_xor_si128(_mm_load_si128(input + i + 7), xout7);
if(SOFT_AES)
{
soft_aes_round(k0, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
soft_aes_round(k1, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
soft_aes_round(k2, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
soft_aes_round(k3, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
soft_aes_round(k4, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
soft_aes_round(k5, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
soft_aes_round(k6, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
soft_aes_round(k7, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
soft_aes_round(k8, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
soft_aes_round(k9, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
}
else
{
aes_round(k0, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
aes_round(k1, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
aes_round(k2, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
aes_round(k3, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
aes_round(k4, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
aes_round(k5, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
aes_round(k6, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
aes_round(k7, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
aes_round(k8, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
aes_round(k9, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
}
}
_mm_store_si128(output + 4, xout0);
_mm_store_si128(output + 5, xout1);
_mm_store_si128(output + 6, xout2);
_mm_store_si128(output + 7, xout3);
_mm_store_si128(output + 8, xout4);
_mm_store_si128(output + 9, xout5);
_mm_store_si128(output + 10, xout6);
_mm_store_si128(output + 11, xout7);
}
template<size_t ITERATIONS, size_t MEM, bool PREFETCH, bool SOFT_AES>
void cryptonight_hash(const void* input, size_t len, void* output, cryptonight_ctx* ctx0)
{
keccak((const uint8_t *)input, len, ctx0->hash_state, 200);
// Optim - 99% time boundary
cn_explode_scratchpad<MEM, SOFT_AES>((__m128i*)ctx0->hash_state, (__m128i*)ctx0->long_state);
uint8_t* l0 = ctx0->long_state;
uint64_t* h0 = (uint64_t*)ctx0->hash_state;
uint64_t al0 = h0[0] ^ h0[4];
uint64_t ah0 = h0[1] ^ h0[5];
__m128i bx0 = _mm_set_epi64x(h0[3] ^ h0[7], h0[2] ^ h0[6]);
uint64_t idx0 = h0[0] ^ h0[4];
// Optim - 90% time boundary
for(size_t i = 0; i < ITERATIONS; i++)
{
__m128i cx;
cx = _mm_load_si128((__m128i *)&l0[idx0 & 0x1FFFF0]);
if(SOFT_AES)
cx = soft_aesenc(cx, _mm_set_epi64x(ah0, al0));
else
cx = _mm_aesenc_si128(cx, _mm_set_epi64x(ah0, al0));
_mm_store_si128((__m128i *)&l0[idx0 & 0x1FFFF0], _mm_xor_si128(bx0, cx));
idx0 = _mm_cvtsi128_si64(cx);
bx0 = cx;
if(PREFETCH)
_mm_prefetch((const char*)&l0[idx0 & 0x1FFFF0], _MM_HINT_T0);
uint64_t hi, lo, cl, ch;
cl = ((uint64_t*)&l0[idx0 & 0x1FFFF0])[0];
ch = ((uint64_t*)&l0[idx0 & 0x1FFFF0])[1];
lo = _umul128(idx0, cl, &hi);
al0 += hi;
ah0 += lo;
((uint64_t*)&l0[idx0 & 0x1FFFF0])[0] = al0;
((uint64_t*)&l0[idx0 & 0x1FFFF0])[1] = ah0;
ah0 ^= ch;
al0 ^= cl;
idx0 = al0;
if(PREFETCH)
_mm_prefetch((const char*)&l0[idx0 & 0x1FFFF0], _MM_HINT_T0);
}
// Optim - 90% time boundary
cn_implode_scratchpad<MEM, SOFT_AES>((__m128i*)ctx0->long_state, (__m128i*)ctx0->hash_state);
// Optim - 99% time boundary
keccakf((uint64_t*)ctx0->hash_state, 24);
extra_hashes[ctx0->hash_state[0] & 3](ctx0->hash_state, 200, (char*)output);
}
// This lovely creation will do 2 cn hashes at a time. We have plenty of space on silicon
// to fit temporary vars for two contexts. Function will read len*2 from input and write 64 bytes to output
// We are still limited by L3 cache, so doubling will only work with CPUs where we have more than 2MB to core (Xeons)
template<size_t ITERATIONS, size_t MEM, bool PREFETCH, bool SOFT_AES>
void cryptonight_double_hash(const void* input, size_t len, void* output, cryptonight_ctx* __restrict ctx0, cryptonight_ctx* __restrict ctx1)
{
keccak((const uint8_t *)input, len, ctx0->hash_state, 200);
keccak((const uint8_t *)input+len, len, ctx1->hash_state, 200);
// Optim - 99% time boundary
cn_explode_scratchpad<MEM, SOFT_AES>((__m128i*)ctx0->hash_state, (__m128i*)ctx0->long_state);
cn_explode_scratchpad<MEM, SOFT_AES>((__m128i*)ctx1->hash_state, (__m128i*)ctx1->long_state);
uint8_t* l0 = ctx0->long_state;
uint64_t* h0 = (uint64_t*)ctx0->hash_state;
uint8_t* l1 = ctx1->long_state;
uint64_t* h1 = (uint64_t*)ctx1->hash_state;
__m128i ax0 = _mm_set_epi64x(h0[1] ^ h0[5], h0[0] ^ h0[4]);
__m128i bx0 = _mm_set_epi64x(h0[3] ^ h0[7], h0[2] ^ h0[6]);
__m128i ax1 = _mm_set_epi64x(h1[1] ^ h1[5], h1[0] ^ h1[4]);
__m128i bx1 = _mm_set_epi64x(h1[3] ^ h1[7], h1[2] ^ h1[6]);
uint64_t idx0 = h0[0] ^ h0[4];
uint64_t idx1 = h1[0] ^ h1[4];
// Optim - 90% time boundary
for (size_t i = 0; i < ITERATIONS; i++)
{
__m128i cx;
cx = _mm_load_si128((__m128i *)&l0[idx0 & 0x1FFFF0]);
if(SOFT_AES)
cx = soft_aesenc(cx, ax0);
else
cx = _mm_aesenc_si128(cx, ax0);
_mm_store_si128((__m128i *)&l0[idx0 & 0x1FFFF0], _mm_xor_si128(bx0, cx));
idx0 = _mm_cvtsi128_si64(cx);
bx0 = cx;
if(PREFETCH)
_mm_prefetch((const char*)&l0[idx0 & 0x1FFFF0], _MM_HINT_T0);
cx = _mm_load_si128((__m128i *)&l1[idx1 & 0x1FFFF0]);
if(SOFT_AES)
cx = soft_aesenc(cx, ax1);
else
cx = _mm_aesenc_si128(cx, ax1);
_mm_store_si128((__m128i *)&l1[idx1 & 0x1FFFF0], _mm_xor_si128(bx1, cx));
idx1 = _mm_cvtsi128_si64(cx);
bx1 = cx;
if(PREFETCH)
_mm_prefetch((const char*)&l1[idx1 & 0x1FFFF0], _MM_HINT_T0);
uint64_t hi, lo;
cx = _mm_load_si128((__m128i *)&l0[idx0 & 0x1FFFF0]);
lo = _umul128(idx0, _mm_cvtsi128_si64(cx), &hi);
ax0 = _mm_add_epi64(ax0, _mm_set_epi64x(lo, hi));
_mm_store_si128((__m128i*)&l0[idx0 & 0x1FFFF0], ax0);
ax0 = _mm_xor_si128(ax0, cx);
idx0 = _mm_cvtsi128_si64(ax0);
if(PREFETCH)
_mm_prefetch((const char*)&l0[idx0 & 0x1FFFF0], _MM_HINT_T0);
cx = _mm_load_si128((__m128i *)&l1[idx1 & 0x1FFFF0]);
lo = _umul128(idx1, _mm_cvtsi128_si64(cx), &hi);
ax1 = _mm_add_epi64(ax1, _mm_set_epi64x(lo, hi));
_mm_store_si128((__m128i*)&l1[idx1 & 0x1FFFF0], ax1);
ax1 = _mm_xor_si128(ax1, cx);
idx1 = _mm_cvtsi128_si64(ax1);
if(PREFETCH)
_mm_prefetch((const char*)&l1[idx1 & 0x1FFFF0], _MM_HINT_T0);
}
// Optim - 90% time boundary
cn_implode_scratchpad<MEM, SOFT_AES>((__m128i*)ctx0->long_state, (__m128i*)ctx0->hash_state);
cn_implode_scratchpad<MEM, SOFT_AES>((__m128i*)ctx1->long_state, (__m128i*)ctx1->hash_state);
// Optim - 99% time boundary
keccakf((uint64_t*)ctx0->hash_state, 24);
extra_hashes[ctx0->hash_state[0] & 3](ctx0->hash_state, 200, (char*)output);
keccakf((uint64_t*)ctx1->hash_state, 24);
extra_hashes[ctx1->hash_state[0] & 3](ctx1->hash_state, 200, (char*)output + 32);
}

70
crypto/cryptonight_common.cpp Normal file
View File

@ -0,0 +1,70 @@
/*
* 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
* 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 <http://www.gnu.org/licenses/>.
*/
extern "C"
{
#include "c_groestl.h"
#include "c_blake256.h"
#include "c_jh.h"
#include "c_skein.h"
}
#include "cryptonight.h"
#include "cryptonight_aesni.h"
#include <stdio.h>
#include <stdlib.h>
#ifdef __GNUC__
#include <mm_malloc.h>
#else
#include <malloc.h>
#endif // __GNUC__
#ifdef _WIN32
#include <windows.h>
#else
#include <sys/mman.h>
#include <errno.h>
#include <string.h>
#endif // _WIN32
void do_blake_hash(const void* input, size_t len, char* output) {
blake256_hash((uint8_t*)output, (const uint8_t*)input, len);
}
void do_groestl_hash(const void* input, size_t len, char* output) {
groestl((const uint8_t*)input, len * 8, (uint8_t*)output);
}
void do_jh_hash(const void* input, size_t len, char* output) {
jh_hash(32 * 8, (const uint8_t*)input, 8 * len, (uint8_t*)output);
}
void do_skein_hash(const void* input, size_t len, char* output) {
skein_hash(8 * 32, (const uint8_t*)input, 8 * len, (uint8_t*)output);
}
void (* const extra_hashes[4])(const void *, size_t, char *) = {do_blake_hash, do_groestl_hash, do_jh_hash, do_skein_hash};
void cryptonight_hash_ctx(const void* input, size_t len, void* output, cryptonight_ctx* ctx)
{
cryptonight_hash<0x80000, MEMORY, true, false>(input, len, output, ctx);
}
void cryptonight_hash_ctx_soft(const void* input, size_t len, void* output, cryptonight_ctx* ctx)
{
cryptonight_hash<0x80000, MEMORY, true, true>(input, len, output, ctx);
}

38
crypto/groestl_tables.h Normal file
View File

@ -0,0 +1,38 @@
#ifndef __tables_h
#define __tables_h
const uint32_t T[512] = {0xa5f432c6, 0xc6a597f4, 0x84976ff8, 0xf884eb97, 0x99b05eee, 0xee99c7b0, 0x8d8c7af6, 0xf68df78c, 0xd17e8ff, 0xff0de517, 0xbddc0ad6, 0xd6bdb7dc, 0xb1c816de, 0xdeb1a7c8, 0x54fc6d91, 0x915439fc
, 0x50f09060, 0x6050c0f0, 0x3050702, 0x2030405, 0xa9e02ece, 0xcea987e0, 0x7d87d156, 0x567dac87, 0x192bcce7, 0xe719d52b, 0x62a613b5, 0xb56271a6, 0xe6317c4d, 0x4de69a31, 0x9ab559ec, 0xec9ac3b5
, 0x45cf408f, 0x8f4505cf, 0x9dbca31f, 0x1f9d3ebc, 0x40c04989, 0x894009c0, 0x879268fa, 0xfa87ef92, 0x153fd0ef, 0xef15c53f, 0xeb2694b2, 0xb2eb7f26, 0xc940ce8e, 0x8ec90740, 0xb1de6fb, 0xfb0bed1d
, 0xec2f6e41, 0x41ec822f, 0x67a91ab3, 0xb3677da9, 0xfd1c435f, 0x5ffdbe1c, 0xea256045, 0x45ea8a25, 0xbfdaf923, 0x23bf46da, 0xf7025153, 0x53f7a602, 0x96a145e4, 0xe496d3a1, 0x5bed769b, 0x9b5b2ded
, 0xc25d2875, 0x75c2ea5d, 0x1c24c5e1, 0xe11cd924, 0xaee9d43d, 0x3dae7ae9, 0x6abef24c, 0x4c6a98be, 0x5aee826c, 0x6c5ad8ee, 0x41c3bd7e, 0x7e41fcc3, 0x206f3f5, 0xf502f106, 0x4fd15283, 0x834f1dd1
, 0x5ce48c68, 0x685cd0e4, 0xf4075651, 0x51f4a207, 0x345c8dd1, 0xd134b95c, 0x818e1f9, 0xf908e918, 0x93ae4ce2, 0xe293dfae, 0x73953eab, 0xab734d95, 0x53f59762, 0x6253c4f5, 0x3f416b2a, 0x2a3f5441
, 0xc141c08, 0x80c1014, 0x52f66395, 0x955231f6, 0x65afe946, 0x46658caf, 0x5ee27f9d, 0x9d5e21e2, 0x28784830, 0x30286078, 0xa1f8cf37, 0x37a16ef8, 0xf111b0a, 0xa0f1411, 0xb5c4eb2f, 0x2fb55ec4
, 0x91b150e, 0xe091c1b, 0x365a7e24, 0x2436485a, 0x9bb6ad1b, 0x1b9b36b6, 0x3d4798df, 0xdf3da547, 0x266aa7cd, 0xcd26816a, 0x69bbf54e, 0x4e699cbb, 0xcd4c337f, 0x7fcdfe4c, 0x9fba50ea, 0xea9fcfba
, 0x1b2d3f12, 0x121b242d, 0x9eb9a41d, 0x1d9e3ab9, 0x749cc458, 0x5874b09c, 0x2e724634, 0x342e6872, 0x2d774136, 0x362d6c77, 0xb2cd11dc, 0xdcb2a3cd, 0xee299db4, 0xb4ee7329, 0xfb164d5b, 0x5bfbb616
, 0xf601a5a4, 0xa4f65301, 0x4dd7a176, 0x764decd7, 0x61a314b7, 0xb76175a3, 0xce49347d, 0x7dcefa49, 0x7b8ddf52, 0x527ba48d, 0x3e429fdd, 0xdd3ea142, 0x7193cd5e, 0x5e71bc93, 0x97a2b113, 0x139726a2
, 0xf504a2a6, 0xa6f55704, 0x68b801b9, 0xb96869b8, 0x0, 0x0, 0x2c74b5c1, 0xc12c9974, 0x60a0e040, 0x406080a0, 0x1f21c2e3, 0xe31fdd21, 0xc8433a79, 0x79c8f243, 0xed2c9ab6, 0xb6ed772c
, 0xbed90dd4, 0xd4beb3d9, 0x46ca478d, 0x8d4601ca, 0xd9701767, 0x67d9ce70, 0x4bddaf72, 0x724be4dd, 0xde79ed94, 0x94de3379, 0xd467ff98, 0x98d42b67, 0xe82393b0, 0xb0e87b23, 0x4ade5b85, 0x854a11de
, 0x6bbd06bb, 0xbb6b6dbd, 0x2a7ebbc5, 0xc52a917e, 0xe5347b4f, 0x4fe59e34, 0x163ad7ed, 0xed16c13a, 0xc554d286, 0x86c51754, 0xd762f89a, 0x9ad72f62, 0x55ff9966, 0x6655ccff, 0x94a7b611, 0x119422a7
, 0xcf4ac08a, 0x8acf0f4a, 0x1030d9e9, 0xe910c930, 0x60a0e04, 0x406080a, 0x819866fe, 0xfe81e798, 0xf00baba0, 0xa0f05b0b, 0x44ccb478, 0x7844f0cc, 0xbad5f025, 0x25ba4ad5, 0xe33e754b, 0x4be3963e
, 0xf30eaca2, 0xa2f35f0e, 0xfe19445d, 0x5dfeba19, 0xc05bdb80, 0x80c01b5b, 0x8a858005, 0x58a0a85, 0xadecd33f, 0x3fad7eec, 0xbcdffe21, 0x21bc42df, 0x48d8a870, 0x7048e0d8, 0x40cfdf1, 0xf104f90c
, 0xdf7a1963, 0x63dfc67a, 0xc1582f77, 0x77c1ee58, 0x759f30af, 0xaf75459f, 0x63a5e742, 0x426384a5, 0x30507020, 0x20304050, 0x1a2ecbe5, 0xe51ad12e, 0xe12effd, 0xfd0ee112, 0x6db708bf, 0xbf6d65b7
, 0x4cd45581, 0x814c19d4, 0x143c2418, 0x1814303c, 0x355f7926, 0x26354c5f, 0x2f71b2c3, 0xc32f9d71, 0xe13886be, 0xbee16738, 0xa2fdc835, 0x35a26afd, 0xcc4fc788, 0x88cc0b4f, 0x394b652e, 0x2e395c4b
, 0x57f96a93, 0x93573df9, 0xf20d5855, 0x55f2aa0d, 0x829d61fc, 0xfc82e39d, 0x47c9b37a, 0x7a47f4c9, 0xacef27c8, 0xc8ac8bef, 0xe73288ba, 0xbae76f32, 0x2b7d4f32, 0x322b647d, 0x95a442e6, 0xe695d7a4
, 0xa0fb3bc0, 0xc0a09bfb, 0x98b3aa19, 0x199832b3, 0xd168f69e, 0x9ed12768, 0x7f8122a3, 0xa37f5d81, 0x66aaee44, 0x446688aa, 0x7e82d654, 0x547ea882, 0xabe6dd3b, 0x3bab76e6, 0x839e950b, 0xb83169e
, 0xca45c98c, 0x8cca0345, 0x297bbcc7, 0xc729957b, 0xd36e056b, 0x6bd3d66e, 0x3c446c28, 0x283c5044, 0x798b2ca7, 0xa779558b, 0xe23d81bc, 0xbce2633d, 0x1d273116, 0x161d2c27, 0x769a37ad, 0xad76419a
, 0x3b4d96db, 0xdb3bad4d, 0x56fa9e64, 0x6456c8fa, 0x4ed2a674, 0x744ee8d2, 0x1e223614, 0x141e2822, 0xdb76e492, 0x92db3f76, 0xa1e120c, 0xc0a181e, 0x6cb4fc48, 0x486c90b4, 0xe4378fb8, 0xb8e46b37
, 0x5de7789f, 0x9f5d25e7, 0x6eb20fbd, 0xbd6e61b2, 0xef2a6943, 0x43ef862a, 0xa6f135c4, 0xc4a693f1, 0xa8e3da39, 0x39a872e3, 0xa4f7c631, 0x31a462f7, 0x37598ad3, 0xd337bd59, 0x8b8674f2, 0xf28bff86
, 0x325683d5, 0xd532b156, 0x43c54e8b, 0x8b430dc5, 0x59eb856e, 0x6e59dceb, 0xb7c218da, 0xdab7afc2, 0x8c8f8e01, 0x18c028f, 0x64ac1db1, 0xb16479ac, 0xd26df19c, 0x9cd2236d, 0xe03b7249, 0x49e0923b
, 0xb4c71fd8, 0xd8b4abc7, 0xfa15b9ac, 0xacfa4315, 0x709faf3, 0xf307fd09, 0x256fa0cf, 0xcf25856f, 0xafea20ca, 0xcaaf8fea, 0x8e897df4, 0xf48ef389, 0xe9206747, 0x47e98e20, 0x18283810, 0x10182028
, 0xd5640b6f, 0x6fd5de64, 0x888373f0, 0xf088fb83, 0x6fb1fb4a, 0x4a6f94b1, 0x7296ca5c, 0x5c72b896, 0x246c5438, 0x3824706c, 0xf1085f57, 0x57f1ae08, 0xc7522173, 0x73c7e652, 0x51f36497, 0x975135f3
, 0x2365aecb, 0xcb238d65, 0x7c8425a1, 0xa17c5984, 0x9cbf57e8, 0xe89ccbbf, 0x21635d3e, 0x3e217c63, 0xdd7cea96, 0x96dd377c, 0xdc7f1e61, 0x61dcc27f, 0x86919c0d, 0xd861a91, 0x85949b0f, 0xf851e94
, 0x90ab4be0, 0xe090dbab, 0x42c6ba7c, 0x7c42f8c6, 0xc4572671, 0x71c4e257, 0xaae529cc, 0xccaa83e5, 0xd873e390, 0x90d83b73, 0x50f0906, 0x6050c0f, 0x103f4f7, 0xf701f503, 0x12362a1c, 0x1c123836
, 0xa3fe3cc2, 0xc2a39ffe, 0x5fe18b6a, 0x6a5fd4e1, 0xf910beae, 0xaef94710, 0xd06b0269, 0x69d0d26b, 0x91a8bf17, 0x17912ea8, 0x58e87199, 0x995829e8, 0x2769533a, 0x3a277469, 0xb9d0f727, 0x27b94ed0
, 0x384891d9, 0xd938a948, 0x1335deeb, 0xeb13cd35, 0xb3cee52b, 0x2bb356ce, 0x33557722, 0x22334455, 0xbbd604d2, 0xd2bbbfd6, 0x709039a9, 0xa9704990, 0x89808707, 0x7890e80, 0xa7f2c133, 0x33a766f2
, 0xb6c1ec2d, 0x2db65ac1, 0x22665a3c, 0x3c227866, 0x92adb815, 0x15922aad, 0x2060a9c9, 0xc9208960, 0x49db5c87, 0x874915db, 0xff1ab0aa, 0xaaff4f1a, 0x7888d850, 0x5078a088, 0x7a8e2ba5, 0xa57a518e
, 0x8f8a8903, 0x38f068a, 0xf8134a59, 0x59f8b213, 0x809b9209, 0x980129b, 0x1739231a, 0x1a173439, 0xda751065, 0x65daca75, 0x315384d7, 0xd731b553, 0xc651d584, 0x84c61351, 0xb8d303d0, 0xd0b8bbd3
, 0xc35edc82, 0x82c31f5e, 0xb0cbe229, 0x29b052cb, 0x7799c35a, 0x5a77b499, 0x11332d1e, 0x1e113c33, 0xcb463d7b, 0x7bcbf646, 0xfc1fb7a8, 0xa8fc4b1f, 0xd6610c6d, 0x6dd6da61, 0x3a4e622c, 0x2c3a584e};
#endif /* __tables_h */

5
crypto/hash.h Normal file
View File

@ -0,0 +1,5 @@
#pragma once
typedef unsigned char BitSequence;
typedef unsigned long long DataLength;
typedef enum {SUCCESS = 0, FAIL = 1, BAD_HASHLEN = 2} HashReturn;

153
crypto/int-util.h Normal file
View File

@ -0,0 +1,153 @@
// Copyright(c) 2012 - 2013 The Cryptonote developers
// Distributed under the MIT/X11 software license, see the accompanying
// file COPYING or http://www.opensource.org/licenses/mit-license.php.
#pragma once
#include <assert.h>
#include <stdbool.h>
#include <stdint.h>
#include <string.h>
#if defined(_MSC_VER)
#include <stdlib.h>
static inline uint32_t rol32(uint32_t x, int r) {
static_assert(sizeof(uint32_t) == sizeof(unsigned int), "this code assumes 32-bit integers");
return _rotl(x, r);
}
static inline uint64_t rol64(uint64_t x, int r) {
return _rotl64(x, r);
}
#else
static inline uint32_t rol32(uint32_t x, int r) {
return (x << (r & 31)) | (x >> (-r & 31));
}
static inline uint64_t rol64(uint64_t x, int r) {
return (x << (r & 63)) | (x >> (-r & 63));
}
#endif
static inline uint64_t hi_dword(uint64_t val) {
return val >> 32;
}
static inline uint64_t lo_dword(uint64_t val) {
return val & 0xFFFFFFFF;
}
static inline uint64_t div_with_reminder(uint64_t dividend, uint32_t divisor, uint32_t* remainder) {
dividend |= ((uint64_t)*remainder) << 32;
*remainder = dividend % divisor;
return dividend / divisor;
}
// Long division with 2^32 base
static inline uint32_t div128_32(uint64_t dividend_hi, uint64_t dividend_lo, uint32_t divisor, uint64_t* quotient_hi, uint64_t* quotient_lo) {
uint64_t dividend_dwords[4];
uint32_t remainder = 0;
dividend_dwords[3] = hi_dword(dividend_hi);
dividend_dwords[2] = lo_dword(dividend_hi);
dividend_dwords[1] = hi_dword(dividend_lo);
dividend_dwords[0] = lo_dword(dividend_lo);
*quotient_hi = div_with_reminder(dividend_dwords[3], divisor, &remainder) << 32;
*quotient_hi |= div_with_reminder(dividend_dwords[2], divisor, &remainder);
*quotient_lo = div_with_reminder(dividend_dwords[1], divisor, &remainder) << 32;
*quotient_lo |= div_with_reminder(dividend_dwords[0], divisor, &remainder);
return remainder;
}
#define IDENT32(x) ((uint32_t) (x))
#define IDENT64(x) ((uint64_t) (x))
#define SWAP32(x) ((((uint32_t) (x) & 0x000000ff) << 24) | \
(((uint32_t) (x) & 0x0000ff00) << 8) | \
(((uint32_t) (x) & 0x00ff0000) >> 8) | \
(((uint32_t) (x) & 0xff000000) >> 24))
#define SWAP64(x) ((((uint64_t) (x) & 0x00000000000000ff) << 56) | \
(((uint64_t) (x) & 0x000000000000ff00) << 40) | \
(((uint64_t) (x) & 0x0000000000ff0000) << 24) | \
(((uint64_t) (x) & 0x00000000ff000000) << 8) | \
(((uint64_t) (x) & 0x000000ff00000000) >> 8) | \
(((uint64_t) (x) & 0x0000ff0000000000) >> 24) | \
(((uint64_t) (x) & 0x00ff000000000000) >> 40) | \
(((uint64_t) (x) & 0xff00000000000000) >> 56))
static inline uint32_t ident32(uint32_t x) { return x; }
static inline uint64_t ident64(uint64_t x) { return x; }
static inline uint32_t swap32(uint32_t x) {
x = ((x & 0x00ff00ff) << 8) | ((x & 0xff00ff00) >> 8);
return (x << 16) | (x >> 16);
}
static inline uint64_t swap64(uint64_t x) {
x = ((x & 0x00ff00ff00ff00ff) << 8) | ((x & 0xff00ff00ff00ff00) >> 8);
x = ((x & 0x0000ffff0000ffff) << 16) | ((x & 0xffff0000ffff0000) >> 16);
return (x << 32) | (x >> 32);
}
#if defined(__GNUC__)
#define UNUSED __attribute__((unused))
#else
#define UNUSED
#endif
static inline void mem_inplace_ident(void *mem UNUSED, size_t n UNUSED) { }
#undef UNUSED
static inline void mem_inplace_swap32(void *mem, size_t n) {
size_t i;
for (i = 0; i < n; i++) {
((uint32_t *)mem)[i] = swap32(((const uint32_t *)mem)[i]);
}
}
static inline void mem_inplace_swap64(void *mem, size_t n) {
size_t i;
for (i = 0; i < n; i++) {
((uint64_t *)mem)[i] = swap64(((const uint64_t *)mem)[i]);
}
}
static inline void memcpy_ident32(void *dst, const void *src, size_t n) {
memcpy(dst, src, 4 * n);
}
static inline void memcpy_ident64(void *dst, const void *src, size_t n) {
memcpy(dst, src, 8 * n);
}
static inline void memcpy_swap32(void *dst, const void *src, size_t n) {
size_t i;
for (i = 0; i < n; i++) {
((uint32_t *)dst)[i] = swap32(((const uint32_t *)src)[i]);
}
}
static inline void memcpy_swap64(void *dst, const void *src, size_t n) {
size_t i;
for (i = 0; i < n; i++) {
((uint64_t *)dst)[i] = swap64(((const uint64_t *)src)[i]);
}
}
#define SWAP32LE IDENT32
#define SWAP32BE SWAP32
#define swap32le ident32
#define swap32be swap32
#define mem_inplace_swap32le mem_inplace_ident
#define mem_inplace_swap32be mem_inplace_swap32
#define memcpy_swap32le memcpy_ident32
#define memcpy_swap32be memcpy_swap32
#define SWAP64LE IDENT64
#define SWAP64BE SWAP64
#define swap64le ident64
#define swap64be swap64
#define mem_inplace_swap64le mem_inplace_ident
#define mem_inplace_swap64be mem_inplace_swap64
#define memcpy_swap64le memcpy_ident64
#define memcpy_swap64be memcpy_swap64

179
crypto/skein_port.h Normal file
View File

@ -0,0 +1,179 @@
#ifndef _SKEIN_PORT_H_
#define _SKEIN_PORT_H_
#include <limits.h>
#include <stdint.h>
#include <stddef.h>
#ifndef RETURN_VALUES
# define RETURN_VALUES
# if defined( DLL_EXPORT )
# if defined( _MSC_VER ) || defined ( __INTEL_COMPILER )
# define VOID_RETURN __declspec( dllexport ) void __stdcall
# define INT_RETURN __declspec( dllexport ) int __stdcall
# elif defined( __GNUC__ )
# define VOID_RETURN __declspec( __dllexport__ ) void
# define INT_RETURN __declspec( __dllexport__ ) int
# else
# error Use of the DLL is only available on the Microsoft, Intel and GCC compilers
# endif
# elif defined( DLL_IMPORT )
# if defined( _MSC_VER ) || defined ( __INTEL_COMPILER )
# define VOID_RETURN __declspec( dllimport ) void __stdcall
# define INT_RETURN __declspec( dllimport ) int __stdcall
# elif defined( __GNUC__ )
# define VOID_RETURN __declspec( __dllimport__ ) void
# define INT_RETURN __declspec( __dllimport__ ) int
# else
# error Use of the DLL is only available on the Microsoft, Intel and GCC compilers
# endif
# elif defined( __WATCOMC__ )
# define VOID_RETURN void __cdecl
# define INT_RETURN int __cdecl
# else
# define VOID_RETURN void
# define INT_RETURN int
# endif
#endif
/* These defines are used to declare buffers in a way that allows
faster operations on longer variables to be used. In all these
defines 'size' must be a power of 2 and >= 8
dec_unit_type(size,x) declares a variable 'x' of length
'size' bits
dec_bufr_type(size,bsize,x) declares a buffer 'x' of length 'bsize'
bytes defined as an array of variables
each of 'size' bits (bsize must be a
multiple of size / 8)
ptr_cast(x,size) casts a pointer to a pointer to a
varaiable of length 'size' bits
*/
#define ui_type(size) uint##size##_t
#define dec_unit_type(size,x) typedef ui_type(size) x
#define dec_bufr_type(size,bsize,x) typedef ui_type(size) x[bsize / (size >> 3)]
#define ptr_cast(x,size) ((ui_type(size)*)(x))
typedef unsigned int uint_t; /* native unsigned integer */
typedef uint8_t u08b_t; /* 8-bit unsigned integer */
typedef uint64_t u64b_t; /* 64-bit unsigned integer */
#ifndef RotL_64
#define RotL_64(x,N) (((x) << (N)) | ((x) >> (64-(N))))
#endif
/*
* Skein is "natively" little-endian (unlike SHA-xxx), for optimal
* performance on x86 CPUs. The Skein code requires the following
* definitions for dealing with endianness:
*
* SKEIN_NEED_SWAP: 0 for little-endian, 1 for big-endian
* Skein_Put64_LSB_First
* Skein_Get64_LSB_First
* Skein_Swap64
*
* If SKEIN_NEED_SWAP is defined at compile time, it is used here
* along with the portable versions of Put64/Get64/Swap64, which
* are slow in general.
*
* Otherwise, an "auto-detect" of endianness is attempted below.
* If the default handling doesn't work well, the user may insert
* platform-specific code instead (e.g., for big-endian CPUs).
*
*/
#ifndef SKEIN_NEED_SWAP /* compile-time "override" for endianness? */
#define PLATFORM_BYTE_ORDER IS_LITTLE_ENDIAN
/* special handler for IA64, which may be either endianness (?) */
/* here we assume little-endian, but this may need to be changed */
#if defined(__ia64) || defined(__ia64__) || defined(_M_IA64)
# define PLATFORM_MUST_ALIGN (1)
#ifndef PLATFORM_BYTE_ORDER
# define PLATFORM_BYTE_ORDER IS_LITTLE_ENDIAN
#endif
#endif
#ifndef PLATFORM_MUST_ALIGN
# define PLATFORM_MUST_ALIGN (0)
#endif
#if PLATFORM_BYTE_ORDER == IS_BIG_ENDIAN
/* here for big-endian CPUs */
#define SKEIN_NEED_SWAP (1)
#elif PLATFORM_BYTE_ORDER == IS_LITTLE_ENDIAN
/* here for x86 and x86-64 CPUs (and other detected little-endian CPUs) */
#define SKEIN_NEED_SWAP (0)
#if PLATFORM_MUST_ALIGN == 0 /* ok to use "fast" versions? */
#define Skein_Put64_LSB_First(dst08,src64,bCnt) memcpy(dst08,src64,bCnt)
#define Skein_Get64_LSB_First(dst64,src08,wCnt) memcpy(dst64,src08,8*(wCnt))
#endif
#else
#error "Skein needs endianness setting!"
#endif
#endif /* ifndef SKEIN_NEED_SWAP */
/*
******************************************************************
* Provide any definitions still needed.
******************************************************************
*/
#ifndef Skein_Swap64 /* swap for big-endian, nop for little-endian */
#if SKEIN_NEED_SWAP
#define Skein_Swap64(w64) \
( (( ((u64b_t)(w64)) & 0xFF) << 56) | \
(((((u64b_t)(w64)) >> 8) & 0xFF) << 48) | \
(((((u64b_t)(w64)) >>16) & 0xFF) << 40) | \
(((((u64b_t)(w64)) >>24) & 0xFF) << 32) | \
(((((u64b_t)(w64)) >>32) & 0xFF) << 24) | \
(((((u64b_t)(w64)) >>40) & 0xFF) << 16) | \
(((((u64b_t)(w64)) >>48) & 0xFF) << 8) | \
(((((u64b_t)(w64)) >>56) & 0xFF) ) )
#else
#define Skein_Swap64(w64) (w64)
#endif
#endif /* ifndef Skein_Swap64 */
#ifndef Skein_Put64_LSB_First
void Skein_Put64_LSB_First(u08b_t *dst,const u64b_t *src,size_t bCnt)
#ifdef SKEIN_PORT_CODE /* instantiate the function code here? */
{ /* this version is fully portable (big-endian or little-endian), but slow */
size_t n;
for (n=0;n<bCnt;n++)
dst[n] = (u08b_t) (src[n>>3] >> (8*(n&7)));
}
#else
; /* output only the function prototype */
#endif
#endif /* ifndef Skein_Put64_LSB_First */
#ifndef Skein_Get64_LSB_First
void Skein_Get64_LSB_First(u64b_t *dst,const u08b_t *src,size_t wCnt)
#ifdef SKEIN_PORT_CODE /* instantiate the function code here? */
{ /* this version is fully portable (big-endian or little-endian), but slow */
size_t n;
for (n=0;n<8*wCnt;n+=8)
dst[n/8] = (((u64b_t) src[n ]) ) +
(((u64b_t) src[n+1]) << 8) +
(((u64b_t) src[n+2]) << 16) +
(((u64b_t) src[n+3]) << 24) +
(((u64b_t) src[n+4]) << 32) +
(((u64b_t) src[n+5]) << 40) +
(((u64b_t) src[n+6]) << 48) +
(((u64b_t) src[n+7]) << 56) ;
}
#else
; /* output only the function prototype */
#endif
#endif /* ifndef Skein_Get64_LSB_First */
#endif /* ifndef _SKEIN_PORT_H_ */

197
crypto/soft_aes.c Normal file
View File

@ -0,0 +1,197 @@
/*
* 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
* 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 <http://www.gnu.org/licenses/>.
*/
/*
* The orginal author of this AES implementation is Karl Malbrain.
*/
#ifdef __GNUC__
#include <x86intrin.h>
#else
#include <intrin.h>
#endif // __GNUC__
#include <inttypes.h>
#define TABLE_ALIGN 32
#define WPOLY 0x011b
#define N_COLS 4
#define AES_BLOCK_SIZE 16
#define RC_LENGTH (5 * (AES_BLOCK_SIZE / 4 - 2))
#if defined(_MSC_VER)
#define ALIGN __declspec(align(TABLE_ALIGN))
#elif defined(__GNUC__)
#define ALIGN __attribute__ ((aligned(16)))
#else
#define ALIGN
#endif
#define rf1(r,c) (r)
#define word_in(x,c) (*((uint32_t*)(x)+(c)))
#define word_out(x,c,v) (*((uint32_t*)(x)+(c)) = (v))
#define s(x,c) x[c]
#define si(y,x,c) (s(y,c) = word_in(x, c))
#define so(y,x,c) word_out(y, c, s(x,c))
#define state_in(y,x) si(y,x,0); si(y,x,1); si(y,x,2); si(y,x,3)
#define state_out(y,x) so(y,x,0); so(y,x,1); so(y,x,2); so(y,x,3)
#define round(y,x,k) \
y[0] = (k)[0] ^ (t_fn[0][x[0] & 0xff] ^ t_fn[1][(x[1] >> 8) & 0xff] ^ t_fn[2][(x[2] >> 16) & 0xff] ^ t_fn[3][x[3] >> 24]); \
y[1] = (k)[1] ^ (t_fn[0][x[1] & 0xff] ^ t_fn[1][(x[2] >> 8) & 0xff] ^ t_fn[2][(x[3] >> 16) & 0xff] ^ t_fn[3][x[0] >> 24]); \
y[2] = (k)[2] ^ (t_fn[0][x[2] & 0xff] ^ t_fn[1][(x[3] >> 8) & 0xff] ^ t_fn[2][(x[0] >> 16) & 0xff] ^ t_fn[3][x[1] >> 24]); \
y[3] = (k)[3] ^ (t_fn[0][x[3] & 0xff] ^ t_fn[1][(x[0] >> 8) & 0xff] ^ t_fn[2][(x[1] >> 16) & 0xff] ^ t_fn[3][x[2] >> 24]);
#define to_byte(x) ((x) & 0xff)
#define bval(x,n) to_byte((x) >> (8 * (n)))
#define fwd_var(x,r,c)\
( r == 0 ? ( c == 0 ? s(x,0) : c == 1 ? s(x,1) : c == 2 ? s(x,2) : s(x,3))\
: r == 1 ? ( c == 0 ? s(x,1) : c == 1 ? s(x,2) : c == 2 ? s(x,3) : s(x,0))\
: r == 2 ? ( c == 0 ? s(x,2) : c == 1 ? s(x,3) : c == 2 ? s(x,0) : s(x,1))\
: ( c == 0 ? s(x,3) : c == 1 ? s(x,0) : c == 2 ? s(x,1) : s(x,2)))
#define fwd_rnd(y,x,k,c) (s(y,c) = (k)[c] ^ four_tables(x,t_use(f,n),fwd_var,rf1,c))
#define sb_data(w) {\
w(0x63), w(0x7c), w(0x77), w(0x7b), w(0xf2), w(0x6b), w(0x6f), w(0xc5),\
w(0x30), w(0x01), w(0x67), w(0x2b), w(0xfe), w(0xd7), w(0xab), w(0x76),\
w(0xca), w(0x82), w(0xc9), w(0x7d), w(0xfa), w(0x59), w(0x47), w(0xf0),\
w(0xad), w(0xd4), w(0xa2), w(0xaf), w(0x9c), w(0xa4), w(0x72), w(0xc0),\
w(0xb7), w(0xfd), w(0x93), w(0x26), w(0x36), w(0x3f), w(0xf7), w(0xcc),\
w(0x34), w(0xa5), w(0xe5), w(0xf1), w(0x71), w(0xd8), w(0x31), w(0x15),\
w(0x04), w(0xc7), w(0x23), w(0xc3), w(0x18), w(0x96), w(0x05), w(0x9a),\
w(0x07), w(0x12), w(0x80), w(0xe2), w(0xeb), w(0x27), w(0xb2), w(0x75),\
w(0x09), w(0x83), w(0x2c), w(0x1a), w(0x1b), w(0x6e), w(0x5a), w(0xa0),\
w(0x52), w(0x3b), w(0xd6), w(0xb3), w(0x29), w(0xe3), w(0x2f), w(0x84),\
w(0x53), w(0xd1), w(0x00), w(0xed), w(0x20), w(0xfc), w(0xb1), w(0x5b),\
w(0x6a), w(0xcb), w(0xbe), w(0x39), w(0x4a), w(0x4c), w(0x58), w(0xcf),\
w(0xd0), w(0xef), w(0xaa), w(0xfb), w(0x43), w(0x4d), w(0x33), w(0x85),\
w(0x45), w(0xf9), w(0x02), w(0x7f), w(0x50), w(0x3c), w(0x9f), w(0xa8),\
w(0x51), w(0xa3), w(0x40), w(0x8f), w(0x92), w(0x9d), w(0x38), w(0xf5),\
w(0xbc), w(0xb6), w(0xda), w(0x21), w(0x10), w(0xff), w(0xf3), w(0xd2),\
w(0xcd), w(0x0c), w(0x13), w(0xec), w(0x5f), w(0x97), w(0x44), w(0x17),\
w(0xc4), w(0xa7), w(0x7e), w(0x3d), w(0x64), w(0x5d), w(0x19), w(0x73),\
w(0x60), w(0x81), w(0x4f), w(0xdc), w(0x22), w(0x2a), w(0x90), w(0x88),\
w(0x46), w(0xee), w(0xb8), w(0x14), w(0xde), w(0x5e), w(0x0b), w(0xdb),\
w(0xe0), w(0x32), w(0x3a), w(0x0a), w(0x49), w(0x06), w(0x24), w(0x5c),\
w(0xc2), w(0xd3), w(0xac), w(0x62), w(0x91), w(0x95), w(0xe4), w(0x79),\
w(0xe7), w(0xc8), w(0x37), w(0x6d), w(0x8d), w(0xd5), w(0x4e), w(0xa9),\
w(0x6c), w(0x56), w(0xf4), w(0xea), w(0x65), w(0x7a), w(0xae), w(0x08),\
w(0xba), w(0x78), w(0x25), w(0x2e), w(0x1c), w(0xa6), w(0xb4), w(0xc6),\
w(0xe8), w(0xdd), w(0x74), w(0x1f), w(0x4b), w(0xbd), w(0x8b), w(0x8a),\
w(0x70), w(0x3e), w(0xb5), w(0x66), w(0x48), w(0x03), w(0xf6), w(0x0e),\
w(0x61), w(0x35), w(0x57), w(0xb9), w(0x86), w(0xc1), w(0x1d), w(0x9e),\
w(0xe1), w(0xf8), w(0x98), w(0x11), w(0x69), w(0xd9), w(0x8e), w(0x94),\
w(0x9b), w(0x1e), w(0x87), w(0xe9), w(0xce), w(0x55), w(0x28), w(0xdf),\
w(0x8c), w(0xa1), w(0x89), w(0x0d), w(0xbf), w(0xe6), w(0x42), w(0x68),\
w(0x41), w(0x99), w(0x2d), w(0x0f), w(0xb0), w(0x54), w(0xbb), w(0x16) }
#define rc_data(w) {\
w(0x01), w(0x02), w(0x04), w(0x08), w(0x10),w(0x20), w(0x40), w(0x80),\
w(0x1b), w(0x36) }
#define bytes2word(b0, b1, b2, b3) (((uint32_t)(b3) << 24) | \
((uint32_t)(b2) << 16) | ((uint32_t)(b1) << 8) | (b0))
#define h0(x) (x)
#define w0(p) bytes2word(p, 0, 0, 0)
#define w1(p) bytes2word(0, p, 0, 0)
#define w2(p) bytes2word(0, 0, p, 0)
#define w3(p) bytes2word(0, 0, 0, p)
#define u0(p) bytes2word(f2(p), p, p, f3(p))
#define u1(p) bytes2word(f3(p), f2(p), p, p)
#define u2(p) bytes2word(p, f3(p), f2(p), p)
#define u3(p) bytes2word(p, p, f3(p), f2(p))
#define v0(p) bytes2word(fe(p), f9(p), fd(p), fb(p))
#define v1(p) bytes2word(fb(p), fe(p), f9(p), fd(p))
#define v2(p) bytes2word(fd(p), fb(p), fe(p), f9(p))
#define v3(p) bytes2word(f9(p), fd(p), fb(p), fe(p))
#define f2(x) ((x<<1) ^ (((x>>7) & 1) * WPOLY))
#define f4(x) ((x<<2) ^ (((x>>6) & 1) * WPOLY) ^ (((x>>6) & 2) * WPOLY))
#define f8(x) ((x<<3) ^ (((x>>5) & 1) * WPOLY) ^ (((x>>5) & 2) * WPOLY) ^ (((x>>5) & 4) * WPOLY))
#define f3(x) (f2(x) ^ x)
#define f9(x) (f8(x) ^ x)
#define fb(x) (f8(x) ^ f2(x) ^ x)
#define fd(x) (f8(x) ^ f4(x) ^ x)
#define fe(x) (f8(x) ^ f4(x) ^ f2(x))
#define t_dec(m,n) t_##m##n
#define t_set(m,n) t_##m##n
#define t_use(m,n) t_##m##n
#define d_4(t,n,b,e,f,g,h) ALIGN const t n[4][256] = { b(e), b(f), b(g), b(h) }
#define four_tables(x,tab,vf,rf,c) \
(tab[0][bval(vf(x,0,c),rf(0,c))] \
^ tab[1][bval(vf(x,1,c),rf(1,c))] \
^ tab[2][bval(vf(x,2,c),rf(2,c))] \
^ tab[3][bval(vf(x,3,c),rf(3,c))])
d_4(uint32_t, t_dec(f,n), sb_data, u0, u1, u2, u3);
__m128i soft_aesenc(__m128i in, __m128i key)
{
uint32_t x0, x1, x2, x3;
x0 = _mm_cvtsi128_si32(in);
x1 = _mm_cvtsi128_si32(_mm_shuffle_epi32(in, 0x55));
x2 = _mm_cvtsi128_si32(_mm_shuffle_epi32(in, 0xAA));
x3 = _mm_cvtsi128_si32(_mm_shuffle_epi32(in, 0xFF));
__m128i out = _mm_set_epi32(
(t_fn[0][x3 & 0xff] ^ t_fn[1][(x0 >> 8) & 0xff] ^ t_fn[2][(x1 >> 16) & 0xff] ^ t_fn[3][x2 >> 24]),
(t_fn[0][x2 & 0xff] ^ t_fn[1][(x3 >> 8) & 0xff] ^ t_fn[2][(x0 >> 16) & 0xff] ^ t_fn[3][x1 >> 24]),
(t_fn[0][x1 & 0xff] ^ t_fn[1][(x2 >> 8) & 0xff] ^ t_fn[2][(x3 >> 16) & 0xff] ^ t_fn[3][x0 >> 24]),
(t_fn[0][x0 & 0xff] ^ t_fn[1][(x1 >> 8) & 0xff] ^ t_fn[2][(x2 >> 16) & 0xff] ^ t_fn[3][x3 >> 24]));
return _mm_xor_si128(out, key);
}
uint8_t Sbox[256] = { // forward s-box
0x63, 0x7c, 0x77, 0x7b, 0xf2, 0x6b, 0x6f, 0xc5, 0x30, 0x01, 0x67, 0x2b, 0xfe, 0xd7, 0xab, 0x76,
0xca, 0x82, 0xc9, 0x7d, 0xfa, 0x59, 0x47, 0xf0, 0xad, 0xd4, 0xa2, 0xaf, 0x9c, 0xa4, 0x72, 0xc0,
0xb7, 0xfd, 0x93, 0x26, 0x36, 0x3f, 0xf7, 0xcc, 0x34, 0xa5, 0xe5, 0xf1, 0x71, 0xd8, 0x31, 0x15,
0x04, 0xc7, 0x23, 0xc3, 0x18, 0x96, 0x05, 0x9a, 0x07, 0x12, 0x80, 0xe2, 0xeb, 0x27, 0xb2, 0x75,
0x09, 0x83, 0x2c, 0x1a, 0x1b, 0x6e, 0x5a, 0xa0, 0x52, 0x3b, 0xd6, 0xb3, 0x29, 0xe3, 0x2f, 0x84,
0x53, 0xd1, 0x00, 0xed, 0x20, 0xfc, 0xb1, 0x5b, 0x6a, 0xcb, 0xbe, 0x39, 0x4a, 0x4c, 0x58, 0xcf,
0xd0, 0xef, 0xaa, 0xfb, 0x43, 0x4d, 0x33, 0x85, 0x45, 0xf9, 0x02, 0x7f, 0x50, 0x3c, 0x9f, 0xa8,
0x51, 0xa3, 0x40, 0x8f, 0x92, 0x9d, 0x38, 0xf5, 0xbc, 0xb6, 0xda, 0x21, 0x10, 0xff, 0xf3, 0xd2,
0xcd, 0x0c, 0x13, 0xec, 0x5f, 0x97, 0x44, 0x17, 0xc4, 0xa7, 0x7e, 0x3d, 0x64, 0x5d, 0x19, 0x73,
0x60, 0x81, 0x4f, 0xdc, 0x22, 0x2a, 0x90, 0x88, 0x46, 0xee, 0xb8, 0x14, 0xde, 0x5e, 0x0b, 0xdb,
0xe0, 0x32, 0x3a, 0x0a, 0x49, 0x06, 0x24, 0x5c, 0xc2, 0xd3, 0xac, 0x62, 0x91, 0x95, 0xe4, 0x79,
0xe7, 0xc8, 0x37, 0x6d, 0x8d, 0xd5, 0x4e, 0xa9, 0x6c, 0x56, 0xf4, 0xea, 0x65, 0x7a, 0xae, 0x08,
0xba, 0x78, 0x25, 0x2e, 0x1c, 0xa6, 0xb4, 0xc6, 0xe8, 0xdd, 0x74, 0x1f, 0x4b, 0xbd, 0x8b, 0x8a,
0x70, 0x3e, 0xb5, 0x66, 0x48, 0x03, 0xf6, 0x0e, 0x61, 0x35, 0x57, 0xb9, 0x86, 0xc1, 0x1d, 0x9e,
0xe1, 0xf8, 0x98, 0x11, 0x69, 0xd9, 0x8e, 0x94, 0x9b, 0x1e, 0x87, 0xe9, 0xce, 0x55, 0x28, 0xdf,
0x8c, 0xa1, 0x89, 0x0d, 0xbf, 0xe6, 0x42, 0x68, 0x41, 0x99, 0x2d, 0x0f, 0xb0, 0x54, 0xbb, 0x16};
static inline void sub_word(uint8_t* key)
{
key[0] = Sbox[key[0]];
key[1] = Sbox[key[1]];
key[2] = Sbox[key[2]];
key[3] = Sbox[key[3]];
}
__m128i soft_aeskeygenassist(__m128i key, uint8_t rcon)
{
uint32_t X1 = _mm_cvtsi128_si32(_mm_shuffle_epi32(key, 0x55));
uint32_t X3 = _mm_cvtsi128_si32(_mm_shuffle_epi32(key, 0xFF));
sub_word((uint8_t*)&X1);
sub_word((uint8_t*)&X3);
return _mm_set_epi32(_rotr(X3, 8) ^ rcon, X3,_rotr(X1, 8) ^ rcon, X1);
}

11
donate-level.h Normal file
View File

@ -0,0 +1,11 @@
#pragma once
/*
* Dev donation.
* Percentage of your hashing power that you want to donate to the developer, can be 0.0 if you don't want to do that.
* Example of how it works for the default setting of 1.0:
* You miner will mine into your usual pool for 99 minutes, then switch to the developer's pool for 1.0 minute.
* Switching is instant, and only happens after a successful connection, so you never loose any hashes.
*/
constexpr double fDevDonationLevel = 1.0 / 100.0;

721
executor.cpp Normal file
View File

@ -0,0 +1,721 @@
/*
* 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
* 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 <http://www.gnu.org/licenses/>.
*/
#include <thread>
#include <string>
#include <cmath>
#include <algorithm>
#include <assert.h>
#include <time.h>
#include "executor.h"
#include "jpsock.h"
#include "minethd.h"
#include "jconf.h"
#include "console.h"
#include "donate-level.h"
#ifdef __GNUC__
#include <mm_malloc.h>
#else
#include <malloc.h>
#endif // __GNUC__
#ifdef _WIN32
#define strncasecmp _strnicmp
#endif // _WIN32
executor* executor::oInst = NULL;
executor::executor()
{
my_thd = nullptr;
cpu_ctx = (cryptonight_ctx*)_mm_malloc(sizeof(cryptonight_ctx), 4096);
}
void executor::push_timed_event(ex_event&& ev, size_t sec)
{
std::unique_lock<std::mutex> lck(timed_event_mutex);
lTimedEvents.emplace_back(std::move(ev), sec_to_ticks(sec));
}
void executor::ex_clock_thd()
{
size_t iSwitchPeriod = sec_to_ticks(iDevDonatePeriod);
size_t iDevPortion = (size_t)floor(((double)iSwitchPeriod) * fDevDonationLevel);
//No point in bothering with less than 10 sec
if(iDevPortion < sec_to_ticks(10))
iDevPortion = 0;
//Add 2 seconds to compensate for connect
if(iDevPortion != 0)
iDevPortion += sec_to_ticks(2);
while (true)
{
std::this_thread::sleep_for(std::chrono::milliseconds(size_t(iTickTime)));
push_event(ex_event(EV_PERF_TICK));
// Service timed events
std::unique_lock<std::mutex> lck(timed_event_mutex);
std::list<timed_event>::iterator ev = lTimedEvents.begin();
while (ev != lTimedEvents.end())
{
ev->ticks_left--;
if(ev->ticks_left == 0)
{
push_event(std::move(ev->event));
ev = lTimedEvents.erase(ev);
}
else
ev++;
}
lck.unlock();
if(iDevPortion == 0)
continue;
iSwitchPeriod--;
if(iSwitchPeriod == 0)
{
push_event(ex_event(EV_SWITCH_POOL, usr_pool_id));
iSwitchPeriod = sec_to_ticks(iDevDonatePeriod);
}
else if(iSwitchPeriod == iDevPortion)
{
push_event(ex_event(EV_SWITCH_POOL, dev_pool_id));
}
}
}
void executor::sched_reconnect()
{
long long unsigned int rt = jconf::inst()->GetNetRetry();
printer::inst()->print_msg(L1, "Pool connection lost. Waiting %lld s before retry.", rt);
auto work = minethd::miner_work();
minethd::switch_work(work);
push_timed_event(ex_event(EV_RECONNECT, usr_pool_id), rt);
}
void executor::log_socket_error(std::string&& sError)
{
vSocketLog.emplace_back(std::move(sError));
printer::inst()->print_msg(L1, "SOCKET ERROR - %s", vSocketLog.back().msg.c_str());
}
void executor::log_result_error(std::string&& sError)
{
size_t i = 1, ln = vMineResults.size();
for(; i < ln; i++)
{
if(vMineResults[i].compare(sError))
{
vMineResults[i].increment();
break;
}
}
if(i == ln) //Not found
vMineResults.emplace_back(std::move(sError));
else
sError.clear();
}
void executor::log_result_ok(uint64_t iActualDiff)
{
iPoolHashes += iPoolDiff;
size_t ln = iTopDiff.size() - 1;
if(iActualDiff > iTopDiff[ln])
{
iTopDiff[ln] = iActualDiff;
std::sort(iTopDiff.rbegin(), iTopDiff.rend());
}
vMineResults[0].increment();
}
jpsock* executor::pick_pool_by_id(size_t pool_id)
{
assert(pool_id != invalid_pool_id);
if(pool_id == dev_pool_id)
return dev_pool;
else
return usr_pool;
}
void executor::on_sock_ready(size_t pool_id)
{
jpsock* pool = pick_pool_by_id(pool_id);
if(pool_id == dev_pool_id)
{
if(!pool->cmd_login("", ""))
pool->disconnect();
current_pool_id = dev_pool_id;
printer::inst()->print_msg(L1, "Dev pool logged in. Switching work.");
return;
}
printer::inst()->print_msg(L1, "Connected. Logging in...");
if (!pool->cmd_login(jconf::inst()->GetWalletAddress(), jconf::inst()->GetPoolPwd()))
{
if(!pool->have_sock_error())
{
log_socket_error(pool->get_call_error());
pool->disconnect();
}
}
else
reset_stats();
}
void executor::on_sock_error(size_t pool_id, std::string&& sError)
{
jpsock* pool = pick_pool_by_id(pool_id);
if(pool_id == dev_pool_id)
{
pool->disconnect();
if(current_pool_id != dev_pool_id)
return;
printer::inst()->print_msg(L1, "Dev pool connection error. Switching work.");
on_switch_pool(usr_pool_id);
return;
}
log_socket_error(std::move(sError));
pool->disconnect();
sched_reconnect();
}
void executor::on_pool_have_job(size_t pool_id, pool_job& oPoolJob)
{
if(pool_id != current_pool_id)
return;
jpsock* pool = pick_pool_by_id(pool_id);
minethd::miner_work oWork(oPoolJob.sJobID, oPoolJob.bWorkBlob,
oPoolJob.iWorkLen, oPoolJob.iResumeCnt, oPoolJob.iTarget, pool_id);
minethd::switch_work(oWork);
if(pool_id == dev_pool_id)
return;
if(iPoolDiff != pool->get_current_diff())
{
iPoolDiff = pool->get_current_diff();
printer::inst()->print_msg(L2, "Difficulty changed. Now: %llu.", int_port(iPoolDiff));
}
printer::inst()->print_msg(L3, "New block detected.");
}
void executor::on_miner_result(size_t pool_id, job_result& oResult)
{
jpsock* pool = pick_pool_by_id(pool_id);
*(uint32_t*)(oResult.bWorkBlob + 39) = oResult.iNonce;
if(jconf::inst()->HaveHardwareAes())
cryptonight_hash_ctx(oResult.bWorkBlob, oResult.iWorkLen, oResult.bResult, cpu_ctx);
else
cryptonight_hash_ctx_soft(oResult.bWorkBlob, oResult.iWorkLen, oResult.bResult, cpu_ctx);
bool bVerified = ((uint32_t*)oResult.bResult)[7] < oResult.iTarget;
if(pool_id == dev_pool_id)
{
//Ignore errors silently
if(pool->is_running() && pool->is_logged_in() && bVerified)
pool->cmd_submit(oResult.sJobID, oResult.iNonce, oResult.bResult);
return;
}
if (!bVerified)
{
log_result_error("[GPU COMPUTE ERROR]");
return;
}
if (!pool->is_running() || !pool->is_logged_in())
{
log_result_error("[NETWORK ERROR]");
return;
}
using namespace std::chrono;
size_t t_start = time_point_cast<milliseconds>(high_resolution_clock::now()).time_since_epoch().count();
bool bResult = pool->cmd_submit(oResult.sJobID, oResult.iNonce, oResult.bResult);
size_t t_len = time_point_cast<milliseconds>(high_resolution_clock::now()).time_since_epoch().count() - t_start;
if(t_len > 0xFFFF)
t_len = 0xFFFF;
iPoolCallTimes.push_back((uint16_t)t_len);
if(bResult)
{
uint64_t* targets = (uint64_t*)oResult.bResult;
log_result_ok(jpsock::t64_to_diff(targets[3]));
printer::inst()->print_msg(L3, "Result accepted by the pool.");
}
else
{
if(!pool->have_sock_error())
{
printer::inst()->print_msg(L3, "Result rejected by the pool.");
std::string error = pool->get_call_error();
if(strncasecmp(error.c_str(), "Unauthenticated", 15) == 0)
{
printer::inst()->print_msg(L2, "Your miner was unable to find a share in time. Either the pool difficulty is too high, or the pool timeout is too low.");
pool->disconnect();
}
log_result_error(std::move(error));
}
else
log_result_error("[NETWORK ERROR]");
}
}
void executor::on_reconnect(size_t pool_id)
{
jpsock* pool = pick_pool_by_id(pool_id);
std::string error;
if(pool_id == dev_pool_id)
return;
printer::inst()->print_msg(L1, "Connecting to pool %s ...", jconf::inst()->GetPoolAddress());
if(!pool->connect(jconf::inst()->GetPoolAddress(), error))
{
log_socket_error(std::move(error));
sched_reconnect();
}
}
void executor::on_switch_pool(size_t pool_id)
{
if(pool_id == current_pool_id)
return;
jpsock* pool = pick_pool_by_id(pool_id);
if(pool_id == dev_pool_id)
{
std::string error;
// If it fails, it fails, we carry on on the usr pool
// as we never receive further events
printer::inst()->print_msg(L1, "Connecting to dev pool...");
if(!pool->connect("donate.xmr-stak.net:3333", error))
printer::inst()->print_msg(L1, "Error connecting to dev pool. Staying with user pool.");
}
else
{
printer::inst()->print_msg(L1, "Switching back to user pool.");
current_pool_id = pool_id;
pool_job oPoolJob;
if(!pool->get_current_job(oPoolJob))
{
pool->disconnect();
return;
}
minethd::miner_work oWork(oPoolJob.sJobID, oPoolJob.bWorkBlob,
oPoolJob.iWorkLen, oPoolJob.iResumeCnt, oPoolJob.iTarget, pool_id);
minethd::switch_work(oWork);
if(dev_pool->is_running())
push_timed_event(ex_event(EV_DEV_POOL_EXIT), 5);
}
}
void executor::ex_main()
{
assert(1000 % iTickTime == 0);
minethd::miner_work oWork = minethd::miner_work();
pvThreads = minethd::thread_starter(oWork);
telem = new telemetry(pvThreads->size());
current_pool_id = usr_pool_id;
usr_pool = new jpsock(usr_pool_id);
dev_pool = new jpsock(dev_pool_id);
ex_event ev;
std::thread clock_thd(&executor::ex_clock_thd, this);
//This will connect us to the pool for the first time
push_event(ex_event(EV_RECONNECT, usr_pool_id));
// Place the default success result at postion 0, it needs to
// be here even if our first result is a failure
vMineResults.emplace_back();
// If the user requested it, start the autohash printer
if(jconf::inst()->GetVerboseLevel() >= 4)
push_timed_event(ex_event(EV_HASHRATE_LOOP), jconf::inst()->GetAutohashTime());
size_t cnt = 0, i;
while (true)
{
ev = oEventQ.pop();
switch (ev.iName)
{
case EV_SOCK_READY:
on_sock_ready(ev.iPoolId);
break;
case EV_SOCK_ERROR:
on_sock_error(ev.iPoolId, std::move(ev.sSocketError));
break;
case EV_POOL_HAVE_JOB:
on_pool_have_job(ev.iPoolId, ev.oPoolJob);
break;
case EV_MINER_HAVE_RESULT:
on_miner_result(ev.iPoolId, ev.oJobResult);
break;
case EV_RECONNECT:
on_reconnect(ev.iPoolId);
break;
case EV_SWITCH_POOL:
on_switch_pool(ev.iPoolId);
break;
case EV_DEV_POOL_EXIT:
dev_pool->disconnect();
break;
case EV_PERF_TICK:
for (i = 0; i < pvThreads->size(); i++)
telem->push_perf_value(i, pvThreads->at(i)->iHashCount.load(std::memory_order_relaxed),
pvThreads->at(i)->iTimestamp.load(std::memory_order_relaxed));
if((cnt++ & 0xF) == 0) //Every 16 ticks
{
double fHps = 0.0;
for (i = 0; i < pvThreads->size(); i++)
fHps += telem->calc_telemetry_data(10000, i);
if(fHighestHps < fHps)
fHighestHps = fHps;
}
break;
case EV_USR_HASHRATE:
case EV_USR_RESULTS:
case EV_USR_CONNSTAT:
print_report(ev.iName);
break;
case EV_HTML_HASHRATE:
case EV_HTML_RESULTS:
case EV_HTML_CONNSTAT:
http_report(ev.iName);
break;
case EV_HASHRATE_LOOP:
print_report(EV_USR_HASHRATE);
push_timed_event(ex_event(EV_HASHRATE_LOOP), jconf::inst()->GetAutohashTime());
break;
case EV_INVALID_VAL:
default:
assert(false);
break;
}
}
}
inline const char* hps_format(double h, char* buf, size_t l)
{
if(std::isnormal(h))
{
snprintf(buf, l, " %03.1f", h);
return buf;
}
else if(h == 0.0) //Zero is not normal but we want it
return " 0.0";
else
return " (na)";
}
void executor::hashrate_report(std::string& out)
{
char num[32];
size_t nthd = pvThreads->size();
out.reserve(256 + nthd * 64);
double fTotal[3] = { 0.0, 0.0, 0.0};
size_t i;
out.append("HASHRATE REPORT\n");
out.append("| ID | 10s | 60s | 15m |");
if(nthd != 1)
out.append(" ID | 10s | 60s | 15m |\n");
else
out.append(1, '\n');
for (i = 0; i < nthd; i++)
{
double fHps[3];
fHps[0] = telem->calc_telemetry_data(10000, i);
fHps[1] = telem->calc_telemetry_data(60000, i);
fHps[2] = telem->calc_telemetry_data(900000, i);
snprintf(num, sizeof(num), "| %2u |", (unsigned int)i);
out.append(num);
out.append(hps_format(fHps[0], num, sizeof(num))).append(" |");
out.append(hps_format(fHps[1], num, sizeof(num))).append(" |");
out.append(hps_format(fHps[2], num, sizeof(num))).append(1, ' ');
fTotal[0] += fHps[0];
fTotal[1] += fHps[1];
fTotal[2] += fHps[2];
if((i & 0x1) == 1) //Odd i's
out.append("|\n");
}
if((i & 0x1) == 1) //We had odd number of threads
out.append("|\n");
if(nthd != 1)
out.append("-----------------------------------------------------\n");
else
out.append("---------------------------\n");
out.append("Totals: ");
out.append(hps_format(fTotal[0], num, sizeof(num)));
out.append(hps_format(fTotal[1], num, sizeof(num)));
out.append(hps_format(fTotal[2], num, sizeof(num)));
out.append(" H/s\nHighest: ");
out.append(hps_format(fHighestHps, num, sizeof(num)));
out.append(" H/s\n");
}
char* time_format(char* buf, size_t len, std::chrono::system_clock::time_point time)
{
time_t ctime = std::chrono::system_clock::to_time_t(time);
tm stime;
/*
* Oh for god's sake... this feels like we are back to the 90's...
* and don't get me started on lack strcpy_s because NIH - use non-standard strlcpy...
* And of course C++ implements unsafe version because... reasons
*/
#ifdef _WIN32
localtime_s(&stime, &ctime);
#else
localtime_r(&ctime, &stime);
#endif // __WIN32
strftime(buf, len, "%F %T", &stime);
return buf;
}
void executor::result_report(std::string& out)
{
char num[128];
char date[32];
out.reserve(1024);
size_t iGoodRes = vMineResults[0].count, iTotalRes = iGoodRes;
size_t ln = vMineResults.size();
for(size_t i=1; i < ln; i++)
iTotalRes += vMineResults[i].count;
out.append("RESULT REPORT\n");
if(iTotalRes == 0)
{
out.append("You haven't found any results yet.\n");
return;
}
double dConnSec;
{
using namespace std::chrono;
dConnSec = (double)duration_cast<seconds>(system_clock::now() - tPoolConnTime).count();
}
snprintf(num, sizeof(num), " (%.1f %%)\n", 100.0 * iGoodRes / iTotalRes);
out.append("Difficulty : ").append(std::to_string(iPoolDiff)).append(1, '\n');
out.append("Good results : ").append(std::to_string(iGoodRes)).append(" / ").
append(std::to_string(iTotalRes)).append(num);
if(iPoolCallTimes.size() != 0)
{
// Here we use iPoolCallTimes since it also gets reset when we disconnect
snprintf(num, sizeof(num), "%.1f sec\n", dConnSec / iPoolCallTimes.size());
out.append("Avg result time : ").append(num);
}
out.append("Pool-side hashes : ").append(std::to_string(iPoolHashes)).append(2, '\n');
out.append("Top 10 best results found:\n");
for(size_t i=0; i < 10; i += 2)
{
snprintf(num, sizeof(num), "| %2llu | %16llu | %2llu | %16llu |\n",
int_port(i), int_port(iTopDiff[i]), int_port(i+1), int_port(iTopDiff[i+1]));
out.append(num);
}
out.append("\nError details:\n");
if(ln > 1)
{
out.append("| Count | Error text | Last seen |\n");
for(size_t i=1; i < ln; i++)
{
snprintf(num, sizeof(num), "| %5llu | %-32.32s | %s |\n", int_port(vMineResults[i].count),
vMineResults[i].msg.c_str(), time_format(date, sizeof(date), vMineResults[i].time));
out.append(num);
}
}
else
out.append("Yay! No errors.\n");
}
void executor::connection_report(std::string& out)
{
char num[128];
char date[32];
out.reserve(512);
jpsock* pool = pick_pool_by_id(dev_pool_id + 1);
out.append("CONNECTION REPORT\n");
if (pool->is_running() && pool->is_logged_in())
out.append("Connected since : ").append(time_format(date, sizeof(date), tPoolConnTime)).append(1, '\n');
else
out.append("Connected since : <not connected>\n");
size_t n_calls = iPoolCallTimes.size();
if (n_calls > 1)
{
//Not-really-but-good-enough median
std::nth_element(iPoolCallTimes.begin(), iPoolCallTimes.begin() + n_calls/2, iPoolCallTimes.end());
out.append("Pool ping time : ").append(std::to_string(iPoolCallTimes[n_calls/2])).append(" ms\n");
}
else
out.append("Pool ping time : (n/a)\n");
out.append("\nNetwork error log:\n");
size_t ln = vSocketLog.size();
if(ln > 0)
{
out.append("| Date | Error text |\n");
for(size_t i=0; i < ln; i++)
{
snprintf(num, sizeof(num), "| %s | %-54.54s |\n",
time_format(date, sizeof(date), vSocketLog[i].time), vSocketLog[i].msg.c_str());
out.append(num);
}
}
else
out.append("Yay! No errors.\n");
}
void executor::print_report(ex_event_name ev)
{
std::string out;
switch(ev)
{
case EV_USR_HASHRATE:
hashrate_report(out);
break;
case EV_USR_RESULTS:
result_report(out);
break;
case EV_USR_CONNSTAT:
connection_report(out);
break;
default:
assert(false);
break;
}
printer::inst()->print_str(out.c_str());
}
void executor::http_report(ex_event_name ev)
{
assert(pHttpString != nullptr);
switch(ev)
{
case EV_HTML_HASHRATE:
hashrate_report(*pHttpString);
break;
case EV_HTML_RESULTS:
result_report(*pHttpString);
break;
case EV_HTML_CONNSTAT:
connection_report(*pHttpString);
break;
default:
assert(false);
break;
}
httpReady.set_value();
}
void executor::get_http_report(ex_event_name ev_id, std::string& data)
{
std::lock_guard<std::mutex> lck(httpMutex);
assert(pHttpString == nullptr);
assert(ev_id == EV_HTML_HASHRATE || ev_id == EV_HTML_RESULTS || ev_id == EV_HTML_CONNSTAT);
pHttpString = &data;
httpReady = std::promise<void>();
std::future<void> ready = httpReady.get_future();
push_event(ex_event(ev_id));
ready.wait();
pHttpString = nullptr;
}

174
executor.h Normal file
View File

@ -0,0 +1,174 @@
#pragma once
#include "thdq.hpp"
#include "msgstruct.h"
#include <atomic>
#include <array>
#include <list>
#include <future>
#include "crypto/cryptonight.h"
class jpsock;
class minethd;
class telemetry;
class executor
{
public:
static executor* inst()
{
if (oInst == nullptr) oInst = new executor;
return oInst;
};
void ex_start() { my_thd = new std::thread(&executor::ex_main, this); }
void ex_main();
void get_http_report(ex_event_name ev_id, std::string& data);
inline void push_event(ex_event&& ev) { oEventQ.push(std::move(ev)); }
void push_timed_event(ex_event&& ev, size_t sec);
private:
struct timed_event
{
ex_event event;
size_t ticks_left;
timed_event(ex_event&& ev, size_t ticks) : event(std::move(ev)), ticks_left(ticks) {}
};
cryptonight_ctx* cpu_ctx;
// In miliseconds, has to divide a second (1000ms) into an integer number
constexpr static size_t iTickTime = 500;
// Dev donation time period in seconds. 100 minutes by default.
// We will divide up this period according to the config setting
constexpr static size_t iDevDonatePeriod = 100 * 60;
constexpr static size_t invalid_pool_id = 0;
constexpr static size_t dev_pool_id = 1;
constexpr static size_t usr_pool_id = 2;
std::list<timed_event> lTimedEvents;
std::mutex timed_event_mutex;
thdq<ex_event> oEventQ;
telemetry* telem;
std::vector<minethd*>* pvThreads;
std::thread* my_thd;
size_t current_pool_id;
jpsock* usr_pool;
jpsock* dev_pool;
jpsock* pick_pool_by_id(size_t pool_id);
bool is_dev_time;
executor();
static executor* oInst;
void ex_clock_thd();
void pool_connect(jpsock* pool);
void hashrate_report(std::string& out);
void result_report(std::string& out);
void connection_report(std::string& out);
void http_report(ex_event_name ev);
void print_report(ex_event_name ev);
std::string* pHttpString = nullptr;
std::promise<void> httpReady;
std::mutex httpMutex;
struct sck_error_log
{
std::chrono::system_clock::time_point time;
std::string msg;
sck_error_log(std::string&& err) : msg(std::move(err))
{
time = std::chrono::system_clock::now();
}
};
std::vector<sck_error_log> vSocketLog;
// Element zero is always the success element.
// Keep in mind that this is a tally and not a log like above
struct result_tally
{
std::chrono::system_clock::time_point time;
std::string msg;
size_t count;
result_tally() : msg("[OK]"), count(0)
{
time = std::chrono::system_clock::now();
}
result_tally(std::string&& err) : msg(std::move(err)), count(1)
{
time = std::chrono::system_clock::now();
}
void increment()
{
count++;
time = std::chrono::system_clock::now();
}
bool compare(std::string& err)
{
if(msg == err)
{
increment();
return true;
}
else
return false;
}
};
std::vector<result_tally> vMineResults;
//More result statistics
std::array<size_t, 10> iTopDiff { { } }; //Initialize to zero
std::chrono::system_clock::time_point tPoolConnTime;
size_t iPoolHashes;
uint64_t iPoolDiff;
// Set it to 16 bit so that we can just let it grow
// Maximum realistic growth rate - 5MB / month
std::vector<uint16_t> iPoolCallTimes;
//Those stats are reset if we disconnect
inline void reset_stats()
{
iPoolCallTimes.clear();
tPoolConnTime = std::chrono::system_clock::now();
iPoolHashes = 0;
iPoolDiff = 0;
}
double fHighestHps = 0.0;
void log_socket_error(std::string&& sError);
void log_result_error(std::string&& sError);
void log_result_ok(uint64_t iActualDiff);
void sched_reconnect();
void on_sock_ready(size_t pool_id);
void on_sock_error(size_t pool_id, std::string&& sError);
void on_pool_have_job(size_t pool_id, pool_job& oPoolJob);
void on_miner_result(size_t pool_id, job_result& oResult);
void on_reconnect(size_t pool_id);
void on_switch_pool(size_t pool_id);
inline size_t sec_to_ticks(size_t sec) { return sec * (1000 / iTickTime); }
};

97
httpd.cpp Normal file
View File

@ -0,0 +1,97 @@
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <string>
#include "msgstruct.h"
#include "httpd.h"
#include "console.h"
#include "executor.h"
#include "jconf.h"
#ifdef _WIN32
#include "libmicrohttpd/microhttpd.h"
#define strcasecmp _stricmp
#else
#include <microhttpd.h>
#endif // _WIN32
httpd* httpd::oInst = nullptr;
httpd::httpd()
{
}
int httpd::req_handler(void * cls,
MHD_Connection* connection,
const char* url,
const char* method,
const char* version,
const char* upload_data,
size_t* upload_data_size,
void ** ptr)
{
struct MHD_Response * rsp;
if (strcmp(method, "GET") != 0)
return MHD_NO;
*ptr = nullptr;
std::string str;
if(strcasecmp(url, "/h") == 0 || strcasecmp(url, "/hashrate") == 0)
{
str.append("<html><head><title>Hashrate Report</title></head><body><pre>");
executor::inst()->get_http_report(EV_HTML_HASHRATE, str);
str.append("</pre></body></html>");
rsp = MHD_create_response_from_buffer(str.size(), (void*)str.c_str(), MHD_RESPMEM_MUST_COPY);
}
else if(strcasecmp(url, "/c") == 0 || strcasecmp(url, "/connection") == 0)
{
str.append("<html><head><title>Connection Report</title></head><body><pre>");
executor::inst()->get_http_report(EV_HTML_CONNSTAT, str);
str.append("</pre></body></html>");
rsp = MHD_create_response_from_buffer(str.size(), (void*)str.c_str(), MHD_RESPMEM_MUST_COPY);
}
else if(strcasecmp(url, "/r") == 0 || strcasecmp(url, "/results") == 0)
{
str.append("<html><head><title>Results Report</title></head><body><pre>");
executor::inst()->get_http_report(EV_HTML_RESULTS, str);
str.append("</pre></body></html>");
rsp = MHD_create_response_from_buffer(str.size(), (void*)str.c_str(), MHD_RESPMEM_MUST_COPY);
}
else
{
char buffer[1024];
snprintf(buffer, sizeof(buffer), "<html><head><title>Error</title></head><body>"
"<pre>Unkown url %s - please use /h, /r or /c as url</pre></body></html>", url);
rsp = MHD_create_response_from_buffer(strlen(buffer),
(void*)buffer, MHD_RESPMEM_MUST_COPY);
}
int ret = MHD_queue_response(connection, MHD_HTTP_OK, rsp);
MHD_destroy_response(rsp);
return ret;
}
bool httpd::start_daemon()
{
d = MHD_start_daemon(MHD_USE_THREAD_PER_CONNECTION,
jconf::inst()->GetHttpdPort(), NULL, NULL,
&httpd::req_handler,
NULL, MHD_OPTION_END);
if(d == nullptr)
{
printer::inst()->print_str("HTTP Daemon failed to start.");
return false;
}
return true;
}

31
httpd.h Normal file
View File

@ -0,0 +1,31 @@
#pragma once
struct MHD_Daemon;
struct MHD_Connection;
class httpd
{
public:
static httpd* inst()
{
if (oInst == nullptr) oInst = new httpd;
return oInst;
};
bool start_daemon();
private:
httpd();
static httpd* oInst;
static int req_handler(void * cls,
MHD_Connection* connection,
const char* url,
const char* method,
const char* version,
const char* upload_data,
size_t* upload_data_size,
void ** ptr);
MHD_Daemon *d;
};

353
jconf.cpp Normal file
View File

@ -0,0 +1,353 @@
/*
* 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
* 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 <http://www.gnu.org/licenses/>.
*/
#include "jconf.h"
#include "console.h"
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef _WIN32
#define strcasecmp _stricmp
#include <intrin.h>
#else
#include <cpuid.h>
#endif
#include "rapidjson/document.h"
#include "rapidjson/error/en.h"
#include "jext.h"
#include "console.h"
using namespace rapidjson;
/*
* This enum needs to match index in oConfigValues, otherwise we will get a runtime error
*/
enum configEnum { iGpuThreadNum, aGpuThreadsConf, iPlatformIdx, sPoolAddr, sWalletAddr,
sPoolPwd, iCallTimeout, iNetRetry, iVerboseLevel, iAutohashTime, iHttpdPort, bPreferIpv4 };
struct configVal {
configEnum iName;
const char* sName;
Type iType;
};
//Same order as in configEnum, as per comment above
configVal oConfigValues[] = {
{ iGpuThreadNum, "gpu_thread_num", kNumberType },
{ aGpuThreadsConf, "gpu_threads_conf", kArrayType },
{ iPlatformIdx, "platform_index", kNumberType },
{ sPoolAddr, "pool_address", kStringType },
{ sWalletAddr, "wallet_address", kStringType },
{ sPoolPwd, "pool_password", kStringType },
{ iCallTimeout, "call_timeout", kNumberType },
{ iNetRetry, "retry_time", kNumberType },
{ iVerboseLevel, "verbose_level", kNumberType },
{ iAutohashTime, "h_print_time", kNumberType },
{ iHttpdPort, "httpd_port", kNumberType },
{ bPreferIpv4, "prefer_ipv4", kTrueType }
};
constexpr size_t iConfigCnt = (sizeof(oConfigValues)/sizeof(oConfigValues[0]));
inline bool checkType(Type have, Type want)
{
if(want == have)
return true;
else if(want == kTrueType && have == kFalseType)
return true;
else if(want == kFalseType && have == kTrueType)
return true;
else
return false;
}
struct jconf::opaque_private
{
Document jsonDoc;
const Value* configValues[iConfigCnt]; //Compile time constant
opaque_private()
{
}
};
jconf* jconf::oInst = nullptr;
jconf::jconf()
{
prv = new opaque_private();
}
bool jconf::GetThreadConfig(size_t id, thd_cfg &cfg)
{
if(id >= prv->configValues[aGpuThreadsConf]->Size())
return false;
const Value& oThdConf = prv->configValues[aGpuThreadsConf]->GetArray()[id];
if(!oThdConf.IsObject())
return false;
const Value *idx, *intensity, *w_size, *aff;
idx = GetObjectMember(oThdConf, "index");
intensity = GetObjectMember(oThdConf, "intensity");
w_size = GetObjectMember(oThdConf, "worksize");
aff = GetObjectMember(oThdConf, "affine_to_cpu");
if(idx == nullptr || intensity == nullptr || w_size == nullptr || aff == nullptr)
return false;
if(!idx->IsUint64() || !intensity->IsUint64() || !w_size->IsUint64())
return false;
if(!aff->IsUint64() && !aff->IsBool())
return false;
cfg.index = idx->GetUint64();
cfg.intensity = intensity->GetUint64();
cfg.w_size = w_size->GetUint64();
if(aff->IsNumber())
cfg.cpu_aff = aff->GetInt64();
else
cfg.cpu_aff = -1;
return true;
}
size_t jconf::GetPlatformIdx()
{
return prv->configValues[iPlatformIdx]->GetUint64();
}
const char* jconf::GetPoolAddress()
{
return prv->configValues[sPoolAddr]->GetString();
}
const char* jconf::GetPoolPwd()
{
return prv->configValues[sPoolPwd]->GetString();
}
const char* jconf::GetWalletAddress()
{
return prv->configValues[sWalletAddr]->GetString();
}
bool jconf::PreferIpv4()
{
return prv->configValues[bPreferIpv4]->GetBool();
}
size_t jconf::GetThreadCount()
{
return prv->configValues[aGpuThreadsConf]->Size();
}
uint64_t jconf::GetCallTimeout()
{
return prv->configValues[iCallTimeout]->GetUint64();
}
uint64_t jconf::GetNetRetry()
{
return prv->configValues[iNetRetry]->GetUint64();
}
uint64_t jconf::GetVerboseLevel()
{
return prv->configValues[iVerboseLevel]->GetUint64();
}
uint64_t jconf::GetAutohashTime()
{
return prv->configValues[iAutohashTime]->GetUint64();
}
uint16_t jconf::GetHttpdPort()
{
return prv->configValues[iHttpdPort]->GetUint();
}
bool jconf::check_cpu_features()
{
constexpr int AESNI_BIT = 1 << 25;
constexpr int SSE2_BIT = 1 << 26;
int cpu_info[4];
#ifdef _WIN32
__cpuid(cpu_info, 1);
#else
__cpuid(1, cpu_info[0], cpu_info[1], cpu_info[2], cpu_info[3]);
#endif
bHaveAes = (cpu_info[2] & AESNI_BIT) != 0;
return (cpu_info[3] & SSE2_BIT) != 0;
}
bool jconf::parse_config(const char* sFilename)
{
FILE * pFile;
char * buffer;
size_t flen;
if(!check_cpu_features())
{
printer::inst()->print_msg(L0, "CPU support of SSE2 is required.");
return false;
}
pFile = fopen(sFilename, "rb");
if (pFile == NULL)
{
printer::inst()->print_msg(L0, "Failed to open config file %s.", sFilename);
return false;
}
fseek(pFile,0,SEEK_END);
flen = ftell(pFile);
rewind(pFile);
if(flen >= 64*1024)
{
fclose(pFile);
printer::inst()->print_msg(L0, "Oversized config file - %s.", sFilename);
return false;
}
if(flen <= 16)
{
printer::inst()->print_msg(L0, "File is empty or too short - %s.", sFilename);
return false;
}
buffer = (char*)malloc(flen + 3);
if(fread(buffer+1, flen, 1, pFile) != 1)
{
free(buffer);
fclose(pFile);
printer::inst()->print_msg(L0, "Read error while reading %s.", sFilename);
return false;
}
fclose(pFile);
//Replace Unicode BOM with spaces - we always use UTF-8
unsigned char* ubuffer = (unsigned char*)buffer;
if(ubuffer[1] == 0xEF && ubuffer[2] == 0xBB && ubuffer[3] == 0xBF)
{
buffer[1] = ' ';
buffer[2] = ' ';
buffer[3] = ' ';
}
buffer[0] = '{';
buffer[flen] = '}';
buffer[flen + 1] = '\0';
prv->jsonDoc.Parse<kParseCommentsFlag|kParseTrailingCommasFlag>(buffer, flen+2);
free(buffer);
if(prv->jsonDoc.HasParseError())
{
printer::inst()->print_msg(L0, "JSON config parse error(offset %llu): %s",
int_port(prv->jsonDoc.GetErrorOffset()), GetParseError_En(prv->jsonDoc.GetParseError()));
return false;
}
if(!prv->jsonDoc.IsObject())
{ //This should never happen as we created the root ourselves
printer::inst()->print_msg(L0, "Invalid config file. No root?\n");
return false;
}
for(size_t i = 0; i < iConfigCnt; i++)
{
if(oConfigValues[i].iName != i)
{
printer::inst()->print_msg(L0, "Code error. oConfigValues are not in order.");
return false;
}
prv->configValues[i] = GetObjectMember(prv->jsonDoc, oConfigValues[i].sName);
if(prv->configValues[i] == nullptr)
{
printer::inst()->print_msg(L0, "Invalid config file. Missing value \"%s\".", oConfigValues[i].sName);
return false;
}
if(!checkType(prv->configValues[i]->GetType(), oConfigValues[i].iType))
{
printer::inst()->print_msg(L0, "Invalid config file. Value \"%s\" has unexpected type.", oConfigValues[i].sName);
return false;
}
}
size_t n_thd = prv->configValues[aGpuThreadsConf]->Size();
if(prv->configValues[iGpuThreadNum]->GetUint64() != n_thd)
{
printer::inst()->print_msg(L0,
"Invalid config file. Your GPU config array has %llu members, while you want to use %llu threads.",
int_port(n_thd), int_port(prv->configValues[iGpuThreadNum]->GetUint64()));
return false;
}
thd_cfg c;
for(size_t i=0; i < n_thd; i++)
{
if(!GetThreadConfig(i, c))
{
printer::inst()->print_msg(L0, "Thread %llu has invalid config.", int_port(i));
return false;
}
}
if(!prv->configValues[iCallTimeout]->IsUint64() || !prv->configValues[iNetRetry]->IsUint64())
{
printer::inst()->print_msg(L0,
"Invalid config file. call_timeout and retry_time need to be positive integers.");
return false;
}
if(!prv->configValues[iVerboseLevel]->IsUint64() || !prv->configValues[iAutohashTime]->IsUint64())
{
printer::inst()->print_msg(L0,
"Invalid config file. verbose_level and h_print_time need to be positive integers.");
return false;
}
if(!prv->configValues[iHttpdPort]->IsUint() || prv->configValues[iHttpdPort]->GetUint() > 0xFFFF)
{
printer::inst()->print_msg(L0,
"Invalid config file. httpd_port has to be in the range 0 to 65535.");
return false;
}
#ifdef _WIN32
if(GetSlowMemSetting() == no_mlck)
{
printer::inst()->print_msg(L0, "On Windows large pages need mlock. Please use another option.\n");
return false;
}
#endif // _WIN32
printer::inst()->set_verbose_level(prv->configValues[iVerboseLevel]->GetUint64());
return true;
}

53
jconf.h Normal file
View File

@ -0,0 +1,53 @@
#pragma once
#include <stdlib.h>
#include <string>
class jconf
{
public:
static jconf* inst()
{
if (oInst == nullptr) oInst = new jconf;
return oInst;
};
bool parse_config(const char* sFilename);
struct thd_cfg {
size_t index;
size_t intensity;
size_t w_size;
long long cpu_aff;
};
size_t GetThreadCount();
bool GetThreadConfig(size_t id, thd_cfg &cfg);
size_t GetPlatformIdx();
const char* GetPoolAddress();
const char* GetPoolPwd();
const char* GetWalletAddress();
uint64_t GetVerboseLevel();
uint64_t GetAutohashTime();
uint64_t GetCallTimeout();
uint64_t GetNetRetry();
uint16_t GetHttpdPort();
bool PreferIpv4();
inline bool HaveHardwareAes() { return bHaveAes; }
private:
jconf();
static jconf* oInst;
bool check_cpu_features();
struct opaque_private;
opaque_private* prv;
bool bHaveAes;
};

13
jext.h Normal file
View File

@ -0,0 +1,13 @@
#pragma once
using namespace rapidjson;
/* This macro brings rapidjson more in line with other libs */
inline const Value* GetObjectMember(const Value& obj, const char* key)
{
Value::ConstMemberIterator itr = obj.FindMember(key);
if (itr != obj.MemberEnd())
return &itr->value;
else
return nullptr;
}

671
jpsock.cpp Normal file
View File

@ -0,0 +1,671 @@
/*
* 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
* 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 <http://www.gnu.org/licenses/>.
*/
#include <stdarg.h>
#include <assert.h>
#include "jpsock.h"
#include "executor.h"
#include "jconf.h"
#include "rapidjson/document.h"
#include "jext.h"
#include "socks.h"
#define AGENTID_STR "xmr-stak-cpu/1.0"
using namespace rapidjson;
struct jpsock::call_rsp
{
bool bHaveResponse;
uint64_t iCallId;
Value* pCallData;
std::string sCallErr;
call_rsp(Value* val) : pCallData(val)
{
bHaveResponse = false;
iCallId = 0;
sCallErr.clear();
}
};
typedef GenericDocument<UTF8<>, MemoryPoolAllocator<>, MemoryPoolAllocator<>> MemDocument;
/*
*
* !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
* ASSUMPTION - only one calling thread. Multiple calling threads would require better
* thread safety. The calling thread is assumed to be the executor thread.
* If there is a reason to call the pool outside of the executor context, consider
* doing it via an executor event.
* !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
*
* Call values and allocators are for the calling thread (executor). When processing
* a call, the recv thread will make a copy of the call response and then erase its copy.
*/
struct jpsock::opaque_private
{
addrinfo *pSockAddr;
addrinfo *pAddrRoot;
SOCKET hSocket;
Value oCallValue;
MemoryPoolAllocator<> callAllocator;
MemoryPoolAllocator<> recvAllocator;
MemoryPoolAllocator<> parseAllocator;
MemDocument jsonDoc;
call_rsp oCallRsp;
opaque_private(uint8_t* bCallMem, uint8_t* bRecvMem, uint8_t* bParseMem) :
callAllocator(bCallMem, jpsock::iJsonMemSize),
recvAllocator(bRecvMem, jpsock::iJsonMemSize),
parseAllocator(bParseMem, jpsock::iJsonMemSize),
jsonDoc(&recvAllocator, jpsock::iJsonMemSize, &parseAllocator),
oCallRsp(nullptr)
{
hSocket = INVALID_SOCKET;
pSockAddr = nullptr;
}
};
struct jpsock::opq_json_val
{
const Value* val;
opq_json_val(const Value* val) : val(val) {}
};
jpsock::jpsock(size_t id) : pool_id(id)
{
sock_init();
bJsonCallMem = (uint8_t*)malloc(iJsonMemSize);
bJsonRecvMem = (uint8_t*)malloc(iJsonMemSize);
bJsonParseMem = (uint8_t*)malloc(iJsonMemSize);
prv = new opaque_private(bJsonCallMem, bJsonRecvMem, bJsonParseMem);
oRecvThd = nullptr;
bRunning = false;
bLoggedIn = false;
iJobDiff = 0;
memset(&oCurrentJob, 0, sizeof(oCurrentJob));
}
jpsock::~jpsock()
{
delete prv;
prv = nullptr;
free(bJsonCallMem);
free(bJsonRecvMem);
free(bJsonParseMem);
}
std::string&& jpsock::get_call_error()
{
return std::move(prv->oCallRsp.sCallErr);
}
inline bool jpsock::set_socket_error(const char* a)
{
if(!bHaveSocketError)
{
bHaveSocketError = true;
sSocketError.assign(a);
}
return false;
}
inline bool jpsock::set_socket_error(const char* a, const char* b)
{
if(!bHaveSocketError)
{
bHaveSocketError = true;
size_t ln_a = strlen(a);
size_t ln_b = strlen(b);
sSocketError.reserve(ln_a + ln_b + 2);
sSocketError.assign(a, ln_a);
sSocketError.append(b, ln_b);
}
return false;
}
bool jpsock::set_socket_error_strerr(const char* a)
{
char sSockErrText[512];
return set_socket_error(a, sock_strerror(sSockErrText, sizeof(sSockErrText)));
}
bool jpsock::set_socket_error_strerr(const char* a, int res)
{
char sSockErrText[512];
return set_socket_error(a, sock_gai_strerror(res, sSockErrText, sizeof(sSockErrText)));
}
void jpsock::jpsock_thread()
{
jpsock_thd_main();
executor::inst()->push_event(ex_event(std::move(sSocketError), pool_id));
// If a call is wating, send an error to end it
bool bCallWaiting = false;
std::unique_lock<std::mutex> mlock(call_mutex);
if(prv->oCallRsp.pCallData != nullptr)
{
prv->oCallRsp.bHaveResponse = true;
prv->oCallRsp.iCallId = 0;
prv->oCallRsp.pCallData = nullptr;
bCallWaiting = true;
}
mlock.unlock();
if(bCallWaiting)
call_cond.notify_one();
bRunning = false;
bLoggedIn = false;
std::unique_lock<std::mutex>(job_mutex);
memset(&oCurrentJob, 0, sizeof(oCurrentJob));
}
bool jpsock::jpsock_thd_main()
{
int ret = ::connect(prv->hSocket, prv->pSockAddr->ai_addr, (int)prv->pSockAddr->ai_addrlen);
freeaddrinfo(prv->pAddrRoot);
prv->pAddrRoot = nullptr;
if (ret != 0)
return set_socket_error_strerr("CONNECT error: ");
executor::inst()->push_event(ex_event(EV_SOCK_READY, pool_id));
char buf[iSockBufferSize];
size_t datalen = 0;
while (true)
{
ret = recv(prv->hSocket, buf + datalen, sizeof(buf) - datalen, 0);
if(ret == 0)
return set_socket_error("RECEIVE error: socket closed");
if(ret == SOCKET_ERROR || ret < 0)
return set_socket_error("RECEIVE error: ");
datalen += ret;
if (datalen >= sizeof(buf))
{
sock_close(prv->hSocket);
return set_socket_error("RECEIVE error: data overflow");
}
char* lnend;
char* lnstart = buf;
while ((lnend = (char*)memchr(lnstart, '\n', datalen)) != nullptr)
{
lnend++;
int lnlen = lnend - lnstart;
if (!process_line(lnstart, lnlen))
{
sock_close(prv->hSocket);
return false;
}
datalen -= lnlen;
lnstart = lnend;
}
//Got leftover data? Move it to the front
if (datalen > 0 && buf != lnstart)
memmove(buf, lnstart, datalen);
}
}
bool jpsock::process_line(char* line, size_t len)
{
prv->jsonDoc.SetNull();
prv->parseAllocator.Clear();
prv->callAllocator.Clear();
/*NULL terminate the line instead of '\n', parsing will add some more NULLs*/
line[len-1] = '\0';
//printf("RECV: %s\n", line);
if (prv->jsonDoc.ParseInsitu(line).HasParseError())
return set_socket_error("PARSE error: Invalid JSON");
if (!prv->jsonDoc.IsObject())
return set_socket_error("PARSE error: Invalid root");
const Value* mt;
if (prv->jsonDoc.HasMember("method"))
{
mt = GetObjectMember(prv->jsonDoc, "method");
if(!mt->IsString())
return set_socket_error("PARSE error: Protocol error 1");
if(strcmp(mt->GetString(), "job") != 0)
return set_socket_error("PARSE error: Unsupported server method ", mt->GetString());
mt = GetObjectMember(prv->jsonDoc, "params");
if(mt == nullptr || !mt->IsObject())
return set_socket_error("PARSE error: Protocol error 2");
opq_json_val v(mt);
return process_pool_job(&v);
}
else
{
uint64_t iCallId;
mt = GetObjectMember(prv->jsonDoc, "id");
if (mt == nullptr || !mt->IsUint64())
return set_socket_error("PARSE error: Protocol error 3");
iCallId = mt->GetUint64();
mt = GetObjectMember(prv->jsonDoc, "error");
const char* sError = nullptr;
size_t iErrorLn = 0;
if (mt == nullptr || mt->IsNull())
{
/* If there was no error we need a result */
if ((mt = GetObjectMember(prv->jsonDoc, "result")) == nullptr)
return set_socket_error("PARSE error: Protocol error 7");
}
else
{
if(!mt->IsObject())
return set_socket_error("PARSE error: Protocol error 5");
const Value* msg = GetObjectMember(*mt, "message");
if(msg == nullptr || !msg->IsString())
return set_socket_error("PARSE error: Protocol error 6");
iErrorLn = msg->GetStringLength();
sError = msg->GetString();
}
std::unique_lock<std::mutex> mlock(call_mutex);
if (prv->oCallRsp.pCallData == nullptr)
{
/*Server sent us a call reply without us making a call*/
mlock.unlock();
return set_socket_error("PARSE error: Unexpected call response");
}
prv->oCallRsp.bHaveResponse = true;
prv->oCallRsp.iCallId = iCallId;
if(sError != nullptr)
{
prv->oCallRsp.pCallData = nullptr;
prv->oCallRsp.sCallErr.assign(sError, iErrorLn);
}
else
prv->oCallRsp.pCallData->CopyFrom(*mt, prv->callAllocator);
mlock.unlock();
call_cond.notify_one();
return true;
}
}
bool jpsock::process_pool_job(const opq_json_val* params)
{
if (!params->val->IsObject())
return set_socket_error("PARSE error: Job error 1");
const Value * blob, *jobid, *target;
jobid = GetObjectMember(*params->val, "job_id");
blob = GetObjectMember(*params->val, "blob");
target = GetObjectMember(*params->val, "target");
if (jobid == nullptr || blob == nullptr || target == nullptr ||
!jobid->IsString() || !blob->IsString() || !target->IsString())
{
return set_socket_error("PARSE error: Job error 2");
}
if (jobid->GetStringLength() >= sizeof(pool_job::sJobID)) // Note >=
return set_socket_error("PARSE error: Job error 3");
uint32_t iWorkLn = blob->GetStringLength() / 2;
if (iWorkLn > sizeof(pool_job::bWorkBlob))
return set_socket_error("PARSE error: Invalid job legth. Are you sure you are mining the correct coin?");
pool_job oPoolJob;
if (!hex2bin(blob->GetString(), iWorkLn * 2, oPoolJob.bWorkBlob))
return set_socket_error("PARSE error: Job error 4");
oPoolJob.iWorkLen = iWorkLn;
memset(oPoolJob.sJobID, 0, sizeof(pool_job::sJobID));
memcpy(oPoolJob.sJobID, jobid->GetString(), jobid->GetStringLength()); //Bounds checking at proto error 3
size_t target_slen = target->GetStringLength();
if(target_slen <= 8)
{
uint32_t iTempInt;
char sTempStr[] = "00000000"; // Little-endian CPU FTW
memcpy(sTempStr, target->GetString(), target_slen);
hex2bin(sTempStr, 8, (unsigned char*)&iTempInt);
oPoolJob.iTarget = iTempInt;
}
else
return set_socket_error("PARSE error: Job error 5");
iJobDiff = t32_to_diff(oPoolJob.iTarget);
executor::inst()->push_event(ex_event(oPoolJob, pool_id));
std::unique_lock<std::mutex>(job_mutex);
oCurrentJob = oPoolJob;
return true;
}
bool jpsock::connect(const char* sAddr, std::string& sConnectError)
{
if(prv_connect(sAddr))
return true;
sConnectError = std::move(sSocketError);
return false;
}
bool jpsock::prv_connect(const char* sAddr)
{
char sAddrMb[256];
char *sTmp, *sPort;
bHaveSocketError = false;
sSocketError.clear();
iJobDiff = 0;
size_t ln = strlen(sAddr);
if (ln >= sizeof(sAddrMb))
return set_socket_error("CONNECT error: Pool address overflow.");
memcpy(sAddrMb, sAddr, ln);
sAddrMb[ln] = '\0';
if ((sTmp = strstr(sAddrMb, "//")) != nullptr)
memmove(sAddrMb, sTmp, strlen(sTmp) + 1);
if ((sPort = strchr(sAddrMb, ':')) == nullptr)
return set_socket_error("CONNECT error: Pool port number not specified, please use format <hostname>:<port>.");
sPort[0] = '\0';
sPort++;
addrinfo hints = { 0 };
hints.ai_family = AF_UNSPEC;
hints.ai_socktype = SOCK_STREAM;
hints.ai_protocol = IPPROTO_TCP;
prv->pAddrRoot = nullptr;
int err;
if ((err = getaddrinfo(sAddrMb, sPort, &hints, &prv->pAddrRoot)) != 0)
return set_socket_error_strerr("CONNECT error: GetAddrInfo: ", err);
addrinfo *ptr = prv->pAddrRoot;
addrinfo *ipv4 = nullptr, *ipv6 = nullptr;
while (ptr != nullptr)
{
if (ptr->ai_family == AF_INET)
ipv4 = ptr;
if (ptr->ai_family == AF_INET6)
ipv6 = ptr;
ptr = ptr->ai_next;
}
if (ipv4 == nullptr && ipv6 == nullptr)
{
freeaddrinfo(prv->pAddrRoot);
prv->pAddrRoot = nullptr;
return set_socket_error("CONNECT error: I found some DNS records but no IPv4 or IPv6 addresses.");
}
else if (ipv4 != nullptr && ipv6 == nullptr)
prv->pSockAddr = ipv4;
else if (ipv4 == nullptr && ipv6 != nullptr)
prv->pSockAddr = ipv6;
else if (ipv4 != nullptr && ipv6 != nullptr)
{
if(jconf::inst()->PreferIpv4())
prv->pSockAddr = ipv4;
else
prv->pSockAddr = ipv6;
}
prv->hSocket = socket(prv->pSockAddr->ai_family, prv->pSockAddr->ai_socktype, prv->pSockAddr->ai_protocol);
if (prv->hSocket == INVALID_SOCKET)
{
freeaddrinfo(prv->pAddrRoot);
prv->pAddrRoot = nullptr;
return set_socket_error_strerr("CONNECT error: Socket creation failed ");
}
bRunning = true;
oRecvThd = new std::thread(&jpsock::jpsock_thread, this);
return true;
}
void jpsock::disconnect()
{
if(prv->hSocket != INVALID_SOCKET)
{
sock_close(prv->hSocket);
prv->hSocket = INVALID_SOCKET;
}
if(oRecvThd != nullptr)
{
oRecvThd->join();
delete oRecvThd;
oRecvThd = nullptr;
}
}
bool jpsock::cmd_ret_wait(const char* sPacket, opq_json_val& poResult)
{
//printf("SEND: %s\n", sPacket);
/*Set up the call rsp for the call reply*/
prv->oCallValue.SetNull();
prv->callAllocator.Clear();
std::unique_lock<std::mutex> mlock(call_mutex);
prv->oCallRsp = call_rsp(&prv->oCallValue);
mlock.unlock();
int pos = 0, slen = strlen(sPacket);
while (pos != slen)
{
int ret = send(prv->hSocket, sPacket + pos, slen - pos, 0);
if (ret == SOCKET_ERROR)
{
set_socket_error_strerr("SEND error: ");
disconnect(); //This will join the other thread;
return false;
}
else
pos += ret;
}
//Success is true if the server approves, result is true if there was no socket error
bool bSuccess;
mlock.lock();
bool bResult = call_cond.wait_for(mlock, std::chrono::seconds(jconf::inst()->GetCallTimeout()),
[&]() { return prv->oCallRsp.bHaveResponse; });
bSuccess = prv->oCallRsp.pCallData != nullptr;
prv->oCallRsp.pCallData = nullptr;
mlock.unlock();
if(bHaveSocketError)
return false;
//This means that there was no socket error, but the server is not taking to us
if(!bResult)
{
set_socket_error("CALL error: Timeout while waiting for a reply");
disconnect();
return false;
}
if(bSuccess)
poResult.val = &prv->oCallValue;
return bSuccess;
}
bool jpsock::cmd_login(const char* sLogin, const char* sPassword)
{
char cmd_buffer[1024];
snprintf(cmd_buffer, sizeof(cmd_buffer), "{\"method\":\"login\",\"params\":{\"login\":\"%s\",\"pass\":\"%s\",\"agent\":\"" AGENTID_STR "\"},\"id\":1}\n",
sLogin, sPassword);
opq_json_val oResult(nullptr);
/*Normal error conditions (failed login etc..) will end here*/
if (!cmd_ret_wait(cmd_buffer, oResult))
return false;
if (!oResult.val->IsObject())
{
set_socket_error("PARSE error: Login protocol error 1");
disconnect();
return false;
}
const Value* id = GetObjectMember(*oResult.val, "id");
const Value* job = GetObjectMember(*oResult.val, "job");
if (id == nullptr || job == nullptr || !id->IsString())
{
set_socket_error("PARSE error: Login protocol error 2");
disconnect();
return false;
}
if (id->GetStringLength() >= sizeof(sMinerId))
{
set_socket_error("PARSE error: Login protocol error 3");
disconnect();
return false;
}
memset(sMinerId, 0, sizeof(sMinerId));
memcpy(sMinerId, id->GetString(), id->GetStringLength());
opq_json_val v(job);
if(!process_pool_job(&v))
{
disconnect();
return false;
}
bLoggedIn = true;
return true;
}
bool jpsock::cmd_submit(const char* sJobId, uint32_t iNonce, const uint8_t* bResult)
{
char cmd_buffer[1024];
char sNonce[9];
char sResult[65];
bin2hex((unsigned char*)&iNonce, 4, sNonce);
sNonce[8] = '\0';
bin2hex(bResult, 32, sResult);
sResult[64] = '\0';
snprintf(cmd_buffer, sizeof(cmd_buffer), "{\"method\":\"submit\",\"params\":{\"id\":\"%s\",\"job_id\":\"%s\",\"nonce\":\"%s\",\"result\":\"%s\"},\"id\":1}\n",
sMinerId, sJobId, sNonce, sResult);
opq_json_val oResult(nullptr);
return cmd_ret_wait(cmd_buffer, oResult);
}
bool jpsock::get_current_job(pool_job& job)
{
std::unique_lock<std::mutex>(job_mutex);
if(oCurrentJob.iWorkLen == 0)
return false;
oCurrentJob.iResumeCnt++;
job = oCurrentJob;
return true;
}
inline unsigned char hf_hex2bin(char c, bool &err)
{
if (c >= '0' && c <= '9')
return c - '0';
else if (c >= 'a' && c <= 'f')
return c - 'a' + 0xA;
else if (c >= 'A' && c <= 'F')
return c - 'A' + 0xA;
err = true;
return 0;
}
bool jpsock::hex2bin(const char* in, unsigned int len, unsigned char* out)
{
bool error = false;
for (unsigned int i = 0; i < len; i += 2)
{
out[i / 2] = (hf_hex2bin(in[i], error) << 4) | hf_hex2bin(in[i + 1], error);
if (error) return false;
}
return true;
}
inline char hf_bin2hex(unsigned char c)
{
if (c <= 0x9)
return '0' + c;
else
return 'a' - 0xA + c;
}
void jpsock::bin2hex(const unsigned char* in, unsigned int len, char* out)
{
for (unsigned int i = 0; i < len; i++)
{
out[i * 2] = hf_bin2hex((in[i] & 0xF0) >> 4);
out[i * 2 + 1] = hf_bin2hex(in[i] & 0x0F);
}
}

97
jpsock.h Normal file
View File

@ -0,0 +1,97 @@
#pragma once
#include <mutex>
#include <atomic>
#include <condition_variable>
#include <thread>
#include <string>
#include "msgstruct.h"
/* Our pool can have two kinds of errors:
- Parsing or connection error
Those are fatal errors (we drop the connection if we encounter them).
After they are constructed from const char* strings from various places.
(can be from read-only mem), we passs them in an exectutor message
once the recv thread expires.
- Call error
This error happens when the "server says no". Usually because the job was
outdated, or we somehow got the hash wrong. It isn't fatal.
We parse it in-situ in the network buffer, after that we copy it to a
std::string. Executor will move the buffer via an r-value ref.
*/
class jpsock
{
public:
jpsock(size_t id);
~jpsock();
bool connect(const char* sAddr, std::string& sConnectError);
void disconnect();
bool cmd_login(const char* sLogin, const char* sPassword);
bool cmd_submit(const char* sJobId, uint32_t iNonce, const uint8_t* bResult);
static bool hex2bin(const char* in, unsigned int len, unsigned char* out);
static void bin2hex(const unsigned char* in, unsigned int len, char* out);
inline bool is_running() { return bRunning; }
inline bool is_logged_in() { return bLoggedIn; }
std::string&& get_call_error();
bool have_sock_error() { return bHaveSocketError; }
inline static uint64_t t32_to_t64(uint32_t t) { return 0xFFFFFFFFFFFFFFFFULL / (0xFFFFFFFFULL / ((uint64_t)t)); }
inline static uint64_t t64_to_diff(uint64_t t) { return 0xFFFFFFFFFFFFFFFFULL / t; }
inline static uint64_t t32_to_diff(uint32_t t) { return 0xFFFFFFFF / t; }
inline static uint64_t diff_to_t64(uint64_t d) { return 0xFFFFFFFFFFFFFFFFULL / d; }
inline uint64_t get_current_diff() { return iJobDiff; }
bool get_current_job(pool_job& job);
size_t pool_id;
private:
std::atomic<bool> bRunning;
std::atomic<bool> bLoggedIn;
uint8_t* bJsonRecvMem;
uint8_t* bJsonParseMem;
uint8_t* bJsonCallMem;
static constexpr size_t iJsonMemSize = 4096;
static constexpr size_t iSockBufferSize = 4096;
struct call_rsp;
struct opaque_private;
struct opq_json_val;
void jpsock_thread();
bool jpsock_thd_main();
bool process_line(char* line, size_t len);
bool process_pool_job(const opq_json_val* params);
bool cmd_ret_wait(const char* sPacket, opq_json_val& poResult);
char sMinerId[64];
std::atomic<uint64_t> iJobDiff;
std::string sSocketError;
std::atomic<bool> bHaveSocketError;
bool set_socket_error(const char* a);
bool set_socket_error(const char* a, const char* b);
bool set_socket_error_strerr(const char* a);
bool set_socket_error_strerr(const char* a, int res);
bool prv_connect(const char* sAddr);
std::mutex call_mutex;
std::condition_variable call_cond;
std::thread* oRecvThd;
std::mutex job_mutex;
pool_job oCurrentJob;
opaque_private* prv;
};

2875
libmicrohttpd/microhttpd.h Normal file
View File

File diff suppressed because it is too large Load Diff

255
minethd.cpp Normal file
View File

@ -0,0 +1,255 @@
/*
* 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
* 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 <http://www.gnu.org/licenses/>.
*/
#include <assert.h>
#include <cmath>
#include <chrono>
#include <thread>
#include "console.h"
#ifdef _WIN32
#include <windows.h>
void thd_setaffinity(std::thread::native_handle_type h, uint64_t cpu_id)
{
SetThreadAffinityMask(h, 1 << cpu_id);
}
#else
#include <pthread.h>
void thd_setaffinity(std::thread::native_handle_type h, uint64_t cpu_id)
{
cpu_set_t mn;
CPU_ZERO(&mn);
CPU_SET(cpu_id, &mn);
pthread_setaffinity_np(h, sizeof(cpu_set_t), &mn);
}
#endif // _WIN32
#include "executor.h"
#include "minethd.h"
#include "jconf.h"
#include "crypto/cryptonight.h"
telemetry::telemetry(size_t iThd)
{
ppHashCounts = new uint64_t*[iThd];
ppTimestamps = new uint64_t*[iThd];
iBucketTop = new uint32_t[iThd];
for (size_t i = 0; i < iThd; i++)
{
ppHashCounts[i] = new uint64_t[iBucketSize];
ppTimestamps[i] = new uint64_t[iBucketSize];
iBucketTop[i] = 0;
memset(ppHashCounts[0], 0, sizeof(uint64_t) * iBucketSize);
memset(ppTimestamps[0], 0, sizeof(uint64_t) * iBucketSize);
}
}
double telemetry::calc_telemetry_data(size_t iLastMilisec, size_t iThread)
{
using namespace std::chrono;
uint64_t iTimeNow = time_point_cast<milliseconds>(high_resolution_clock::now()).time_since_epoch().count();
uint64_t iEarliestHashCnt = 0;
uint64_t iEarliestStamp = 0;
uint64_t iLastestStamp = 0;
uint64_t iLastestHashCnt = 0;
bool bHaveFullSet = false;
//Start at 1, buckettop points to next empty
for (size_t i = 1; i < iBucketSize; i++)
{
size_t idx = (iBucketTop[iThread] - i) & iBucketMask; //overflow expected here
if (ppTimestamps[iThread][idx] == 0)
break; //That means we don't have the data yet
if (iLastestStamp == 0)
{
iLastestStamp = ppTimestamps[iThread][idx];
iLastestHashCnt = ppHashCounts[iThread][idx];
}
if (iTimeNow - ppTimestamps[iThread][idx] > iLastMilisec)
{
bHaveFullSet = true;
break; //We are out of the requested time period
}
iEarliestStamp = ppTimestamps[iThread][idx];
iEarliestHashCnt = ppHashCounts[iThread][idx];
}
if (!bHaveFullSet || iEarliestStamp == 0 || iLastestStamp == 0)
return nan("");
double fHashes, fTime;
fHashes = iLastestHashCnt - iEarliestHashCnt;
fTime = iLastestStamp - iEarliestStamp;
fTime /= 1000.0;
return fHashes / fTime;
}
void telemetry::push_perf_value(size_t iThd, uint64_t iHashCount, uint64_t iTimestamp)
{
size_t iTop = iBucketTop[iThd];
ppHashCounts[iThd][iTop] = iHashCount;
ppTimestamps[iThd][iTop] = iTimestamp;
iBucketTop[iThd] = (iTop + 1) & iBucketMask;
}
minethd::minethd(miner_work& pWork, size_t iNo, GpuContext* ctx)
{
oWork = pWork;
bQuit = 0;
iThreadNo = (uint8_t)iNo;
iJobNo = 0;
iHashCount = 0;
iTimestamp = 0;
pGpuCtx = ctx;
oWorkThd = std::thread(&minethd::work_main, this);
}
bool minethd::init_gpus()
{
size_t i, n = jconf::inst()->GetThreadCount();
printer::inst()->print_msg(L1, "Compiling code and initializing GPUs. This will take a while...");
vGpuData.resize(n);
jconf::thd_cfg cfg;
for(i = 0; i < n; i++)
{
jconf::inst()->GetThreadConfig(i, cfg);
vGpuData[i].deviceIdx = cfg.index;
vGpuData[i].rawIntensity = cfg.intensity;
vGpuData[i].workSize = cfg.w_size;
}
return InitOpenCL(vGpuData.data(), n, jconf::inst()->GetPlatformIdx()) == ERR_SUCCESS;
}
std::atomic<uint64_t> minethd::iGlobalJobNo;
std::atomic<uint64_t> minethd::iConsumeCnt; //Threads get jobs as they are initialized
minethd::miner_work minethd::oGlobalWork;
uint64_t minethd::iThreadCount = 0;
std::vector<GpuContext> minethd::vGpuData;
std::vector<minethd*>* minethd::thread_starter(miner_work& pWork)
{
iGlobalJobNo = 0;
iConsumeCnt = 0;
std::vector<minethd*>* pvThreads = new std::vector<minethd*>;
size_t i, n = jconf::inst()->GetThreadCount();
pvThreads->reserve(n);
jconf::thd_cfg cfg;
for (i = 0; i < n; i++)
{
jconf::inst()->GetThreadConfig(i, cfg);
minethd* thd = new minethd(pWork, i, &vGpuData[i]);
if(cfg.cpu_aff >= 0)
thd_setaffinity(thd->oWorkThd.native_handle(), cfg.cpu_aff);
pvThreads->push_back(thd);
if(cfg.cpu_aff >= 0)
printer::inst()->print_msg(L1, "Starting GPU thread, affinity: %d.", (int)cfg.cpu_aff);
else
printer::inst()->print_msg(L1, "Starting GPU thread, no affinity.");
}
iThreadCount = n;
return pvThreads;
}
void minethd::switch_work(miner_work& pWork)
{
// iConsumeCnt is a basic lock-like polling mechanism just in case we happen to push work
// faster than threads can consume them. This should never happen in real life.
// Pool cant physically send jobs faster than every 250ms or so due to net latency.
while (iConsumeCnt.load(std::memory_order_seq_cst) < iThreadCount)
std::this_thread::sleep_for(std::chrono::milliseconds(100));
oGlobalWork = pWork;
iConsumeCnt.store(0, std::memory_order_seq_cst);
iGlobalJobNo++;
}
void minethd::consume_work()
{
memcpy(&oWork, &oGlobalWork, sizeof(miner_work));
iJobNo++;
iConsumeCnt++;
if(!oWork.bStall)
{
pGpuCtx->Nonce = calc_start_nonce(oWork.iResumeCnt);
XMRSetJob(pGpuCtx, oWork.bWorkBlob, oWork.iWorkSize, oWork.iTarget);
}
}
void minethd::work_main()
{
uint64_t iCount = 0;
iConsumeCnt++;
while (bQuit == 0)
{
if (oWork.bStall)
{
/* We are stalled here because the executor didn't find a job for us yet,
either because of network latency, or a socket problem. Since we are
raison d'etre of this software it us sensible to just wait until we have something*/
while (iGlobalJobNo.load(std::memory_order_relaxed) == iJobNo)
std::this_thread::sleep_for(std::chrono::milliseconds(100));
consume_work();
continue;
}
assert(sizeof(job_result::sJobID) == sizeof(pool_job::sJobID));
while(iGlobalJobNo.load(std::memory_order_relaxed) == iJobNo)
{
cl_uint results[0x100] = { 0 };
XMRRunJob(pGpuCtx, results);
for(size_t i = 0; i < results[0xFF]; i++)
{
executor::inst()->push_event(ex_event(job_result(oWork.sJobID, oWork.bWorkBlob,
oWork.iWorkSize, oWork.iTarget, results[i]), oWork.iPoolId));
}
iCount += pGpuCtx->rawIntensity;
using namespace std::chrono;
uint64_t iStamp = time_point_cast<milliseconds>(high_resolution_clock::now()).time_since_epoch().count();
iHashCount.store(iCount, std::memory_order_relaxed);
iTimestamp.store(iStamp, std::memory_order_relaxed);
std::this_thread::yield();
}
consume_work();
}
}

133
minethd.h Normal file
View File

@ -0,0 +1,133 @@
#pragma once
#include <thread>
#include <atomic>
#include "amd_gpu/gpu.h"
class telemetry
{
public:
telemetry(size_t iThd);
void push_perf_value(size_t iThd, uint64_t iHashCount, uint64_t iTimestamp);
double calc_telemetry_data(size_t iLastMilisec, size_t iThread);
private:
constexpr static size_t iBucketSize = 2 << 11; //Power of 2 to simplify calculations
constexpr static size_t iBucketMask = iBucketSize - 1;
uint32_t* iBucketTop;
uint64_t** ppHashCounts;
uint64_t** ppTimestamps;
};
class minethd
{
public:
struct miner_work
{
char sJobID[64];
uint8_t bWorkBlob[88];
uint32_t iWorkSize;
uint32_t iResumeCnt;
uint32_t iTarget;
bool bStall;
size_t iPoolId;
miner_work() : iWorkSize(0), bStall(true), iPoolId(0) { }
miner_work(const char* sJobID, const uint8_t* bWork, uint32_t iWorkSize, uint32_t iResumeCnt,
uint64_t iTarget, size_t iPoolId) : iWorkSize(iWorkSize), iResumeCnt(iResumeCnt),
iTarget(iTarget), bStall(false), iPoolId(iPoolId)
{
assert(iWorkSize <= sizeof(bWorkBlob));
memcpy(this->sJobID, sJobID, sizeof(miner_work::sJobID));
memcpy(this->bWorkBlob, bWork, iWorkSize);
}
miner_work(miner_work const&) = delete;
miner_work& operator=(miner_work const& from)
{
assert(this != &from);
iWorkSize = from.iWorkSize;
iResumeCnt = from.iResumeCnt;
iTarget = from.iTarget;
bStall = from.bStall;
iPoolId = from.iPoolId;
assert(iWorkSize <= sizeof(bWorkBlob));
memcpy(sJobID, from.sJobID, sizeof(sJobID));
memcpy(bWorkBlob, from.bWorkBlob, iWorkSize);
return *this;
}
miner_work(miner_work&& from) : iWorkSize(from.iWorkSize), iTarget(from.iTarget),
bStall(from.bStall), iPoolId(from.iPoolId)
{
assert(iWorkSize <= sizeof(bWorkBlob));
memcpy(sJobID, from.sJobID, sizeof(sJobID));
memcpy(bWorkBlob, from.bWorkBlob, iWorkSize);
}
miner_work& operator=(miner_work&& from)
{
assert(this != &from);
iWorkSize = from.iWorkSize;
iResumeCnt = from.iResumeCnt;
iTarget = from.iTarget;
bStall = from.bStall;
iPoolId = from.iPoolId;
assert(iWorkSize <= sizeof(bWorkBlob));
memcpy(sJobID, from.sJobID, sizeof(sJobID));
memcpy(bWorkBlob, from.bWorkBlob, iWorkSize);
return *this;
}
};
static void switch_work(miner_work& pWork);
static std::vector<minethd*>* thread_starter(miner_work& pWork);
static bool init_gpus();
std::atomic<uint64_t> iHashCount;
std::atomic<uint64_t> iTimestamp;
private:
minethd(miner_work& pWork, size_t iNo, GpuContext* ctx);
// We use the top 8 bits of the nonce for thread and resume
// This allows us to resume up to 64 threads 4 times before
// we get nonce collisions
// Bottom 24 bits allow for an hour of work at 4000 H/s
inline uint32_t calc_start_nonce(uint32_t resume)
{ return (resume * iThreadCount + iThreadNo) << 24; }
void work_main();
void double_work_main();
void consume_work();
static std::atomic<uint64_t> iGlobalJobNo;
static std::atomic<uint64_t> iConsumeCnt;
static uint64_t iThreadCount;
uint64_t iJobNo;
static miner_work oGlobalWork;
miner_work oWork;
std::thread oWorkThd;
uint8_t iThreadNo;
bool bQuit;
bool bNoPrefetch;
//Mutable ptr to vector below, different for each thread
GpuContext* pGpuCtx;
// WARNING - this vector (but not its contents) must be immutable
// once the threads are started
static std::vector<GpuContext> vGpuData;
};

139
msgstruct.h Normal file
View File

@ -0,0 +1,139 @@
#pragma once
#include <string.h>
#include <assert.h>
// Structures that we use to pass info between threads constructors are here just to make
// the stack allocation take up less space, heap is a shared resouce that needs locks too of course
struct pool_job
{
char sJobID[64];
uint8_t bWorkBlob[88];
uint32_t iTarget;
uint32_t iWorkLen;
uint32_t iResumeCnt;
pool_job() : iWorkLen(0), iResumeCnt(0) {}
pool_job(const char* sJobID, uint64_t iTarget, const uint8_t* bWorkBlob, uint32_t iWorkLen) :
iTarget(iTarget), iWorkLen(iWorkLen), iResumeCnt(0)
{
assert(iWorkLen <= sizeof(pool_job::bWorkBlob));
memcpy(this->sJobID, sJobID, sizeof(pool_job::sJobID));
memcpy(this->bWorkBlob, bWorkBlob, iWorkLen);
}
};
struct job_result
{
uint8_t bResult[32];
char sJobID[64];
uint8_t bWorkBlob[88];
uint32_t iTarget;
uint32_t iWorkLen;
uint32_t iNonce;
job_result() {}
job_result(const char* sJobID, uint8_t* bWorkBlob, uint32_t iWorkLen, uint32_t iTarget, uint32_t iNonce) :
iTarget(iTarget), iWorkLen(iWorkLen), iNonce(iNonce)
{
assert(iWorkLen <= sizeof(job_result::bWorkBlob));
memcpy(this->sJobID, sJobID, sizeof(job_result::sJobID));
memcpy(this->bWorkBlob, bWorkBlob, iWorkLen);
memset(this->bResult, 0, sizeof(job_result::bResult));
}
};
enum ex_event_name { EV_INVALID_VAL, EV_SOCK_READY, EV_SOCK_ERROR,
EV_POOL_HAVE_JOB, EV_MINER_HAVE_RESULT, EV_PERF_TICK, EV_RECONNECT,
EV_SWITCH_POOL, EV_DEV_POOL_EXIT, EV_USR_HASHRATE, EV_USR_RESULTS, EV_USR_CONNSTAT,
EV_HASHRATE_LOOP, EV_HTML_HASHRATE, EV_HTML_RESULTS, EV_HTML_CONNSTAT };
/*
This is how I learned to stop worrying and love c++11 =).
Ghosts of endless heap allocations have finally been exorcised. Thanks
to the nifty magic of move semantics, string will only be allocated
once on the heap. Considering that it makes a jorney across stack,
heap alloced queue, to another stack before being finally processed
I think it is kind of nifty, don't you?
Also note that for non-arg events we only copy two qwords
*/
struct ex_event
{
ex_event_name iName;
size_t iPoolId;
union
{
pool_job oPoolJob;
job_result oJobResult;
std::string sSocketError;
};
ex_event() { iName = EV_INVALID_VAL; iPoolId = 0;}
ex_event(std::string&& err, size_t id) : iName(EV_SOCK_ERROR), iPoolId(id), sSocketError(std::move(err)) { }
ex_event(job_result dat, size_t id) : iName(EV_MINER_HAVE_RESULT), iPoolId(id), oJobResult(dat) {}
ex_event(pool_job dat, size_t id) : iName(EV_POOL_HAVE_JOB), iPoolId(id), oPoolJob(dat) {}
ex_event(ex_event_name ev, size_t id = 0) : iName(ev), iPoolId(id) {}
// Delete the copy operators to make sure we are moving only what is needed
ex_event(ex_event const&) = delete;
ex_event& operator=(ex_event const&) = delete;
ex_event(ex_event&& from)
{
iName = from.iName;
iPoolId = from.iPoolId;
switch(iName)
{
case EV_SOCK_ERROR:
new (&sSocketError) std::string(std::move(from.sSocketError));
break;
case EV_MINER_HAVE_RESULT:
oJobResult = from.oJobResult;
break;
case EV_POOL_HAVE_JOB:
oPoolJob = from.oPoolJob;
break;
default:
break;
}
}
ex_event& operator=(ex_event&& from)
{
assert(this != &from);
if(iName == EV_SOCK_ERROR)
sSocketError.~basic_string();
iName = from.iName;
iPoolId = from.iPoolId;
switch(iName)
{
case EV_SOCK_ERROR:
new (&sSocketError) std::string();
sSocketError = std::move(from.sSocketError);
break;
case EV_MINER_HAVE_RESULT:
oJobResult = from.oJobResult;
break;
case EV_POOL_HAVE_JOB:
oPoolJob = from.oPoolJob;
break;
default:
break;
}
return *this;
}
~ex_event()
{
if(iName == EV_SOCK_ERROR)
sSocketError.~basic_string();
}
};

96
opencl/blake256.cl Normal file
View File

@ -0,0 +1,96 @@
/*
* blake256 kernel implementation.
*
* ==========================(LICENSE BEGIN)============================
* Copyright (c) 2014 djm34
* Copyright (c) 2014 tpruvot
* 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.
*
* ===========================(LICENSE END)=============================
*
* @author djm34
*/
__constant static const int sigma[16][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 },
{ 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 }
};
__constant static const sph_u32 c_IV256[8] = {
0x6A09E667, 0xBB67AE85,
0x3C6EF372, 0xA54FF53A,
0x510E527F, 0x9B05688C,
0x1F83D9AB, 0x5BE0CD19
};
/* Second part (64-80) msg never change, store it */
__constant static const sph_u32 c_Padding[16] = {
0, 0, 0, 0,
0x80000000, 0, 0, 0,
0, 0, 0, 0,
0, 1, 0, 640,
};
__constant static const sph_u32 c_u256[16] = {
0x243F6A88, 0x85A308D3,
0x13198A2E, 0x03707344,
0xA4093822, 0x299F31D0,
0x082EFA98, 0xEC4E6C89,
0x452821E6, 0x38D01377,
0xBE5466CF, 0x34E90C6C,
0xC0AC29B7, 0xC97C50DD,
0x3F84D5B5, 0xB5470917
};
#define GS(a,b,c,d,x) { \
const sph_u32 idx1 = sigma[r][x]; \
const sph_u32 idx2 = sigma[r][x+1]; \
v[a] += (m[idx1] ^ c_u256[idx2]) + v[b]; \
v[d] ^= v[a]; \
v[d] = rotate(v[d], 16U); \
v[c] += v[d]; \
v[b] ^= v[c]; \
v[b] = rotate(v[b], 20U); \
\
v[a] += (m[idx2] ^ c_u256[idx1]) + v[b]; \
v[d] ^= v[a]; \
v[d] = rotate(v[d], 24U); \
v[c] += v[d]; \
v[b] ^= v[c]; \
v[b] = rotate(v[b], 25U); \
}

912
opencl/cryptonight.cl Normal file
View File

@ -0,0 +1,912 @@
/*
* 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
* 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 <http://www.gnu.org/licenses/>.
*/
#pragma OPENCL EXTENSION cl_amd_media_ops2 : enable
#include "opencl/wolf-aes.cl"
#include "opencl/wolf-skein.cl"
#include "opencl/jh.cl"
#include "opencl/blake256.cl"
#include "opencl/groestl256.cl"
static const __constant ulong keccakf_rndc[24] =
{
0x0000000000000001, 0x0000000000008082, 0x800000000000808a,
0x8000000080008000, 0x000000000000808b, 0x0000000080000001,
0x8000000080008081, 0x8000000000008009, 0x000000000000008a,
0x0000000000000088, 0x0000000080008009, 0x000000008000000a,
0x000000008000808b, 0x800000000000008b, 0x8000000000008089,
0x8000000000008003, 0x8000000000008002, 0x8000000000000080,
0x000000000000800a, 0x800000008000000a, 0x8000000080008081,
0x8000000000008080, 0x0000000080000001, 0x8000000080008008
};
static const __constant uchar sbox[256] =
{
0x63, 0x7C, 0x77, 0x7B, 0xF2, 0x6B, 0x6F, 0xC5, 0x30, 0x01, 0x67, 0x2B, 0xFE, 0xD7, 0xAB, 0x76,
0xCA, 0x82, 0xC9, 0x7D, 0xFA, 0x59, 0x47, 0xF0, 0xAD, 0xD4, 0xA2, 0xAF, 0x9C, 0xA4, 0x72, 0xC0,
0xB7, 0xFD, 0x93, 0x26, 0x36, 0x3F, 0xF7, 0xCC, 0x34, 0xA5, 0xE5, 0xF1, 0x71, 0xD8, 0x31, 0x15,
0x04, 0xC7, 0x23, 0xC3, 0x18, 0x96, 0x05, 0x9A, 0x07, 0x12, 0x80, 0xE2, 0xEB, 0x27, 0xB2, 0x75,
0x09, 0x83, 0x2C, 0x1A, 0x1B, 0x6E, 0x5A, 0xA0, 0x52, 0x3B, 0xD6, 0xB3, 0x29, 0xE3, 0x2F, 0x84,
0x53, 0xD1, 0x00, 0xED, 0x20, 0xFC, 0xB1, 0x5B, 0x6A, 0xCB, 0xBE, 0x39, 0x4A, 0x4C, 0x58, 0xCF,
0xD0, 0xEF, 0xAA, 0xFB, 0x43, 0x4D, 0x33, 0x85, 0x45, 0xF9, 0x02, 0x7F, 0x50, 0x3C, 0x9F, 0xA8,
0x51, 0xA3, 0x40, 0x8F, 0x92, 0x9D, 0x38, 0xF5, 0xBC, 0xB6, 0xDA, 0x21, 0x10, 0xFF, 0xF3, 0xD2,
0xCD, 0x0C, 0x13, 0xEC, 0x5F, 0x97, 0x44, 0x17, 0xC4, 0xA7, 0x7E, 0x3D, 0x64, 0x5D, 0x19, 0x73,
0x60, 0x81, 0x4F, 0xDC, 0x22, 0x2A, 0x90, 0x88, 0x46, 0xEE, 0xB8, 0x14, 0xDE, 0x5E, 0x0B, 0xDB,
0xE0, 0x32, 0x3A, 0x0A, 0x49, 0x06, 0x24, 0x5C, 0xC2, 0xD3, 0xAC, 0x62, 0x91, 0x95, 0xE4, 0x79,
0xE7, 0xC8, 0x37, 0x6D, 0x8D, 0xD5, 0x4E, 0xA9, 0x6C, 0x56, 0xF4, 0xEA, 0x65, 0x7A, 0xAE, 0x08,
0xBA, 0x78, 0x25, 0x2E, 0x1C, 0xA6, 0xB4, 0xC6, 0xE8, 0xDD, 0x74, 0x1F, 0x4B, 0xBD, 0x8B, 0x8A,
0x70, 0x3E, 0xB5, 0x66, 0x48, 0x03, 0xF6, 0x0E, 0x61, 0x35, 0x57, 0xB9, 0x86, 0xC1, 0x1D, 0x9E,
0xE1, 0xF8, 0x98, 0x11, 0x69, 0xD9, 0x8E, 0x94, 0x9B, 0x1E, 0x87, 0xE9, 0xCE, 0x55, 0x28, 0xDF,
0x8C, 0xA1, 0x89, 0x0D, 0xBF, 0xE6, 0x42, 0x68, 0x41, 0x99, 0x2D, 0x0F, 0xB0, 0x54, 0xBB, 0x16
};
void keccakf1600(ulong *s)
{
for(int i = 0; i < 24; ++i)
{
ulong bc[5], tmp1, tmp2;
bc[0] = s[0] ^ s[5] ^ s[10] ^ s[15] ^ s[20] ^ rotate(s[2] ^ s[7] ^ s[12] ^ s[17] ^ s[22], 1UL);
bc[1] = s[1] ^ s[6] ^ s[11] ^ s[16] ^ s[21] ^ rotate(s[3] ^ s[8] ^ s[13] ^ s[18] ^ s[23], 1UL);
bc[2] = s[2] ^ s[7] ^ s[12] ^ s[17] ^ s[22] ^ rotate(s[4] ^ s[9] ^ s[14] ^ s[19] ^ s[24], 1UL);
bc[3] = s[3] ^ s[8] ^ s[13] ^ s[18] ^ s[23] ^ rotate(s[0] ^ s[5] ^ s[10] ^ s[15] ^ s[20], 1UL);
bc[4] = s[4] ^ s[9] ^ s[14] ^ s[19] ^ s[24] ^ rotate(s[1] ^ s[6] ^ s[11] ^ s[16] ^ s[21], 1UL);
tmp1 = s[1] ^ bc[0];
s[0] ^= bc[4];
s[1] = rotate(s[6] ^ bc[0], 44UL);
s[6] = rotate(s[9] ^ bc[3], 20UL);
s[9] = rotate(s[22] ^ bc[1], 61UL);
s[22] = rotate(s[14] ^ bc[3], 39UL);
s[14] = rotate(s[20] ^ bc[4], 18UL);
s[20] = rotate(s[2] ^ bc[1], 62UL);
s[2] = rotate(s[12] ^ bc[1], 43UL);
s[12] = rotate(s[13] ^ bc[2], 25UL);
s[13] = rotate(s[19] ^ bc[3], 8UL);
s[19] = rotate(s[23] ^ bc[2], 56UL);
s[23] = rotate(s[15] ^ bc[4], 41UL);
s[15] = rotate(s[4] ^ bc[3], 27UL);
s[4] = rotate(s[24] ^ bc[3], 14UL);
s[24] = rotate(s[21] ^ bc[0], 2UL);
s[21] = rotate(s[8] ^ bc[2], 55UL);
s[8] = rotate(s[16] ^ bc[0], 35UL);
s[16] = rotate(s[5] ^ bc[4], 36UL);
s[5] = rotate(s[3] ^ bc[2], 28UL);
s[3] = rotate(s[18] ^ bc[2], 21UL);
s[18] = rotate(s[17] ^ bc[1], 15UL);
s[17] = rotate(s[11] ^ bc[0], 10UL);
s[11] = rotate(s[7] ^ bc[1], 6UL);
s[7] = rotate(s[10] ^ bc[4], 3UL);
s[10] = rotate(tmp1, 1UL);
tmp1 = s[0]; tmp2 = s[1]; s[0] = bitselect(s[0] ^ s[2], s[0], s[1]); s[1] = bitselect(s[1] ^ s[3], s[1], s[2]); s[2] = bitselect(s[2] ^ s[4], s[2], s[3]); s[3] = bitselect(s[3] ^ tmp1, s[3], s[4]); s[4] = bitselect(s[4] ^ tmp2, s[4], tmp1);
tmp1 = s[5]; tmp2 = s[6]; s[5] = bitselect(s[5] ^ s[7], s[5], s[6]); s[6] = bitselect(s[6] ^ s[8], s[6], s[7]); s[7] = bitselect(s[7] ^ s[9], s[7], s[8]); s[8] = bitselect(s[8] ^ tmp1, s[8], s[9]); s[9] = bitselect(s[9] ^ tmp2, s[9], tmp1);
tmp1 = s[10]; tmp2 = s[11]; s[10] = bitselect(s[10] ^ s[12], s[10], s[11]); s[11] = bitselect(s[11] ^ s[13], s[11], s[12]); s[12] = bitselect(s[12] ^ s[14], s[12], s[13]); s[13] = bitselect(s[13] ^ tmp1, s[13], s[14]); s[14] = bitselect(s[14] ^ tmp2, s[14], tmp1);
tmp1 = s[15]; tmp2 = s[16]; s[15] = bitselect(s[15] ^ s[17], s[15], s[16]); s[16] = bitselect(s[16] ^ s[18], s[16], s[17]); s[17] = bitselect(s[17] ^ s[19], s[17], s[18]); s[18] = bitselect(s[18] ^ tmp1, s[18], s[19]); s[19] = bitselect(s[19] ^ tmp2, s[19], tmp1);
tmp1 = s[20]; tmp2 = s[21]; s[20] = bitselect(s[20] ^ s[22], s[20], s[21]); s[21] = bitselect(s[21] ^ s[23], s[21], s[22]); s[22] = bitselect(s[22] ^ s[24], s[22], s[23]); s[23] = bitselect(s[23] ^ tmp1, s[23], s[24]); s[24] = bitselect(s[24] ^ tmp2, s[24], tmp1);
s[0] ^= keccakf_rndc[i];
}
}
static const __constant uint keccakf_rotc[24] =
{
1, 3, 6, 10, 15, 21, 28, 36, 45, 55, 2, 14,
27, 41, 56, 8, 25, 43, 62, 18, 39, 61, 20, 44
};
static const __constant uint keccakf_piln[24] =
{
10, 7, 11, 17, 18, 3, 5, 16, 8, 21, 24, 4,
15, 23, 19, 13, 12, 2, 20, 14, 22, 9, 6, 1
};
void keccakf1600_1(ulong *st)
{
int i, round;
ulong t, bc[5];
#pragma unroll 1
for(round = 0; round < 24; ++round)
{
// Theta
bc[0] = st[0] ^ st[5] ^ st[10] ^ st[15] ^ st[20];
bc[1] = st[1] ^ st[6] ^ st[11] ^ st[16] ^ st[21];
bc[2] = st[2] ^ st[7] ^ st[12] ^ st[17] ^ st[22];
bc[3] = st[3] ^ st[8] ^ st[13] ^ st[18] ^ st[23];
bc[4] = st[4] ^ st[9] ^ st[14] ^ st[19] ^ st[24];
#pragma unroll 1
for (i = 0; i < 5; ++i) {
t = bc[(i + 4) % 5] ^ rotate(bc[(i + 1) % 5], 1UL);
st[i ] ^= t;
st[i + 5] ^= t;
st[i + 10] ^= t;
st[i + 15] ^= t;
st[i + 20] ^= t;
}
// Rho Pi
t = st[1];
#pragma unroll
for (i = 0; i < 24; ++i) {
bc[0] = st[keccakf_piln[i]];
st[keccakf_piln[i]] = rotate(t, (ulong)keccakf_rotc[i]);
t = bc[0];
}
//ulong tmp1 = st[0]; ulong tmp2 = st[1]; st[0] = bitselect(st[0] ^ st[2], st[0], st[1]); st[1] = bitselect(st[1] ^ st[3], st[1], st[2]); st[2] = bitselect(st[2] ^ st[4], st[2], st[3]); st[3] = bitselect(st[3] ^ tmp1, st[3], st[4]); st[4] = bitselect(st[4] ^ tmp2, st[4], tmp1);
//tmp1 = st[5]; tmp2 = st[6]; st[5] = bitselect(st[5] ^ st[7], st[5], st[6]); st[6] = bitselect(st[6] ^ st[8], st[6], st[7]); st[7] = bitselect(st[7] ^ st[9], st[7], st[8]); st[8] = bitselect(st[8] ^ tmp1, st[8], st[9]); st[9] = bitselect(st[9] ^ tmp2, st[9], tmp1);
//tmp1 = st[10]; tmp2 = st[11]; st[10] = bitselect(st[10] ^ st[12], st[10], st[11]); st[11] = bitselect(st[11] ^ st[13], st[11], st[12]); st[12] = bitselect(st[12] ^ st[14], st[12], st[13]); st[13] = bitselect(st[13] ^ tmp1, st[13], st[14]); st[14] = bitselect(st[14] ^ tmp2, st[14], tmp1);
//tmp1 = st[15]; tmp2 = st[16]; st[15] = bitselect(st[15] ^ st[17], st[15], st[16]); st[16] = bitselect(st[16] ^ st[18], st[16], st[17]); st[17] = bitselect(st[17] ^ st[19], st[17], st[18]); st[18] = bitselect(st[18] ^ tmp1, st[18], st[19]); st[19] = bitselect(st[19] ^ tmp2, st[19], tmp1);
//tmp1 = st[20]; tmp2 = st[21]; st[20] = bitselect(st[20] ^ st[22], st[20], st[21]); st[21] = bitselect(st[21] ^ st[23], st[21], st[22]); st[22] = bitselect(st[22] ^ st[24], st[22], st[23]); st[23] = bitselect(st[23] ^ tmp1, st[23], st[24]); st[24] = bitselect(st[24] ^ tmp2, st[24], tmp1);
#pragma unroll 1
for(int i = 0; i < 25; i += 5)
{
ulong tmp[5];
#pragma unroll 1
for(int x = 0; x < 5; ++x)
tmp[x] = bitselect(st[i + x] ^ st[i + ((x + 2) % 5)], st[i + x], st[i + ((x + 1) % 5)]);
#pragma unroll 1
for(int x = 0; x < 5; ++x) st[i + x] = tmp[x];
}
// Iota
st[0] ^= keccakf_rndc[round];
}
}
void keccakf1600_2(ulong *st)
{
int i, round;
ulong t, bc[5];
#pragma unroll 1
for(round = 0; round < 24; ++round)
{
// Theta
//bc[0] = st[0] ^ st[5] ^ st[10] ^ st[15] ^ st[20];
//bc[1] = st[1] ^ st[6] ^ st[11] ^ st[16] ^ st[21];
//bc[2] = st[2] ^ st[7] ^ st[12] ^ st[17] ^ st[22];
//bc[3] = st[3] ^ st[8] ^ st[13] ^ st[18] ^ st[23];
//bc[4] = st[4] ^ st[9] ^ st[14] ^ st[19] ^ st[24];
/*
#pragma unroll
for (i = 0; i < 5; ++i) {
t = bc[(i + 4) % 5] ^ rotate(bc[(i + 1) % 5], 1UL);
st[i ] ^= t;
st[i + 5] ^= t;
st[i + 10] ^= t;
st[i + 15] ^= t;
st[i + 20] ^= t;
}
*/
bc[0] = st[0] ^ st[5] ^ st[10] ^ st[15] ^ st[20] ^ rotate(st[2] ^ st[7] ^ st[12] ^ st[17] ^ st[22], 1UL);
bc[1] = st[1] ^ st[6] ^ st[11] ^ st[16] ^ st[21] ^ rotate(st[3] ^ st[8] ^ st[13] ^ st[18] ^ st[23], 1UL);
bc[2] = st[2] ^ st[7] ^ st[12] ^ st[17] ^ st[22] ^ rotate(st[4] ^ st[9] ^ st[14] ^ st[19] ^ st[24], 1UL);
bc[3] = st[3] ^ st[8] ^ st[13] ^ st[18] ^ st[23] ^ rotate(st[0] ^ st[5] ^ st[10] ^ st[15] ^ st[20], 1UL);
bc[4] = st[4] ^ st[9] ^ st[14] ^ st[19] ^ st[24] ^ rotate(st[1] ^ st[6] ^ st[11] ^ st[16] ^ st[21], 1UL);
st[0] ^= bc[4];
st[5] ^= bc[4];
st[10] ^= bc[4];
st[15] ^= bc[4];
st[20] ^= bc[4];
st[1] ^= bc[0];
st[6] ^= bc[0];
st[11] ^= bc[0];
st[16] ^= bc[0];
st[21] ^= bc[0];
st[2] ^= bc[1];
st[7] ^= bc[1];
st[12] ^= bc[1];
st[17] ^= bc[1];
st[22] ^= bc[1];
st[3] ^= bc[2];
st[8] ^= bc[2];
st[13] ^= bc[2];
st[18] ^= bc[2];
st[23] ^= bc[2];
st[4] ^= bc[3];
st[9] ^= bc[3];
st[14] ^= bc[3];
st[19] ^= bc[3];
st[24] ^= bc[3];
// Rho Pi
t = st[1];
#pragma unroll
for (i = 0; i < 24; ++i) {
bc[0] = st[keccakf_piln[i]];
st[keccakf_piln[i]] = rotate(t, (ulong)keccakf_rotc[i]);
t = bc[0];
}
/*ulong tmp1 = st[1] ^ bc[0];
st[0] ^= bc[4];
st[1] = rotate(st[6] ^ bc[0], 44UL);
st[6] = rotate(st[9] ^ bc[3], 20UL);
st[9] = rotate(st[22] ^ bc[1], 61UL);
st[22] = rotate(st[14] ^ bc[3], 39UL);
st[14] = rotate(st[20] ^ bc[4], 18UL);
st[20] = rotate(st[2] ^ bc[1], 62UL);
st[2] = rotate(st[12] ^ bc[1], 43UL);
st[12] = rotate(st[13] ^ bc[2], 25UL);
st[13] = rotate(st[19] ^ bc[3], 8UL);
st[19] = rotate(st[23] ^ bc[2], 56UL);
st[23] = rotate(st[15] ^ bc[4], 41UL);
st[15] = rotate(st[4] ^ bc[3], 27UL);
st[4] = rotate(st[24] ^ bc[3], 14UL);
st[24] = rotate(st[21] ^ bc[0], 2UL);
st[21] = rotate(st[8] ^ bc[2], 55UL);
st[8] = rotate(st[16] ^ bc[0], 35UL);
st[16] = rotate(st[5] ^ bc[4], 36UL);
st[5] = rotate(st[3] ^ bc[2], 28UL);
st[3] = rotate(st[18] ^ bc[2], 21UL);
st[18] = rotate(st[17] ^ bc[1], 15UL);
st[17] = rotate(st[11] ^ bc[0], 10UL);
st[11] = rotate(st[7] ^ bc[1], 6UL);
st[7] = rotate(st[10] ^ bc[4], 3UL);
st[10] = rotate(tmp1, 1UL);
*/
//ulong tmp1 = st[0]; ulong tmp2 = st[1]; st[0] = bitselect(st[0] ^ st[2], st[0], st[1]); st[1] = bitselect(st[1] ^ st[3], st[1], st[2]); st[2] = bitselect(st[2] ^ st[4], st[2], st[3]); st[3] = bitselect(st[3] ^ tmp1, st[3], st[4]); st[4] = bitselect(st[4] ^ tmp2, st[4], tmp1);
//tmp1 = st[5]; tmp2 = st[6]; st[5] = bitselect(st[5] ^ st[7], st[5], st[6]); st[6] = bitselect(st[6] ^ st[8], st[6], st[7]); st[7] = bitselect(st[7] ^ st[9], st[7], st[8]); st[8] = bitselect(st[8] ^ tmp1, st[8], st[9]); st[9] = bitselect(st[9] ^ tmp2, st[9], tmp1);
//tmp1 = st[10]; tmp2 = st[11]; st[10] = bitselect(st[10] ^ st[12], st[10], st[11]); st[11] = bitselect(st[11] ^ st[13], st[11], st[12]); st[12] = bitselect(st[12] ^ st[14], st[12], st[13]); st[13] = bitselect(st[13] ^ tmp1, st[13], st[14]); st[14] = bitselect(st[14] ^ tmp2, st[14], tmp1);
//tmp1 = st[15]; tmp2 = st[16]; st[15] = bitselect(st[15] ^ st[17], st[15], st[16]); st[16] = bitselect(st[16] ^ st[18], st[16], st[17]); st[17] = bitselect(st[17] ^ st[19], st[17], st[18]); st[18] = bitselect(st[18] ^ tmp1, st[18], st[19]); st[19] = bitselect(st[19] ^ tmp2, st[19], tmp1);
//tmp1 = st[20]; tmp2 = st[21]; st[20] = bitselect(st[20] ^ st[22], st[20], st[21]); st[21] = bitselect(st[21] ^ st[23], st[21], st[22]); st[22] = bitselect(st[22] ^ st[24], st[22], st[23]); st[23] = bitselect(st[23] ^ tmp1, st[23], st[24]); st[24] = bitselect(st[24] ^ tmp2, st[24], tmp1);
#pragma unroll
for(int i = 0; i < 25; i += 5)
{
ulong tmp1 = st[i], tmp2 = st[i + 1];
st[i] = bitselect(st[i] ^ st[i + 2], st[i], st[i + 1]);
st[i + 1] = bitselect(st[i + 1] ^ st[i + 3], st[i + 1], st[i + 2]);
st[i + 2] = bitselect(st[i + 2] ^ st[i + 4], st[i + 2], st[i + 3]);
st[i + 3] = bitselect(st[i + 3] ^ tmp1, st[i + 3], st[i + 4]);
st[i + 4] = bitselect(st[i + 4] ^ tmp2, st[i + 4], tmp1);
}
// Iota
st[0] ^= keccakf_rndc[round];
}
}
void CNKeccak(ulong *output, ulong *input)
{
ulong st[25];
// Copy 72 bytes
for(int i = 0; i < 9; ++i) st[i] = input[i];
// Last four and '1' bit for padding
//st[9] = as_ulong((uint2)(((uint *)input)[18], 0x00000001U));
st[9] = (input[9] & 0x00000000FFFFFFFFUL) | 0x0000000100000000UL;
for(int i = 10; i < 25; ++i) st[i] = 0x00UL;
// Last bit of padding
st[16] = 0x8000000000000000UL;
keccakf1600_1(st);
for(int i = 0; i < 25; ++i) output[i] = st[i];
}
static const __constant uchar rcon[8] = { 0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40 };
#pragma OPENCL EXTENSION cl_amd_media_ops2 : enable
#define BYTE(x, y) (amd_bfe((x), (y) << 3U, 8U))
#define SubWord(inw) ((sbox[BYTE(inw, 3)] << 24) | (sbox[BYTE(inw, 2)] << 16) | (sbox[BYTE(inw, 1)] << 8) | sbox[BYTE(inw, 0)])
void AESExpandKey256(uint *keybuf)
{
//#pragma unroll 4
for(uint c = 8, i = 1; c < 60; ++c)
{
// For 256-bit keys, an sbox permutation is done every other 4th uint generated, AND every 8th
uint t = ((!(c & 7)) || ((c & 7) == 4)) ? SubWord(keybuf[c - 1]) : keybuf[c - 1];
// If the uint we're generating has an index that is a multiple of 8, rotate and XOR with the round constant,
// then XOR this with previously generated uint. If it's 4 after a multiple of 8, only the sbox permutation
// is done, followed by the XOR. If neither are true, only the XOR with the previously generated uint is done.
keybuf[c] = keybuf[c - 8] ^ ((!(c & 7)) ? rotate(t, 24U) ^ as_uint((uchar4)(rcon[i++], 0U, 0U, 0U)) : t);
}
}
#define IDX(x) ((x) * (get_global_size(0)))
__attribute__((reqd_work_group_size(WORKSIZE, 8, 1)))
__kernel void cn0(__global ulong *input, __global uint4 *Scratchpad, __global ulong *states)
{
ulong State[25];
uint ExpandedKey1[256];
__local uint AES0[256], AES1[256], AES2[256], AES3[256];
uint4 text;
states += (25 * (get_global_id(0) - get_global_offset(0)));
Scratchpad += ((get_global_id(0) - get_global_offset(0)));
for(int i = get_local_id(0); i < 256; i += WORKSIZE)
{
const uint tmp = AES0_C[i];
AES0[i] = tmp;
AES1[i] = rotate(tmp, 8U);
AES2[i] = rotate(tmp, 16U);
AES3[i] = rotate(tmp, 24U);
}
((ulong8 *)State)[0] = vload8(0, input);
State[8] = input[8];
State[9] = input[9];
State[10] = input[10];
((uint *)State)[9] &= 0x00FFFFFFU;
((uint *)State)[9] |= ((get_global_id(0)) & 0xFF) << 24;
((uint *)State)[10] &= 0xFF000000U;
((uint *)State)[10] |= ((get_global_id(0) >> 8));
for(int i = 11; i < 25; ++i) State[i] = 0x00UL;
// Last bit of padding
State[16] = 0x8000000000000000UL;
keccakf1600_2(State);
mem_fence(CLK_GLOBAL_MEM_FENCE);
#pragma unroll
for(int i = 0; i < 25; ++i) states[i] = State[i];
text = vload4(get_local_id(1) + 4, (__global uint *)(states));
#pragma unroll
for(int i = 0; i < 4; ++i) ((ulong *)ExpandedKey1)[i] = states[i];
AESExpandKey256(ExpandedKey1);
mem_fence(CLK_LOCAL_MEM_FENCE);
#pragma unroll 2
for(int i = 0; i < 0x4000; ++i)
{
#pragma unroll
for(int j = 0; j < 10; ++j)
text = AES_Round(AES0, AES1, AES2, AES3, text, ((uint4 *)ExpandedKey1)[j]);
Scratchpad[IDX((i << 3) + get_local_id(1))] = text;
}
mem_fence(CLK_GLOBAL_MEM_FENCE);
}
__attribute__((reqd_work_group_size(WORKSIZE, 1, 1)))
__kernel void cn1(__global uint4 *Scratchpad, __global ulong *states)
{
ulong a[2], b[2];
__local uint AES0[256], AES1[256], AES2[256], AES3[256];
Scratchpad += ((get_global_id(0) - get_global_offset(0)));
states += (25 * (get_global_id(0) - get_global_offset(0)));
for(int i = get_local_id(0); i < 256; i += WORKSIZE)
{
const uint tmp = AES0_C[i];
AES0[i] = tmp;
AES1[i] = rotate(tmp, 8U);
AES2[i] = rotate(tmp, 16U);
AES3[i] = rotate(tmp, 24U);
}
a[0] = states[0] ^ states[4];
b[0] = states[2] ^ states[6];
a[1] = states[1] ^ states[5];
b[1] = states[3] ^ states[7];
uint4 b_x = ((uint4 *)b)[0];
mem_fence(CLK_LOCAL_MEM_FENCE);
#pragma unroll 8
for(int i = 0; i < 0x80000; ++i)
{
ulong c[2];
((uint4 *)c)[0] = Scratchpad[IDX((a[0] & 0x1FFFF0) >> 4)];
((uint4 *)c)[0] = AES_Round(AES0, AES1, AES2, AES3, ((uint4 *)c)[0], ((uint4 *)a)[0]);
//b_x ^= ((uint4 *)c)[0];
Scratchpad[IDX((a[0] & 0x1FFFF0) >> 4)] = b_x ^ ((uint4 *)c)[0];
uint4 tmp;
tmp = Scratchpad[IDX((c[0] & 0x1FFFF0) >> 4)];
a[1] += c[0] * as_ulong2(tmp).s0;
a[0] += mul_hi(c[0], as_ulong2(tmp).s0);
Scratchpad[IDX((c[0] & 0x1FFFF0) >> 4)] = ((uint4 *)a)[0];
((uint4 *)a)[0] ^= tmp;
b_x = ((uint4 *)c)[0];
}
mem_fence(CLK_GLOBAL_MEM_FENCE);
}
__attribute__((reqd_work_group_size(WORKSIZE, 8, 1)))
__kernel void cn2(__global uint4 *Scratchpad, __global ulong *states, __global uint *Branch0, __global uint *Branch1, __global uint *Branch2, __global uint *Branch3)
{
__local uint AES0[256], AES1[256], AES2[256], AES3[256];
uint ExpandedKey2[256];
ulong State[25];
uint4 text;
Scratchpad += ((get_global_id(0) - get_global_offset(0)));
states += (25 * (get_global_id(0) - get_global_offset(0)));
for(int i = get_local_id(0); i < 256; i += WORKSIZE)
{
const uint tmp = AES0_C[i];
AES0[i] = tmp;
AES1[i] = rotate(tmp, 8U);
AES2[i] = rotate(tmp, 16U);
AES3[i] = rotate(tmp, 24U);
}
#if defined(__Tahiti__) || defined(__Pitcairn__)
for(int i = 0; i < 4; ++i) ((ulong *)ExpandedKey2)[i] = states[i + 4];
text = vload4(get_local_id(1) + 4, (__global uint *)states);
#else
text = vload4(get_local_id(1) + 4, (__global uint *)states);
((uint8 *)ExpandedKey2)[0] = vload8(1, (__global uint *)states);
#endif
AESExpandKey256(ExpandedKey2);
barrier(CLK_LOCAL_MEM_FENCE);
#pragma unroll 2
for(int i = 0; i < 0x4000; ++i)
{
text ^= Scratchpad[IDX((i << 3) + get_local_id(1))];
#pragma unroll
for(int j = 0; j < 10; ++j)
text = AES_Round(AES0, AES1, AES2, AES3, text, ((uint4 *)ExpandedKey2)[j]);
}
vstore2(as_ulong2(text), get_local_id(1) + 4, states);
barrier(CLK_GLOBAL_MEM_FENCE);
if(!get_local_id(1))
{
for(int i = 0; i < 25; ++i) State[i] = states[i];
keccakf1600_2(State);
for(int i = 0; i < 25; ++i) states[i] = State[i];
switch(State[0] & 3)
{
case 0:
Branch0[atomic_inc(Branch0 + get_global_size(0))] = get_global_id(0) - get_global_offset(0);
break;
case 1:
Branch1[atomic_inc(Branch1 + get_global_size(0))] = get_global_id(0) - get_global_offset(0);
break;
case 2:
Branch2[atomic_inc(Branch2 + get_global_size(0))] = get_global_id(0) - get_global_offset(0);
break;
case 3:
Branch3[atomic_inc(Branch3 + get_global_size(0))] = get_global_id(0) - get_global_offset(0);
break;
}
}
mem_fence(CLK_GLOBAL_MEM_FENCE);
}
/*
__kernel void cryptonight(__global ulong *input, __global uint4 *Scratchpad, __global ulong *states, __global uint *Branch0, __global uint *Branch1, __global uint *Branch2, __global uint *Branch3, ulong ThreadCount)
{
uchar State[200];
__local uint AES0[256], AES1[256], AES2[256], AES3[256];
uchar ExpandedKey1[256], ExpandedKey2[256];
ulong inbuf[10], a[2], b[2];
uint4 text[8];
for(int i = 0; i < 256; ++i)
{
const uint tmp = AES0_C[i];
AES0[i] = tmp;
AES1[i] = rotate(tmp, 8U);
AES2[i] = rotate(tmp, 16U);
AES3[i] = rotate(tmp, 24U);
}
((ulong8 *)inbuf)[0] = vload8(0, input);
inbuf[8] = input[8];
inbuf[9] = (ulong)((__global uint *)input)[18];
((uint *)(((uchar *)inbuf) + 39))[0] = get_global_id(0);
CNKeccak((ulong *)State, inbuf);
a[0] = ((ulong *)State)[0] ^ ((ulong *)State)[4];
b[0] = ((ulong *)State)[2] ^ ((ulong *)State)[6];
a[1] = ((ulong *)State)[1] ^ ((ulong *)State)[5];
b[1] = ((ulong *)State)[3] ^ ((ulong *)State)[7];
for(uint i = 0; i < 8; ++i) text[i] = vload4(i + 4, (uint *)(State));
for(int i = 0; i < 4; ++i) ((ulong *)ExpandedKey1)[i] = ((ulong *)State)[i];
for(int i = 0; i < 4; ++i) ((ulong *)ExpandedKey2)[i] = ((ulong *)State)[i + 4];
AESExpandKey256(ExpandedKey1);
AESExpandKey256(ExpandedKey2);
mem_fence(CLK_LOCAL_MEM_FENCE);
Scratchpad += ((1 << 17) * (get_global_id(0) - get_global_offset(0)));
//#pragma unroll 1
for(int i = 0; i < (1 << 17); i += 8)
{
#pragma unroll
for(int j = 0; j < 10; ++j)
{
#pragma unroll
for(int x = 0; x < 8; ++x)
text[x] = AES_Round(AES0, AES1, AES2, AES3, text[x], ((uint4 *)ExpandedKey1)[j]);
}
for(int j = 0; j < 8; ++j) *(Scratchpad + i + j) = text[j];
}
uint4 b_x = ((uint4 *)b)[0];
//#pragma unroll 1
for(int i = 0; i < 0x80000; ++i)
{
ulong c[2];
((uint4 *)c)[0] = Scratchpad[(a[0] & 0x1FFFF0) >> 4];
((uint4 *)c)[0] = AES_Round(AES0, AES1, AES2, AES3, ((uint4 *)c)[0], ((uint4 *)a)[0]);
b_x ^= ((uint4 *)c)[0];
Scratchpad[(a[0] & 0x1FFFF0) >> 4] = b_x;
uint4 tmp;
tmp = Scratchpad[(c[0] & 0x1FFFF0) >> 4];
a[1] += c[0] * as_ulong2(tmp).s0;
a[0] += mul_hi(c[0], as_ulong2(tmp).s0);
Scratchpad[(c[0] & 0x1FFFF0) >> 4] = ((uint4 *)a)[0];
((uint4 *)a)[0] ^= tmp;
b_x = ((uint4 *)c)[0];
}
for(uint i = 0; i < 8; ++i) text[i] = vload4(i + 4, (uint *)(State));
for(int i = 0; i < (1 << 17); i += 8)
{
#pragma unroll
for(int j = 0; j < 8; ++j) text[j] ^= Scratchpad[i + j];
#pragma unroll 1
for(int j = 0; j < 10; ++j)
{
#pragma unroll
for(int x = 0; x < 8; ++x)
text[x] = AES_Round(AES0, AES1, AES2, AES3, text[x], ((uint4 *)ExpandedKey2)[j]);
}
}
for(uint i = 0; i < 8; ++i) vstore4(text[i], i + 4, (uint *)(State));
keccakf1600((ulong *)State);
states += (25 * (get_global_id(0) - get_global_offset(0)));
for(int i = 0; i < 25; ++i) states[i] = ((ulong *)State)[i];
switch(State[0] & 3)
{
case 0:
Branch0[atomic_inc(Branch0 + ThreadCount)] = get_global_id(0) - get_global_offset(0);
break;
case 1:
Branch1[atomic_inc(Branch1 + ThreadCount)] = get_global_id(0) - get_global_offset(0);
break;
case 2:
Branch2[atomic_inc(Branch2 + ThreadCount)] = get_global_id(0) - get_global_offset(0);
break;
case 3:
Branch3[atomic_inc(Branch3 + ThreadCount)] = get_global_id(0) - get_global_offset(0);
break;
}
}
*/
#define VSWAP8(x) (((x) >> 56) | (((x) >> 40) & 0x000000000000FF00UL) | (((x) >> 24) & 0x0000000000FF0000UL) \
| (((x) >> 8) & 0x00000000FF000000UL) | (((x) << 8) & 0x000000FF00000000UL) \
| (((x) << 24) & 0x0000FF0000000000UL) | (((x) << 40) & 0x00FF000000000000UL) | (((x) << 56) & 0xFF00000000000000UL))
#define VSWAP4(x) ((((x) >> 24) & 0xFFU) | (((x) >> 8) & 0xFF00U) | (((x) << 8) & 0xFF0000U) | (((x) << 24) & 0xFF000000U))
__kernel void Skein(__global ulong *states, __global uint *BranchBuf, __global uint *output, uint Target, ulong Threads)
{
const ulong idx = get_global_id(0) - get_global_offset(0);
if(idx >= Threads) return;
states += 25 * BranchBuf[idx];
// skein
ulong8 h = vload8(0, SKEIN512_256_IV);
// Type field begins with final bit, first bit, then six bits of type; the last 96
// bits are input processed (including in the block to be processed with that tweak)
// The output transform is only one run of UBI, since we need only 256 bits of output
// The tweak for the output transform is Type = Output with the Final bit set
// T[0] for the output is 8, and I don't know why - should be message size...
ulong t[3] = { 0x00UL, 0x7000000000000000UL, 0x00UL };
ulong8 p, m;
for(uint i = 0; i < 4; ++i)
{
if(i < 3) t[0] += 0x40UL;
else t[0] += 0x08UL;
t[2] = t[0] ^ t[1];
m = (i < 3) ? vload8(i, states) : (ulong8)(states[24], 0UL, 0UL, 0UL, 0UL, 0UL, 0UL, 0UL);
const ulong h8 = h.s0 ^ h.s1 ^ h.s2 ^ h.s3 ^ h.s4 ^ h.s5 ^ h.s6 ^ h.s7 ^ SKEIN_KS_PARITY;
p = Skein512Block(m, h, h8, t);
h = m ^ p;
if(i < 2) t[1] = 0x3000000000000000UL;
else t[1] = 0xB000000000000000UL;
}
t[0] = 0x08UL;
t[1] = 0xFF00000000000000UL;
t[2] = t[0] ^ t[1];
p = (ulong8)(0);
const ulong h8 = h.s0 ^ h.s1 ^ h.s2 ^ h.s3 ^ h.s4 ^ h.s5 ^ h.s6 ^ h.s7 ^ SKEIN_KS_PARITY;
p = Skein512Block(p, h, h8, t);
//vstore8(p, 0, output);
if(as_uint16(p).s7 <= Target) output[atomic_inc(output + 0xFF)] = BranchBuf[idx] + get_global_offset(0);
mem_fence(CLK_GLOBAL_MEM_FENCE);
}
#define SWAP8(x) as_ulong(as_uchar8(x).s76543210)
__kernel void JH(__global ulong *states, __global uint *BranchBuf, __global uint *output, uint Target, ulong Threads)
{
const uint idx = get_global_id(0) - get_global_offset(0);
if(idx >= Threads) return;
states += 25 * BranchBuf[idx];
sph_u64 h0h = 0xEBD3202C41A398EBUL, h0l = 0xC145B29C7BBECD92UL, h1h = 0xFAC7D4609151931CUL, h1l = 0x038A507ED6820026UL, h2h = 0x45B92677269E23A4UL, h2l = 0x77941AD4481AFBE0UL, h3h = 0x7A176B0226ABB5CDUL, h3l = 0xA82FFF0F4224F056UL;
sph_u64 h4h = 0x754D2E7F8996A371UL, h4l = 0x62E27DF70849141DUL, h5h = 0x948F2476F7957627UL, h5l = 0x6C29804757B6D587UL, h6h = 0x6C0D8EAC2D275E5CUL, h6l = 0x0F7A0557C6508451UL, h7h = 0xEA12247067D3E47BUL, h7l = 0x69D71CD313ABE389UL;
sph_u64 tmp;
for(int i = 0; i < 5; ++i)
{
ulong input[8];
if(i < 3)
{
for(int x = 0; x < 8; ++x) input[x] = (states[(i << 3) + x]);
}
else if(i == 3)
{
input[0] = (states[24]);
input[1] = 0x80UL;
for(int x = 2; x < 8; ++x) input[x] = 0x00UL;
}
else
{
input[7] = 0x4006000000000000UL;
for(int x = 0; x < 7; ++x) input[x] = 0x00UL;
}
h0h ^= input[0];
h0l ^= input[1];
h1h ^= input[2];
h1l ^= input[3];
h2h ^= input[4];
h2l ^= input[5];
h3h ^= input[6];
h3l ^= input[7];
E8;
h4h ^= input[0];
h4l ^= input[1];
h5h ^= input[2];
h5l ^= input[3];
h6h ^= input[4];
h6l ^= input[5];
h7h ^= input[6];
h7l ^= input[7];
}
//output[0] = h6h;
//output[1] = h6l;
//output[2] = h7h;
//output[3] = h7l;
if(as_uint2(h7l).s1 <= Target) output[atomic_inc(output + 0xFF)] = BranchBuf[idx] + get_global_offset(0);
}
#define SWAP4(x) as_uint(as_uchar4(x).s3210)
__kernel void Blake(__global ulong *states, __global uint *BranchBuf, __global uint *output, uint Target, ulong Threads)
{
const uint idx = get_global_id(0) - get_global_offset(0);
if(idx >= Threads) return;
states += 25 * BranchBuf[idx];
unsigned int m[16];
unsigned int v[16];
uint h[8];
((uint8 *)h)[0] = vload8(0U, c_IV256);
for(uint i = 0, bitlen = 0; i < 4; ++i)
{
if(i < 3)
{
((uint16 *)m)[0] = vload16(i, (__global uint *)states);
for(int i = 0; i < 16; ++i) m[i] = SWAP4(m[i]);
bitlen += 512;
}
else
{
m[0] = SWAP4(((__global uint *)states)[48]);
m[1] = SWAP4(((__global uint *)states)[49]);
m[2] = 0x80000000U;
for(int i = 3; i < 13; ++i) m[i] = 0x00U;
m[13] = 1U;
m[14] = 0U;
m[15] = 0x640;
bitlen += 64;
}
((uint16 *)v)[0].lo = ((uint8 *)h)[0];
((uint16 *)v)[0].hi = vload8(0U, c_u256);
//v[12] ^= (i < 3) ? (i + 1) << 9 : 1600U;
//v[13] ^= (i < 3) ? (i + 1) << 9 : 1600U;
v[12] ^= bitlen;
v[13] ^= bitlen;
for(int r = 0; r < 14; r++)
{
GS(0, 4, 0x8, 0xC, 0x0);
GS(1, 5, 0x9, 0xD, 0x2);
GS(2, 6, 0xA, 0xE, 0x4);
GS(3, 7, 0xB, 0xF, 0x6);
GS(0, 5, 0xA, 0xF, 0x8);
GS(1, 6, 0xB, 0xC, 0xA);
GS(2, 7, 0x8, 0xD, 0xC);
GS(3, 4, 0x9, 0xE, 0xE);
}
((uint8 *)h)[0] ^= ((uint8 *)v)[0] ^ ((uint8 *)v)[1];
}
for(int i = 0; i < 8; ++i) h[i] = SWAP4(h[i]);
//for(int i = 0; i < 4; ++i) output[i] = ((ulong *)h)[i];
if(h[7] <= Target) output[atomic_inc(output + 0xFF)] = BranchBuf[idx] + get_global_offset(0);
}
__kernel void Groestl(__global ulong *states, __global uint *BranchBuf, __global uint *output, uint Target, ulong Threads)
{
const uint idx = get_global_id(0) - get_global_offset(0);
if(idx >= Threads) return;
states += 25 * BranchBuf[idx];
ulong State[8];
for(int i = 0; i < 7; ++i) State[i] = 0UL;
State[7] = 0x0001000000000000UL;
for(uint i = 0; i < 4; ++i)
{
ulong H[8], M[8];
if(i < 3)
{
((ulong8 *)M)[0] = vload8(i, states);
}
else
{
M[0] = states[24];
M[1] = 0x80UL;
for(int x = 2; x < 7; ++x) M[x] = 0UL;
M[7] = 0x0400000000000000UL;
}
for(int x = 0; x < 8; ++x) H[x] = M[x] ^ State[x];
PERM_SMALL_P(H);
PERM_SMALL_Q(M);
for(int x = 0; x < 8; ++x) State[x] ^= H[x] ^ M[x];
}
ulong tmp[8];
for(int i = 0; i < 8; ++i) tmp[i] = State[i];
PERM_SMALL_P(State);
for(int i = 0; i < 8; ++i) State[i] ^= tmp[i];
//for(int i = 0; i < 4; ++i) output[i] = State[i + 4];
if(as_uint2(State[7]).s1 <= Target) output[atomic_inc(output + 0xFF)] = BranchBuf[idx] + get_global_offset(0);
}

289
opencl/groestl256.cl Normal file
View File

@ -0,0 +1,289 @@
/* $Id: groestl.c 260 2011-07-21 01:02:38Z tp $ */
/*
* Groestl256
*
* ==========================(LICENSE BEGIN)============================
* Copyright (c) 2014 djm34
* Copyright (c) 2007-2010 Projet RNRT SAPHIR
*
* 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.
*
* ===========================(LICENSE END)=============================
*
* @author Thomas Pornin <thomas.pornin@cryptolog.com>
*/
#define SPH_C64(x) x
#define SPH_ROTL64(x, y) rotate((x), (ulong)(y))
#define C64e(x) ((SPH_C64(x) >> 56) \
| ((SPH_C64(x) >> 40) & SPH_C64(0x000000000000FF00)) \
| ((SPH_C64(x) >> 24) & SPH_C64(0x0000000000FF0000)) \
| ((SPH_C64(x) >> 8) & SPH_C64(0x00000000FF000000)) \
| ((SPH_C64(x) << 8) & SPH_C64(0x000000FF00000000)) \
| ((SPH_C64(x) << 24) & SPH_C64(0x0000FF0000000000)) \
| ((SPH_C64(x) << 40) & SPH_C64(0x00FF000000000000)) \
| ((SPH_C64(x) << 56) & SPH_C64(0xFF00000000000000)))
#define B64_0(x) ((x) & 0xFF)
#define B64_1(x) (((x) >> 8) & 0xFF)
#define B64_2(x) (((x) >> 16) & 0xFF)
#define B64_3(x) (((x) >> 24) & 0xFF)
#define B64_4(x) (((x) >> 32) & 0xFF)
#define B64_5(x) (((x) >> 40) & 0xFF)
#define B64_6(x) (((x) >> 48) & 0xFF)
#define B64_7(x) ((x) >> 56)
#define R64 SPH_ROTL64
#define PC64(j, r) ((sph_u64)((j) + (r)))
#define QC64(j, r) (((sph_u64)(r) << 56) ^ (~((sph_u64)(j) << 56)))
static const __constant ulong T0_G[] =
{
0xc6a597f4a5f432c6UL, 0xf884eb9784976ff8UL, 0xee99c7b099b05eeeUL, 0xf68df78c8d8c7af6UL,
0xff0de5170d17e8ffUL, 0xd6bdb7dcbddc0ad6UL, 0xdeb1a7c8b1c816deUL, 0x915439fc54fc6d91UL,
0x6050c0f050f09060UL, 0x0203040503050702UL, 0xcea987e0a9e02eceUL, 0x567dac877d87d156UL,
0xe719d52b192bcce7UL, 0xb56271a662a613b5UL, 0x4de69a31e6317c4dUL, 0xec9ac3b59ab559ecUL,
0x8f4505cf45cf408fUL, 0x1f9d3ebc9dbca31fUL, 0x894009c040c04989UL, 0xfa87ef92879268faUL,
0xef15c53f153fd0efUL, 0xb2eb7f26eb2694b2UL, 0x8ec90740c940ce8eUL, 0xfb0bed1d0b1de6fbUL,
0x41ec822fec2f6e41UL, 0xb3677da967a91ab3UL, 0x5ffdbe1cfd1c435fUL, 0x45ea8a25ea256045UL,
0x23bf46dabfdaf923UL, 0x53f7a602f7025153UL, 0xe496d3a196a145e4UL, 0x9b5b2ded5bed769bUL,
0x75c2ea5dc25d2875UL, 0xe11cd9241c24c5e1UL, 0x3dae7ae9aee9d43dUL, 0x4c6a98be6abef24cUL,
0x6c5ad8ee5aee826cUL, 0x7e41fcc341c3bd7eUL, 0xf502f1060206f3f5UL, 0x834f1dd14fd15283UL,
0x685cd0e45ce48c68UL, 0x51f4a207f4075651UL, 0xd134b95c345c8dd1UL, 0xf908e9180818e1f9UL,
0xe293dfae93ae4ce2UL, 0xab734d9573953eabUL, 0x6253c4f553f59762UL, 0x2a3f54413f416b2aUL,
0x080c10140c141c08UL, 0x955231f652f66395UL, 0x46658caf65afe946UL, 0x9d5e21e25ee27f9dUL,
0x3028607828784830UL, 0x37a16ef8a1f8cf37UL, 0x0a0f14110f111b0aUL, 0x2fb55ec4b5c4eb2fUL,
0x0e091c1b091b150eUL, 0x2436485a365a7e24UL, 0x1b9b36b69bb6ad1bUL, 0xdf3da5473d4798dfUL,
0xcd26816a266aa7cdUL, 0x4e699cbb69bbf54eUL, 0x7fcdfe4ccd4c337fUL, 0xea9fcfba9fba50eaUL,
0x121b242d1b2d3f12UL, 0x1d9e3ab99eb9a41dUL, 0x5874b09c749cc458UL, 0x342e68722e724634UL,
0x362d6c772d774136UL, 0xdcb2a3cdb2cd11dcUL, 0xb4ee7329ee299db4UL, 0x5bfbb616fb164d5bUL,
0xa4f65301f601a5a4UL, 0x764decd74dd7a176UL, 0xb76175a361a314b7UL, 0x7dcefa49ce49347dUL,
0x527ba48d7b8ddf52UL, 0xdd3ea1423e429fddUL, 0x5e71bc937193cd5eUL, 0x139726a297a2b113UL,
0xa6f55704f504a2a6UL, 0xb96869b868b801b9UL, 0x0000000000000000UL, 0xc12c99742c74b5c1UL,
0x406080a060a0e040UL, 0xe31fdd211f21c2e3UL, 0x79c8f243c8433a79UL, 0xb6ed772ced2c9ab6UL,
0xd4beb3d9bed90dd4UL, 0x8d4601ca46ca478dUL, 0x67d9ce70d9701767UL, 0x724be4dd4bddaf72UL,
0x94de3379de79ed94UL, 0x98d42b67d467ff98UL, 0xb0e87b23e82393b0UL, 0x854a11de4ade5b85UL,
0xbb6b6dbd6bbd06bbUL, 0xc52a917e2a7ebbc5UL, 0x4fe59e34e5347b4fUL, 0xed16c13a163ad7edUL,
0x86c51754c554d286UL, 0x9ad72f62d762f89aUL, 0x6655ccff55ff9966UL, 0x119422a794a7b611UL,
0x8acf0f4acf4ac08aUL, 0xe910c9301030d9e9UL, 0x0406080a060a0e04UL, 0xfe81e798819866feUL,
0xa0f05b0bf00baba0UL, 0x7844f0cc44ccb478UL, 0x25ba4ad5bad5f025UL, 0x4be3963ee33e754bUL,
0xa2f35f0ef30eaca2UL, 0x5dfeba19fe19445dUL, 0x80c01b5bc05bdb80UL, 0x058a0a858a858005UL,
0x3fad7eecadecd33fUL, 0x21bc42dfbcdffe21UL, 0x7048e0d848d8a870UL, 0xf104f90c040cfdf1UL,
0x63dfc67adf7a1963UL, 0x77c1ee58c1582f77UL, 0xaf75459f759f30afUL, 0x426384a563a5e742UL,
0x2030405030507020UL, 0xe51ad12e1a2ecbe5UL, 0xfd0ee1120e12effdUL, 0xbf6d65b76db708bfUL,
0x814c19d44cd45581UL, 0x1814303c143c2418UL, 0x26354c5f355f7926UL, 0xc32f9d712f71b2c3UL,
0xbee16738e13886beUL, 0x35a26afda2fdc835UL, 0x88cc0b4fcc4fc788UL, 0x2e395c4b394b652eUL,
0x93573df957f96a93UL, 0x55f2aa0df20d5855UL, 0xfc82e39d829d61fcUL, 0x7a47f4c947c9b37aUL,
0xc8ac8befacef27c8UL, 0xbae76f32e73288baUL, 0x322b647d2b7d4f32UL, 0xe695d7a495a442e6UL,
0xc0a09bfba0fb3bc0UL, 0x199832b398b3aa19UL, 0x9ed12768d168f69eUL, 0xa37f5d817f8122a3UL,
0x446688aa66aaee44UL, 0x547ea8827e82d654UL, 0x3bab76e6abe6dd3bUL, 0x0b83169e839e950bUL,
0x8cca0345ca45c98cUL, 0xc729957b297bbcc7UL, 0x6bd3d66ed36e056bUL, 0x283c50443c446c28UL,
0xa779558b798b2ca7UL, 0xbce2633de23d81bcUL, 0x161d2c271d273116UL, 0xad76419a769a37adUL,
0xdb3bad4d3b4d96dbUL, 0x6456c8fa56fa9e64UL, 0x744ee8d24ed2a674UL, 0x141e28221e223614UL,
0x92db3f76db76e492UL, 0x0c0a181e0a1e120cUL, 0x486c90b46cb4fc48UL, 0xb8e46b37e4378fb8UL,
0x9f5d25e75de7789fUL, 0xbd6e61b26eb20fbdUL, 0x43ef862aef2a6943UL, 0xc4a693f1a6f135c4UL,
0x39a872e3a8e3da39UL, 0x31a462f7a4f7c631UL, 0xd337bd5937598ad3UL, 0xf28bff868b8674f2UL,
0xd532b156325683d5UL, 0x8b430dc543c54e8bUL, 0x6e59dceb59eb856eUL, 0xdab7afc2b7c218daUL,
0x018c028f8c8f8e01UL, 0xb16479ac64ac1db1UL, 0x9cd2236dd26df19cUL, 0x49e0923be03b7249UL,
0xd8b4abc7b4c71fd8UL, 0xacfa4315fa15b9acUL, 0xf307fd090709faf3UL, 0xcf25856f256fa0cfUL,
0xcaaf8feaafea20caUL, 0xf48ef3898e897df4UL, 0x47e98e20e9206747UL, 0x1018202818283810UL,
0x6fd5de64d5640b6fUL, 0xf088fb83888373f0UL, 0x4a6f94b16fb1fb4aUL, 0x5c72b8967296ca5cUL,
0x3824706c246c5438UL, 0x57f1ae08f1085f57UL, 0x73c7e652c7522173UL, 0x975135f351f36497UL,
0xcb238d652365aecbUL, 0xa17c59847c8425a1UL, 0xe89ccbbf9cbf57e8UL, 0x3e217c6321635d3eUL,
0x96dd377cdd7cea96UL, 0x61dcc27fdc7f1e61UL, 0x0d861a9186919c0dUL, 0x0f851e9485949b0fUL,
0xe090dbab90ab4be0UL, 0x7c42f8c642c6ba7cUL, 0x71c4e257c4572671UL, 0xccaa83e5aae529ccUL,
0x90d83b73d873e390UL, 0x06050c0f050f0906UL, 0xf701f5030103f4f7UL, 0x1c12383612362a1cUL,
0xc2a39ffea3fe3cc2UL, 0x6a5fd4e15fe18b6aUL, 0xaef94710f910beaeUL, 0x69d0d26bd06b0269UL,
0x17912ea891a8bf17UL, 0x995829e858e87199UL, 0x3a2774692769533aUL, 0x27b94ed0b9d0f727UL,
0xd938a948384891d9UL, 0xeb13cd351335deebUL, 0x2bb356ceb3cee52bUL, 0x2233445533557722UL,
0xd2bbbfd6bbd604d2UL, 0xa9704990709039a9UL, 0x07890e8089808707UL, 0x33a766f2a7f2c133UL,
0x2db65ac1b6c1ec2dUL, 0x3c22786622665a3cUL, 0x15922aad92adb815UL, 0xc92089602060a9c9UL,
0x874915db49db5c87UL, 0xaaff4f1aff1ab0aaUL, 0x5078a0887888d850UL, 0xa57a518e7a8e2ba5UL,
0x038f068a8f8a8903UL, 0x59f8b213f8134a59UL, 0x0980129b809b9209UL, 0x1a1734391739231aUL,
0x65daca75da751065UL, 0xd731b553315384d7UL, 0x84c61351c651d584UL, 0xd0b8bbd3b8d303d0UL,
0x82c31f5ec35edc82UL, 0x29b052cbb0cbe229UL, 0x5a77b4997799c35aUL, 0x1e113c3311332d1eUL,
0x7bcbf646cb463d7bUL, 0xa8fc4b1ffc1fb7a8UL, 0x6dd6da61d6610c6dUL, 0x2c3a584e3a4e622cUL
};
static const __constant ulong T4_G[] =
{
0xA5F432C6C6A597F4UL, 0x84976FF8F884EB97UL, 0x99B05EEEEE99C7B0UL, 0x8D8C7AF6F68DF78CUL,
0x0D17E8FFFF0DE517UL, 0xBDDC0AD6D6BDB7DCUL, 0xB1C816DEDEB1A7C8UL, 0x54FC6D91915439FCUL,
0x50F090606050C0F0UL, 0x0305070202030405UL, 0xA9E02ECECEA987E0UL, 0x7D87D156567DAC87UL,
0x192BCCE7E719D52BUL, 0x62A613B5B56271A6UL, 0xE6317C4D4DE69A31UL, 0x9AB559ECEC9AC3B5UL,
0x45CF408F8F4505CFUL, 0x9DBCA31F1F9D3EBCUL, 0x40C04989894009C0UL, 0x879268FAFA87EF92UL,
0x153FD0EFEF15C53FUL, 0xEB2694B2B2EB7F26UL, 0xC940CE8E8EC90740UL, 0x0B1DE6FBFB0BED1DUL,
0xEC2F6E4141EC822FUL, 0x67A91AB3B3677DA9UL, 0xFD1C435F5FFDBE1CUL, 0xEA25604545EA8A25UL,
0xBFDAF92323BF46DAUL, 0xF702515353F7A602UL, 0x96A145E4E496D3A1UL, 0x5BED769B9B5B2DEDUL,
0xC25D287575C2EA5DUL, 0x1C24C5E1E11CD924UL, 0xAEE9D43D3DAE7AE9UL, 0x6ABEF24C4C6A98BEUL,
0x5AEE826C6C5AD8EEUL, 0x41C3BD7E7E41FCC3UL, 0x0206F3F5F502F106UL, 0x4FD15283834F1DD1UL,
0x5CE48C68685CD0E4UL, 0xF407565151F4A207UL, 0x345C8DD1D134B95CUL, 0x0818E1F9F908E918UL,
0x93AE4CE2E293DFAEUL, 0x73953EABAB734D95UL, 0x53F597626253C4F5UL, 0x3F416B2A2A3F5441UL,
0x0C141C08080C1014UL, 0x52F66395955231F6UL, 0x65AFE94646658CAFUL, 0x5EE27F9D9D5E21E2UL,
0x2878483030286078UL, 0xA1F8CF3737A16EF8UL, 0x0F111B0A0A0F1411UL, 0xB5C4EB2F2FB55EC4UL,
0x091B150E0E091C1BUL, 0x365A7E242436485AUL, 0x9BB6AD1B1B9B36B6UL, 0x3D4798DFDF3DA547UL,
0x266AA7CDCD26816AUL, 0x69BBF54E4E699CBBUL, 0xCD4C337F7FCDFE4CUL, 0x9FBA50EAEA9FCFBAUL,
0x1B2D3F12121B242DUL, 0x9EB9A41D1D9E3AB9UL, 0x749CC4585874B09CUL, 0x2E724634342E6872UL,
0x2D774136362D6C77UL, 0xB2CD11DCDCB2A3CDUL, 0xEE299DB4B4EE7329UL, 0xFB164D5B5BFBB616UL,
0xF601A5A4A4F65301UL, 0x4DD7A176764DECD7UL, 0x61A314B7B76175A3UL, 0xCE49347D7DCEFA49UL,
0x7B8DDF52527BA48DUL, 0x3E429FDDDD3EA142UL, 0x7193CD5E5E71BC93UL, 0x97A2B113139726A2UL,
0xF504A2A6A6F55704UL, 0x68B801B9B96869B8UL, 0x0000000000000000UL, 0x2C74B5C1C12C9974UL,
0x60A0E040406080A0UL, 0x1F21C2E3E31FDD21UL, 0xC8433A7979C8F243UL, 0xED2C9AB6B6ED772CUL,
0xBED90DD4D4BEB3D9UL, 0x46CA478D8D4601CAUL, 0xD970176767D9CE70UL, 0x4BDDAF72724BE4DDUL,
0xDE79ED9494DE3379UL, 0xD467FF9898D42B67UL, 0xE82393B0B0E87B23UL, 0x4ADE5B85854A11DEUL,
0x6BBD06BBBB6B6DBDUL, 0x2A7EBBC5C52A917EUL, 0xE5347B4F4FE59E34UL, 0x163AD7EDED16C13AUL,
0xC554D28686C51754UL, 0xD762F89A9AD72F62UL, 0x55FF99666655CCFFUL, 0x94A7B611119422A7UL,
0xCF4AC08A8ACF0F4AUL, 0x1030D9E9E910C930UL, 0x060A0E040406080AUL, 0x819866FEFE81E798UL,
0xF00BABA0A0F05B0BUL, 0x44CCB4787844F0CCUL, 0xBAD5F02525BA4AD5UL, 0xE33E754B4BE3963EUL,
0xF30EACA2A2F35F0EUL, 0xFE19445D5DFEBA19UL, 0xC05BDB8080C01B5BUL, 0x8A858005058A0A85UL,
0xADECD33F3FAD7EECUL, 0xBCDFFE2121BC42DFUL, 0x48D8A8707048E0D8UL, 0x040CFDF1F104F90CUL,
0xDF7A196363DFC67AUL, 0xC1582F7777C1EE58UL, 0x759F30AFAF75459FUL, 0x63A5E742426384A5UL,
0x3050702020304050UL, 0x1A2ECBE5E51AD12EUL, 0x0E12EFFDFD0EE112UL, 0x6DB708BFBF6D65B7UL,
0x4CD45581814C19D4UL, 0x143C24181814303CUL, 0x355F792626354C5FUL, 0x2F71B2C3C32F9D71UL,
0xE13886BEBEE16738UL, 0xA2FDC83535A26AFDUL, 0xCC4FC78888CC0B4FUL, 0x394B652E2E395C4BUL,
0x57F96A9393573DF9UL, 0xF20D585555F2AA0DUL, 0x829D61FCFC82E39DUL, 0x47C9B37A7A47F4C9UL,
0xACEF27C8C8AC8BEFUL, 0xE73288BABAE76F32UL, 0x2B7D4F32322B647DUL, 0x95A442E6E695D7A4UL,
0xA0FB3BC0C0A09BFBUL, 0x98B3AA19199832B3UL, 0xD168F69E9ED12768UL, 0x7F8122A3A37F5D81UL,
0x66AAEE44446688AAUL, 0x7E82D654547EA882UL, 0xABE6DD3B3BAB76E6UL, 0x839E950B0B83169EUL,
0xCA45C98C8CCA0345UL, 0x297BBCC7C729957BUL, 0xD36E056B6BD3D66EUL, 0x3C446C28283C5044UL,
0x798B2CA7A779558BUL, 0xE23D81BCBCE2633DUL, 0x1D273116161D2C27UL, 0x769A37ADAD76419AUL,
0x3B4D96DBDB3BAD4DUL, 0x56FA9E646456C8FAUL, 0x4ED2A674744EE8D2UL, 0x1E223614141E2822UL,
0xDB76E49292DB3F76UL, 0x0A1E120C0C0A181EUL, 0x6CB4FC48486C90B4UL, 0xE4378FB8B8E46B37UL,
0x5DE7789F9F5D25E7UL, 0x6EB20FBDBD6E61B2UL, 0xEF2A694343EF862AUL, 0xA6F135C4C4A693F1UL,
0xA8E3DA3939A872E3UL, 0xA4F7C63131A462F7UL, 0x37598AD3D337BD59UL, 0x8B8674F2F28BFF86UL,
0x325683D5D532B156UL, 0x43C54E8B8B430DC5UL, 0x59EB856E6E59DCEBUL, 0xB7C218DADAB7AFC2UL,
0x8C8F8E01018C028FUL, 0x64AC1DB1B16479ACUL, 0xD26DF19C9CD2236DUL, 0xE03B724949E0923BUL,
0xB4C71FD8D8B4ABC7UL, 0xFA15B9ACACFA4315UL, 0x0709FAF3F307FD09UL, 0x256FA0CFCF25856FUL,
0xAFEA20CACAAF8FEAUL, 0x8E897DF4F48EF389UL, 0xE920674747E98E20UL, 0x1828381010182028UL,
0xD5640B6F6FD5DE64UL, 0x888373F0F088FB83UL, 0x6FB1FB4A4A6F94B1UL, 0x7296CA5C5C72B896UL,
0x246C54383824706CUL, 0xF1085F5757F1AE08UL, 0xC752217373C7E652UL, 0x51F36497975135F3UL,
0x2365AECBCB238D65UL, 0x7C8425A1A17C5984UL, 0x9CBF57E8E89CCBBFUL, 0x21635D3E3E217C63UL,
0xDD7CEA9696DD377CUL, 0xDC7F1E6161DCC27FUL, 0x86919C0D0D861A91UL, 0x85949B0F0F851E94UL,
0x90AB4BE0E090DBABUL, 0x42C6BA7C7C42F8C6UL, 0xC457267171C4E257UL, 0xAAE529CCCCAA83E5UL,
0xD873E39090D83B73UL, 0x050F090606050C0FUL, 0x0103F4F7F701F503UL, 0x12362A1C1C123836UL,
0xA3FE3CC2C2A39FFEUL, 0x5FE18B6A6A5FD4E1UL, 0xF910BEAEAEF94710UL, 0xD06B026969D0D26BUL,
0x91A8BF1717912EA8UL, 0x58E87199995829E8UL, 0x2769533A3A277469UL, 0xB9D0F72727B94ED0UL,
0x384891D9D938A948UL, 0x1335DEEBEB13CD35UL, 0xB3CEE52B2BB356CEUL, 0x3355772222334455UL,
0xBBD604D2D2BBBFD6UL, 0x709039A9A9704990UL, 0x8980870707890E80UL, 0xA7F2C13333A766F2UL,
0xB6C1EC2D2DB65AC1UL, 0x22665A3C3C227866UL, 0x92ADB81515922AADUL, 0x2060A9C9C9208960UL,
0x49DB5C87874915DBUL, 0xFF1AB0AAAAFF4F1AUL, 0x7888D8505078A088UL, 0x7A8E2BA5A57A518EUL,
0x8F8A8903038F068AUL, 0xF8134A5959F8B213UL, 0x809B92090980129BUL, 0x1739231A1A173439UL,
0xDA75106565DACA75UL, 0x315384D7D731B553UL, 0xC651D58484C61351UL, 0xB8D303D0D0B8BBD3UL,
0xC35EDC8282C31F5EUL, 0xB0CBE22929B052CBUL, 0x7799C35A5A77B499UL, 0x11332D1E1E113C33UL,
0xCB463D7B7BCBF646UL, 0xFC1FB7A8A8FC4B1FUL, 0xD6610C6D6DD6DA61UL, 0x3A4E622C2C3A584EUL
};
#define RSTT(d, a, b0, b1, b2, b3, b4, b5, b6, b7) do { \
t[d] = T0_G[B64_0(a[b0])] \
^ R64(T0_G[B64_1(a[b1])], 8) \
^ R64(T0_G[B64_2(a[b2])], 16) \
^ R64(T0_G[B64_3(a[b3])], 24) \
^ T4_G[B64_4(a[b4])] \
^ R64(T4_G[B64_5(a[b5])], 8) \
^ R64(T4_G[B64_6(a[b6])], 16) \
^ R64(T4_G[B64_7(a[b7])], 24); \
} while (0)
#define ROUND_SMALL_P(a, r) do { \
ulong t[8]; \
a[0] ^= PC64(0x00, r); \
a[1] ^= PC64(0x10, r); \
a[2] ^= PC64(0x20, r); \
a[3] ^= PC64(0x30, r); \
a[4] ^= PC64(0x40, r); \
a[5] ^= PC64(0x50, r); \
a[6] ^= PC64(0x60, r); \
a[7] ^= PC64(0x70, r); \
RSTT(0, a, 0, 1, 2, 3, 4, 5, 6, 7); \
RSTT(1, a, 1, 2, 3, 4, 5, 6, 7, 0); \
RSTT(2, a, 2, 3, 4, 5, 6, 7, 0, 1); \
RSTT(3, a, 3, 4, 5, 6, 7, 0, 1, 2); \
RSTT(4, a, 4, 5, 6, 7, 0, 1, 2, 3); \
RSTT(5, a, 5, 6, 7, 0, 1, 2, 3, 4); \
RSTT(6, a, 6, 7, 0, 1, 2, 3, 4, 5); \
RSTT(7, a, 7, 0, 1, 2, 3, 4, 5, 6); \
a[0] = t[0]; \
a[1] = t[1]; \
a[2] = t[2]; \
a[3] = t[3]; \
a[4] = t[4]; \
a[5] = t[5]; \
a[6] = t[6]; \
a[7] = t[7]; \
} while (0)
#define ROUND_SMALL_Pf(a,r) do { \
a[0] ^= PC64(0x00, r); \
a[1] ^= PC64(0x10, r); \
a[2] ^= PC64(0x20, r); \
a[3] ^= PC64(0x30, r); \
a[4] ^= PC64(0x40, r); \
a[5] ^= PC64(0x50, r); \
a[6] ^= PC64(0x60, r); \
a[7] ^= PC64(0x70, r); \
RSTT(7, a, 7, 0, 1, 2, 3, 4, 5, 6); \
a[7] = t[7]; \
} while (0)
#define ROUND_SMALL_Q(a, r) do { \
ulong t[8]; \
a[0] ^= QC64(0x00, r); \
a[1] ^= QC64(0x10, r); \
a[2] ^= QC64(0x20, r); \
a[3] ^= QC64(0x30, r); \
a[4] ^= QC64(0x40, r); \
a[5] ^= QC64(0x50, r); \
a[6] ^= QC64(0x60, r); \
a[7] ^= QC64(0x70, r); \
RSTT(0, a, 1, 3, 5, 7, 0, 2, 4, 6); \
RSTT(1, a, 2, 4, 6, 0, 1, 3, 5, 7); \
RSTT(2, a, 3, 5, 7, 1, 2, 4, 6, 0); \
RSTT(3, a, 4, 6, 0, 2, 3, 5, 7, 1); \
RSTT(4, a, 5, 7, 1, 3, 4, 6, 0, 2); \
RSTT(5, a, 6, 0, 2, 4, 5, 7, 1, 3); \
RSTT(6, a, 7, 1, 3, 5, 6, 0, 2, 4); \
RSTT(7, a, 0, 2, 4, 6, 7, 1, 3, 5); \
a[0] = t[0]; \
a[1] = t[1]; \
a[2] = t[2]; \
a[3] = t[3]; \
a[4] = t[4]; \
a[5] = t[5]; \
a[6] = t[6]; \
a[7] = t[7]; \
} while (0)
#define PERM_SMALL_P(a) do { \
for (int r = 0; r < 10; r ++) \
ROUND_SMALL_P(a, r); \
} while (0)
#define PERM_SMALL_Pf(a) do { \
for (int r = 0; r < 9; r ++) { \
ROUND_SMALL_P(a, r);} \
ROUND_SMALL_Pf(a,9); \
} while (0)
#define PERM_SMALL_Q(a) do { \
for (int r = 0; r < 10; r ++) \
ROUND_SMALL_Q(a, r); \
} while (0)

272
opencl/jh.cl Normal file
View File

@ -0,0 +1,272 @@
/* $Id: jh.c 255 2011-06-07 19:50:20Z tp $ */
/*
* JH implementation.
*
* ==========================(LICENSE BEGIN)============================
*
* Copyright (c) 2007-2010 Projet RNRT SAPHIR
*
* 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.
*
* ===========================(LICENSE END)=============================
*
* @author Thomas Pornin <thomas.pornin@cryptolog.com>
*/
#define SPH_JH_64 1
#define SPH_LITTLE_ENDIAN 1
#define SPH_C32(x) x
#define SPH_C64(x) x
typedef uint sph_u32;
typedef ulong sph_u64;
/*
* The internal bitslice representation may use either big-endian or
* little-endian (true bitslice operations do not care about the bit
* ordering, and the bit-swapping linear operations in JH happen to
* be invariant through endianness-swapping). The constants must be
* defined according to the chosen endianness; we use some
* byte-swapping macros for that.
*/
#if SPH_LITTLE_ENDIAN
#define C32e(x) ((SPH_C32(x) >> 24) \
| ((SPH_C32(x) >> 8) & SPH_C32(0x0000FF00)) \
| ((SPH_C32(x) << 8) & SPH_C32(0x00FF0000)) \
| ((SPH_C32(x) << 24) & SPH_C32(0xFF000000)))
#define dec32e_aligned sph_dec32le_aligned
#define enc32e sph_enc32le
#define C64e(x) ((SPH_C64(x) >> 56) \
| ((SPH_C64(x) >> 40) & SPH_C64(0x000000000000FF00)) \
| ((SPH_C64(x) >> 24) & SPH_C64(0x0000000000FF0000)) \
| ((SPH_C64(x) >> 8) & SPH_C64(0x00000000FF000000)) \
| ((SPH_C64(x) << 8) & SPH_C64(0x000000FF00000000)) \
| ((SPH_C64(x) << 24) & SPH_C64(0x0000FF0000000000)) \
| ((SPH_C64(x) << 40) & SPH_C64(0x00FF000000000000)) \
| ((SPH_C64(x) << 56) & SPH_C64(0xFF00000000000000)))
#define dec64e_aligned sph_dec64le_aligned
#define enc64e sph_enc64le
#else
#define C32e(x) SPH_C32(x)
#define dec32e_aligned sph_dec32be_aligned
#define enc32e sph_enc32be
#define C64e(x) SPH_C64(x)
#define dec64e_aligned sph_dec64be_aligned
#define enc64e sph_enc64be
#endif
#define Sb(x0, x1, x2, x3, c) do { \
x3 = ~x3; \
x0 ^= (c) & ~x2; \
tmp = (c) ^ (x0 & x1); \
x0 ^= x2 & x3; \
x3 ^= ~x1 & x2; \
x1 ^= x0 & x2; \
x2 ^= x0 & ~x3; \
x0 ^= x1 | x3; \
x3 ^= x1 & x2; \
x1 ^= tmp & x0; \
x2 ^= tmp; \
} while (0)
#define Lb(x0, x1, x2, x3, x4, x5, x6, x7) do { \
x4 ^= x1; \
x5 ^= x2; \
x6 ^= x3 ^ x0; \
x7 ^= x0; \
x0 ^= x5; \
x1 ^= x6; \
x2 ^= x7 ^ x4; \
x3 ^= x4; \
} while (0)
static const __constant ulong C[] =
{
0x67F815DFA2DED572UL, 0x571523B70A15847BUL, 0xF6875A4D90D6AB81UL, 0x402BD1C3C54F9F4EUL,
0x9CFA455CE03A98EAUL, 0x9A99B26699D2C503UL, 0x8A53BBF2B4960266UL, 0x31A2DB881A1456B5UL,
0xDB0E199A5C5AA303UL, 0x1044C1870AB23F40UL, 0x1D959E848019051CUL, 0xDCCDE75EADEB336FUL,
0x416BBF029213BA10UL, 0xD027BBF7156578DCUL, 0x5078AA3739812C0AUL, 0xD3910041D2BF1A3FUL,
0x907ECCF60D5A2D42UL, 0xCE97C0929C9F62DDUL, 0xAC442BC70BA75C18UL, 0x23FCC663D665DFD1UL,
0x1AB8E09E036C6E97UL, 0xA8EC6C447E450521UL, 0xFA618E5DBB03F1EEUL, 0x97818394B29796FDUL,
0x2F3003DB37858E4AUL, 0x956A9FFB2D8D672AUL, 0x6C69B8F88173FE8AUL, 0x14427FC04672C78AUL,
0xC45EC7BD8F15F4C5UL, 0x80BB118FA76F4475UL, 0xBC88E4AEB775DE52UL, 0xF4A3A6981E00B882UL,
0x1563A3A9338FF48EUL, 0x89F9B7D524565FAAUL, 0xFDE05A7C20EDF1B6UL, 0x362C42065AE9CA36UL,
0x3D98FE4E433529CEUL, 0xA74B9A7374F93A53UL, 0x86814E6F591FF5D0UL, 0x9F5AD8AF81AD9D0EUL,
0x6A6234EE670605A7UL, 0x2717B96EBE280B8BUL, 0x3F1080C626077447UL, 0x7B487EC66F7EA0E0UL,
0xC0A4F84AA50A550DUL, 0x9EF18E979FE7E391UL, 0xD48D605081727686UL, 0x62B0E5F3415A9E7EUL,
0x7A205440EC1F9FFCUL, 0x84C9F4CE001AE4E3UL, 0xD895FA9DF594D74FUL, 0xA554C324117E2E55UL,
0x286EFEBD2872DF5BUL, 0xB2C4A50FE27FF578UL, 0x2ED349EEEF7C8905UL, 0x7F5928EB85937E44UL,
0x4A3124B337695F70UL, 0x65E4D61DF128865EUL, 0xE720B95104771BC7UL, 0x8A87D423E843FE74UL,
0xF2947692A3E8297DUL, 0xC1D9309B097ACBDDUL, 0xE01BDC5BFB301B1DUL, 0xBF829CF24F4924DAUL,
0xFFBF70B431BAE7A4UL, 0x48BCF8DE0544320DUL, 0x39D3BB5332FCAE3BUL, 0xA08B29E0C1C39F45UL,
0x0F09AEF7FD05C9E5UL, 0x34F1904212347094UL, 0x95ED44E301B771A2UL, 0x4A982F4F368E3BE9UL,
0x15F66CA0631D4088UL, 0xFFAF52874B44C147UL, 0x30C60AE2F14ABB7EUL, 0xE68C6ECCC5B67046UL,
0x00CA4FBD56A4D5A4UL, 0xAE183EC84B849DDAUL, 0xADD1643045CE5773UL, 0x67255C1468CEA6E8UL,
0x16E10ECBF28CDAA3UL, 0x9A99949A5806E933UL, 0x7B846FC220B2601FUL, 0x1885D1A07FACCED1UL,
0xD319DD8DA15B5932UL, 0x46B4A5AAC01C9A50UL, 0xBA6B04E467633D9FUL, 0x7EEE560BAB19CAF6UL,
0x742128A9EA79B11FUL, 0xEE51363B35F7BDE9UL, 0x76D350755AAC571DUL, 0x01707DA3FEC2463AUL,
0x42D8A498AFC135F7UL, 0x79676B9E20ECED78UL, 0xA8DB3AEA15638341UL, 0x832C83324D3BC3FAUL,
0xF347271C1F3B40A7UL, 0x9A762DB734F04059UL, 0xFD4F21D26C4E3EE7UL, 0xEF5957DC398DFDB8UL,
0xDAEB492B490C9B8DUL, 0x0D70F36849D7A25BUL, 0x84558D7AD0AE3B7DUL, 0x658EF8E4F0E9A5F5UL,
0x533B1036F4A2B8A0UL, 0x5AEC3E759E07A80CUL, 0x4F88E85692946891UL, 0x4CBCBAF8555CB05BUL,
0x7B9487F3993BBBE3UL, 0x5D1C6B72D6F4DA75UL, 0x6DB334DC28ACAE64UL, 0x71DB28B850A5346CUL,
0x2A518D10F2E261F8UL, 0xFC75DD593364DBE3UL, 0xA23FCE43F1BCAC1CUL, 0xB043E8023CD1BB67UL,
0x75A12988CA5B0A33UL, 0x5C5316B44D19347FUL, 0x1E4D790EC3943B92UL, 0x3FAFEEB6D7757479UL,
0x21391ABEF7D4A8EAUL, 0x5127234C097EF45CUL, 0xD23C32BA5324A326UL, 0xADD5A66D4A17A344UL,
0x08C9F2AFA63E1DB5UL, 0x563C6B91983D5983UL, 0x4D608672A17CF84CUL, 0xF6C76E08CC3EE246UL,
0x5E76BCB1B333982FUL, 0x2AE6C4EFA566D62BUL, 0x36D4C1BEE8B6F406UL, 0x6321EFBC1582EE74UL,
0x69C953F40D4EC1FDUL, 0x26585806C45A7DA7UL, 0x16FAE0061614C17EUL, 0x3F9D63283DAF907EUL,
0x0CD29B00E3F2C9D2UL, 0x300CD4B730CEAA5FUL, 0x9832E0F216512A74UL, 0x9AF8CEE3D830EB0DUL,
0x9279F1B57B9EC54BUL, 0xD36886046EE651FFUL, 0x316796E6574D239BUL, 0x05750A17F3A6E6CCUL,
0xCE6C3213D98176B1UL, 0x62A205F88452173CUL, 0x47154778B3CB2BF4UL, 0x486A9323825446FFUL,
0x65655E4E0758DF38UL, 0x8E5086FC897CFCF2UL, 0x86CA0BD0442E7031UL, 0x4E477830A20940F0UL,
0x8338F7D139EEA065UL, 0xBD3A2CE437E95EF7UL, 0x6FF8130126B29721UL, 0xE7DE9FEFD1ED44A3UL,
0xD992257615DFA08BUL, 0xBE42DC12F6F7853CUL, 0x7EB027AB7CECA7D8UL, 0xDEA83EAADA7D8D53UL,
0xD86902BD93CE25AAUL, 0xF908731AFD43F65AUL, 0xA5194A17DAEF5FC0UL, 0x6A21FD4C33664D97UL,
0x701541DB3198B435UL, 0x9B54CDEDBB0F1EEAUL, 0x72409751A163D09AUL, 0xE26F4791BF9D75F6UL
};
#define Ceven_hi(r) (C[((r) << 2) + 0])
#define Ceven_lo(r) (C[((r) << 2) + 1])
#define Codd_hi(r) (C[((r) << 2) + 2])
#define Codd_lo(r) (C[((r) << 2) + 3])
#define S(x0, x1, x2, x3, cb, r) do { \
Sb(x0 ## h, x1 ## h, x2 ## h, x3 ## h, cb ## hi(r)); \
Sb(x0 ## l, x1 ## l, x2 ## l, x3 ## l, cb ## lo(r)); \
} while (0)
#define L(x0, x1, x2, x3, x4, x5, x6, x7) do { \
Lb(x0 ## h, x1 ## h, x2 ## h, x3 ## h, \
x4 ## h, x5 ## h, x6 ## h, x7 ## h); \
Lb(x0 ## l, x1 ## l, x2 ## l, x3 ## l, \
x4 ## l, x5 ## l, x6 ## l, x7 ## l); \
} while (0)
#define Wz(x, c, n) do { \
sph_u64 t = (x ## h & (c)) << (n); \
x ## h = ((x ## h >> (n)) & (c)) | t; \
t = (x ## l & (c)) << (n); \
x ## l = ((x ## l >> (n)) & (c)) | t; \
} while (0)
#define W0(x) Wz(x, SPH_C64(0x5555555555555555), 1)
#define W1(x) Wz(x, SPH_C64(0x3333333333333333), 2)
#define W2(x) Wz(x, SPH_C64(0x0F0F0F0F0F0F0F0F), 4)
#define W3(x) Wz(x, SPH_C64(0x00FF00FF00FF00FF), 8)
#define W4(x) Wz(x, SPH_C64(0x0000FFFF0000FFFF), 16)
#define W5(x) Wz(x, SPH_C64(0x00000000FFFFFFFF), 32)
#define W6(x) do { \
sph_u64 t = x ## h; \
x ## h = x ## l; \
x ## l = t; \
} while (0)
#define SL(ro) SLu(r + ro, ro)
#define SLu(r, ro) do { \
S(h0, h2, h4, h6, Ceven_, r); \
S(h1, h3, h5, h7, Codd_, r); \
L(h0, h2, h4, h6, h1, h3, h5, h7); \
W ## ro(h1); \
W ## ro(h3); \
W ## ro(h5); \
W ## ro(h7); \
} while (0)
#if SPH_SMALL_FOOTPRINT_JH
/*
* The "small footprint" 64-bit version just uses a partially unrolled
* loop.
*/
#define E8 do { \
unsigned r; \
for (r = 0; r < 42; r += 7) { \
SL(0); \
SL(1); \
SL(2); \
SL(3); \
SL(4); \
SL(5); \
SL(6); \
} \
} while (0)
#else
/*
* On a "true 64-bit" architecture, we can unroll at will.
*/
#define E8 do { \
SLu( 0, 0); \
SLu( 1, 1); \
SLu( 2, 2); \
SLu( 3, 3); \
SLu( 4, 4); \
SLu( 5, 5); \
SLu( 6, 6); \
SLu( 7, 0); \
SLu( 8, 1); \
SLu( 9, 2); \
SLu(10, 3); \
SLu(11, 4); \
SLu(12, 5); \
SLu(13, 6); \
SLu(14, 0); \
SLu(15, 1); \
SLu(16, 2); \
SLu(17, 3); \
SLu(18, 4); \
SLu(19, 5); \
SLu(20, 6); \
SLu(21, 0); \
SLu(22, 1); \
SLu(23, 2); \
SLu(24, 3); \
SLu(25, 4); \
SLu(26, 5); \
SLu(27, 6); \
SLu(28, 0); \
SLu(29, 1); \
SLu(30, 2); \
SLu(31, 3); \
SLu(32, 4); \
SLu(33, 5); \
SLu(34, 6); \
SLu(35, 0); \
SLu(36, 1); \
SLu(37, 2); \
SLu(38, 3); \
SLu(39, 4); \
SLu(40, 5); \
SLu(41, 6); \
} while (0)
#endif

87
opencl/wolf-aes.cl Normal file
View File

@ -0,0 +1,87 @@
#ifndef WOLF_AES_CL
#define WOLF_AES_CL
// AES table - the other three are generated on the fly
static const __constant uint AES0_C[256] =
{
0xA56363C6U, 0x847C7CF8U, 0x997777EEU, 0x8D7B7BF6U,
0x0DF2F2FFU, 0xBD6B6BD6U, 0xB16F6FDEU, 0x54C5C591U,
0x50303060U, 0x03010102U, 0xA96767CEU, 0x7D2B2B56U,
0x19FEFEE7U, 0x62D7D7B5U, 0xE6ABAB4DU, 0x9A7676ECU,
0x45CACA8FU, 0x9D82821FU, 0x40C9C989U, 0x877D7DFAU,
0x15FAFAEFU, 0xEB5959B2U, 0xC947478EU, 0x0BF0F0FBU,
0xECADAD41U, 0x67D4D4B3U, 0xFDA2A25FU, 0xEAAFAF45U,
0xBF9C9C23U, 0xF7A4A453U, 0x967272E4U, 0x5BC0C09BU,
0xC2B7B775U, 0x1CFDFDE1U, 0xAE93933DU, 0x6A26264CU,
0x5A36366CU, 0x413F3F7EU, 0x02F7F7F5U, 0x4FCCCC83U,
0x5C343468U, 0xF4A5A551U, 0x34E5E5D1U, 0x08F1F1F9U,
0x937171E2U, 0x73D8D8ABU, 0x53313162U, 0x3F15152AU,
0x0C040408U, 0x52C7C795U, 0x65232346U, 0x5EC3C39DU,
0x28181830U, 0xA1969637U, 0x0F05050AU, 0xB59A9A2FU,
0x0907070EU, 0x36121224U, 0x9B80801BU, 0x3DE2E2DFU,
0x26EBEBCDU, 0x6927274EU, 0xCDB2B27FU, 0x9F7575EAU,
0x1B090912U, 0x9E83831DU, 0x742C2C58U, 0x2E1A1A34U,
0x2D1B1B36U, 0xB26E6EDCU, 0xEE5A5AB4U, 0xFBA0A05BU,
0xF65252A4U, 0x4D3B3B76U, 0x61D6D6B7U, 0xCEB3B37DU,
0x7B292952U, 0x3EE3E3DDU, 0x712F2F5EU, 0x97848413U,
0xF55353A6U, 0x68D1D1B9U, 0x00000000U, 0x2CEDEDC1U,
0x60202040U, 0x1FFCFCE3U, 0xC8B1B179U, 0xED5B5BB6U,
0xBE6A6AD4U, 0x46CBCB8DU, 0xD9BEBE67U, 0x4B393972U,
0xDE4A4A94U, 0xD44C4C98U, 0xE85858B0U, 0x4ACFCF85U,
0x6BD0D0BBU, 0x2AEFEFC5U, 0xE5AAAA4FU, 0x16FBFBEDU,
0xC5434386U, 0xD74D4D9AU, 0x55333366U, 0x94858511U,
0xCF45458AU, 0x10F9F9E9U, 0x06020204U, 0x817F7FFEU,
0xF05050A0U, 0x443C3C78U, 0xBA9F9F25U, 0xE3A8A84BU,
0xF35151A2U, 0xFEA3A35DU, 0xC0404080U, 0x8A8F8F05U,
0xAD92923FU, 0xBC9D9D21U, 0x48383870U, 0x04F5F5F1U,
0xDFBCBC63U, 0xC1B6B677U, 0x75DADAAFU, 0x63212142U,
0x30101020U, 0x1AFFFFE5U, 0x0EF3F3FDU, 0x6DD2D2BFU,
0x4CCDCD81U, 0x140C0C18U, 0x35131326U, 0x2FECECC3U,
0xE15F5FBEU, 0xA2979735U, 0xCC444488U, 0x3917172EU,
0x57C4C493U, 0xF2A7A755U, 0x827E7EFCU, 0x473D3D7AU,
0xAC6464C8U, 0xE75D5DBAU, 0x2B191932U, 0x957373E6U,
0xA06060C0U, 0x98818119U, 0xD14F4F9EU, 0x7FDCDCA3U,
0x66222244U, 0x7E2A2A54U, 0xAB90903BU, 0x8388880BU,
0xCA46468CU, 0x29EEEEC7U, 0xD3B8B86BU, 0x3C141428U,
0x79DEDEA7U, 0xE25E5EBCU, 0x1D0B0B16U, 0x76DBDBADU,
0x3BE0E0DBU, 0x56323264U, 0x4E3A3A74U, 0x1E0A0A14U,
0xDB494992U, 0x0A06060CU, 0x6C242448U, 0xE45C5CB8U,
0x5DC2C29FU, 0x6ED3D3BDU, 0xEFACAC43U, 0xA66262C4U,
0xA8919139U, 0xA4959531U, 0x37E4E4D3U, 0x8B7979F2U,
0x32E7E7D5U, 0x43C8C88BU, 0x5937376EU, 0xB76D6DDAU,
0x8C8D8D01U, 0x64D5D5B1U, 0xD24E4E9CU, 0xE0A9A949U,
0xB46C6CD8U, 0xFA5656ACU, 0x07F4F4F3U, 0x25EAEACFU,
0xAF6565CAU, 0x8E7A7AF4U, 0xE9AEAE47U, 0x18080810U,
0xD5BABA6FU, 0x887878F0U, 0x6F25254AU, 0x722E2E5CU,
0x241C1C38U, 0xF1A6A657U, 0xC7B4B473U, 0x51C6C697U,
0x23E8E8CBU, 0x7CDDDDA1U, 0x9C7474E8U, 0x211F1F3EU,
0xDD4B4B96U, 0xDCBDBD61U, 0x868B8B0DU, 0x858A8A0FU,
0x907070E0U, 0x423E3E7CU, 0xC4B5B571U, 0xAA6666CCU,
0xD8484890U, 0x05030306U, 0x01F6F6F7U, 0x120E0E1CU,
0xA36161C2U, 0x5F35356AU, 0xF95757AEU, 0xD0B9B969U,
0x91868617U, 0x58C1C199U, 0x271D1D3AU, 0xB99E9E27U,
0x38E1E1D9U, 0x13F8F8EBU, 0xB398982BU, 0x33111122U,
0xBB6969D2U, 0x70D9D9A9U, 0x898E8E07U, 0xA7949433U,
0xB69B9B2DU, 0x221E1E3CU, 0x92878715U, 0x20E9E9C9U,
0x49CECE87U, 0xFF5555AAU, 0x78282850U, 0x7ADFDFA5U,
0x8F8C8C03U, 0xF8A1A159U, 0x80898909U, 0x170D0D1AU,
0xDABFBF65U, 0x31E6E6D7U, 0xC6424284U, 0xB86868D0U,
0xC3414182U, 0xB0999929U, 0x772D2D5AU, 0x110F0F1EU,
0xCBB0B07BU, 0xFC5454A8U, 0xD6BBBB6DU, 0x3A16162CU
};
#define BYTE(x, y) (amd_bfe((x), (y) << 3U, 8U))
uint4 AES_Round(const __local uint *AES0, const __local uint *AES1, const __local uint *AES2, const __local uint *AES3, const uint4 X, const uint4 key)
{
uint4 Y;
Y.s0 = AES0[BYTE(X.s0, 0)] ^ AES1[BYTE(X.s1, 1)] ^ AES2[BYTE(X.s2, 2)] ^ AES3[BYTE(X.s3, 3)];
Y.s1 = AES0[BYTE(X.s1, 0)] ^ AES1[BYTE(X.s2, 1)] ^ AES2[BYTE(X.s3, 2)] ^ AES3[BYTE(X.s0, 3)];
Y.s2 = AES0[BYTE(X.s2, 0)] ^ AES1[BYTE(X.s3, 1)] ^ AES2[BYTE(X.s0, 2)] ^ AES3[BYTE(X.s1, 3)];
Y.s3 = AES0[BYTE(X.s3, 0)] ^ AES1[BYTE(X.s0, 1)] ^ AES2[BYTE(X.s1, 2)] ^ AES3[BYTE(X.s2, 3)];
Y ^= key;
return(Y);
}
#endif

111
opencl/wolf-skein.cl Normal file
View File

@ -0,0 +1,111 @@
#ifndef WOLF_SKEIN_CL
#define WOLF_SKEIN_CL
// Vectorized Skein implementation macros and functions by Wolf
#define SKEIN_KS_PARITY 0x1BD11BDAA9FC1A22
static const __constant ulong SKEIN256_IV[8] =
{
0xCCD044A12FDB3E13UL, 0xE83590301A79A9EBUL,
0x55AEA0614F816E6FUL, 0x2A2767A4AE9B94DBUL,
0xEC06025E74DD7683UL, 0xE7A436CDC4746251UL,
0xC36FBAF9393AD185UL, 0x3EEDBA1833EDFC13UL
};
static const __constant ulong SKEIN512_256_IV[8] =
{
0xCCD044A12FDB3E13UL, 0xE83590301A79A9EBUL,
0x55AEA0614F816E6FUL, 0x2A2767A4AE9B94DBUL,
0xEC06025E74DD7683UL, 0xE7A436CDC4746251UL,
0xC36FBAF9393AD185UL, 0x3EEDBA1833EDFC13UL
};
#define SKEIN_INJECT_KEY(p, s) do { \
p += h; \
p.s5 += t[s % 3]; \
p.s6 += t[(s + 1) % 3]; \
p.s7 += s; \
} while(0)
ulong SKEIN_ROT(const uint2 x, const uint y)
{
if(y < 32) return(as_ulong(amd_bitalign(x, x.s10, 32 - y)));
else return(as_ulong(amd_bitalign(x.s10, x, 32 - (y - 32))));
}
void SkeinMix8(ulong4 *pv0, ulong4 *pv1, const uint rc0, const uint rc1, const uint rc2, const uint rc3)
{
*pv0 += *pv1;
(*pv1).s0 = SKEIN_ROT(as_uint2((*pv1).s0), rc0);
(*pv1).s1 = SKEIN_ROT(as_uint2((*pv1).s1), rc1);
(*pv1).s2 = SKEIN_ROT(as_uint2((*pv1).s2), rc2);
(*pv1).s3 = SKEIN_ROT(as_uint2((*pv1).s3), rc3);
*pv1 ^= *pv0;
}
ulong8 SkeinEvenRound(ulong8 p, const ulong8 h, const ulong *t, const uint s)
{
SKEIN_INJECT_KEY(p, s);
ulong4 pv0 = p.even, pv1 = p.odd;
SkeinMix8(&pv0, &pv1, 46, 36, 19, 37);
pv0 = shuffle(pv0, (ulong4)(1, 2, 3, 0));
pv1 = shuffle(pv1, (ulong4)(0, 3, 2, 1));
SkeinMix8(&pv0, &pv1, 33, 27, 14, 42);
pv0 = shuffle(pv0, (ulong4)(1, 2, 3, 0));
pv1 = shuffle(pv1, (ulong4)(0, 3, 2, 1));
SkeinMix8(&pv0, &pv1, 17, 49, 36, 39);
pv0 = shuffle(pv0, (ulong4)(1, 2, 3, 0));
pv1 = shuffle(pv1, (ulong4)(0, 3, 2, 1));
SkeinMix8(&pv0, &pv1, 44, 9, 54, 56);
return(shuffle2(pv0, pv1, (ulong8)(1, 4, 2, 7, 3, 6, 0, 5)));
}
ulong8 SkeinOddRound(ulong8 p, const ulong8 h, const ulong *t, const uint s)
{
SKEIN_INJECT_KEY(p, s);
ulong4 pv0 = p.even, pv1 = p.odd;
SkeinMix8(&pv0, &pv1, 39, 30, 34, 24);
pv0 = shuffle(pv0, (ulong4)(1, 2, 3, 0));
pv1 = shuffle(pv1, (ulong4)(0, 3, 2, 1));
SkeinMix8(&pv0, &pv1, 13, 50, 10, 17);
pv0 = shuffle(pv0, (ulong4)(1, 2, 3, 0));
pv1 = shuffle(pv1, (ulong4)(0, 3, 2, 1));
SkeinMix8(&pv0, &pv1, 25, 29, 39, 43);
pv0 = shuffle(pv0, (ulong4)(1, 2, 3, 0));
pv1 = shuffle(pv1, (ulong4)(0, 3, 2, 1));
SkeinMix8(&pv0, &pv1, 8, 35, 56, 22);
return(shuffle2(pv0, pv1, (ulong8)(1, 4, 2, 7, 3, 6, 0, 5)));
}
ulong8 Skein512Block(ulong8 p, ulong8 h, ulong h8, const ulong *t)
{
#pragma unroll
for(int i = 0; i < 18; ++i)
{
p = SkeinEvenRound(p, h, t, i);
++i;
ulong tmp = h.s0;
h = shuffle(h, (ulong8)(1, 2, 3, 4, 5, 6, 7, 0));
h.s7 = h8;
h8 = tmp;
p = SkeinOddRound(p, h, t, i);
tmp = h.s0;
h = shuffle(h, (ulong8)(1, 2, 3, 4, 5, 6, 7, 0));
h.s7 = h8;
h8 = tmp;
}
SKEIN_INJECT_KEY(p, 18);
return(p);
}
#endif

271
rapidjson/allocators.h Normal file
View File

@ -0,0 +1,271 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_ALLOCATORS_H_
#define RAPIDJSON_ALLOCATORS_H_
#include "rapidjson.h"
RAPIDJSON_NAMESPACE_BEGIN
///////////////////////////////////////////////////////////////////////////////
// Allocator
/*! \class rapidjson::Allocator
\brief Concept for allocating, resizing and freeing memory block.
Note that Malloc() and Realloc() are non-static but Free() is static.
So if an allocator need to support Free(), it needs to put its pointer in
the header of memory block.
\code
concept Allocator {
static const bool kNeedFree; //!< Whether this allocator needs to call Free().
// Allocate a memory block.
// \param size of the memory block in bytes.
// \returns pointer to the memory block.
void* Malloc(size_t size);
// Resize a memory block.
// \param originalPtr The pointer to current memory block. Null pointer is permitted.
// \param originalSize The current size in bytes. (Design issue: since some allocator may not book-keep this, explicitly pass to it can save memory.)
// \param newSize the new size in bytes.
void* Realloc(void* originalPtr, size_t originalSize, size_t newSize);
// Free a memory block.
// \param pointer to the memory block. Null pointer is permitted.
static void Free(void *ptr);
};
\endcode
*/
///////////////////////////////////////////////////////////////////////////////
// CrtAllocator
//! C-runtime library allocator.
/*! This class is just wrapper for standard C library memory routines.
\note implements Allocator concept
*/
class CrtAllocator {
public:
static const bool kNeedFree = true;
void* Malloc(size_t size) {
if (size) // behavior of malloc(0) is implementation defined.
return std::malloc(size);
else
return NULL; // standardize to returning NULL.
}
void* Realloc(void* originalPtr, size_t originalSize, size_t newSize) {
(void)originalSize;
if (newSize == 0) {
std::free(originalPtr);
return NULL;
}
return std::realloc(originalPtr, newSize);
}
static void Free(void *ptr) { std::free(ptr); }
};
///////////////////////////////////////////////////////////////////////////////
// MemoryPoolAllocator
//! Default memory allocator used by the parser and DOM.
/*! This allocator allocate memory blocks from pre-allocated memory chunks.
It does not free memory blocks. And Realloc() only allocate new memory.
The memory chunks are allocated by BaseAllocator, which is CrtAllocator by default.
User may also supply a buffer as the first chunk.
If the user-buffer is full then additional chunks are allocated by BaseAllocator.
The user-buffer is not deallocated by this allocator.
\tparam BaseAllocator the allocator type for allocating memory chunks. Default is CrtAllocator.
\note implements Allocator concept
*/
template <typename BaseAllocator = CrtAllocator>
class MemoryPoolAllocator {
public:
static const bool kNeedFree = false; //!< Tell users that no need to call Free() with this allocator. (concept Allocator)
//! Constructor with chunkSize.
/*! \param chunkSize The size of memory chunk. The default is kDefaultChunkSize.
\param baseAllocator The allocator for allocating memory chunks.
*/
MemoryPoolAllocator(size_t chunkSize = kDefaultChunkCapacity, BaseAllocator* baseAllocator = 0) :
chunkHead_(0), chunk_capacity_(chunkSize), userBuffer_(0), baseAllocator_(baseAllocator), ownBaseAllocator_(0)
{
}
//! Constructor with user-supplied buffer.
/*! The user buffer will be used firstly. When it is full, memory pool allocates new chunk with chunk size.
The user buffer will not be deallocated when this allocator is destructed.
\param buffer User supplied buffer.
\param size Size of the buffer in bytes. It must at least larger than sizeof(ChunkHeader).
\param chunkSize The size of memory chunk. The default is kDefaultChunkSize.
\param baseAllocator The allocator for allocating memory chunks.
*/
MemoryPoolAllocator(void *buffer, size_t size, size_t chunkSize = kDefaultChunkCapacity, BaseAllocator* baseAllocator = 0) :
chunkHead_(0), chunk_capacity_(chunkSize), userBuffer_(buffer), baseAllocator_(baseAllocator), ownBaseAllocator_(0)
{
RAPIDJSON_ASSERT(buffer != 0);
RAPIDJSON_ASSERT(size > sizeof(ChunkHeader));
chunkHead_ = reinterpret_cast<ChunkHeader*>(buffer);
chunkHead_->capacity = size - sizeof(ChunkHeader);
chunkHead_->size = 0;
chunkHead_->next = 0;
}
//! Destructor.
/*! This deallocates all memory chunks, excluding the user-supplied buffer.
*/
~MemoryPoolAllocator() {
Clear();
RAPIDJSON_DELETE(ownBaseAllocator_);
}
//! Deallocates all memory chunks, excluding the user-supplied buffer.
void Clear() {
while (chunkHead_ && chunkHead_ != userBuffer_) {
ChunkHeader* next = chunkHead_->next;
baseAllocator_->Free(chunkHead_);
chunkHead_ = next;
}
if (chunkHead_ && chunkHead_ == userBuffer_)
chunkHead_->size = 0; // Clear user buffer
}
//! Computes the total capacity of allocated memory chunks.
/*! \return total capacity in bytes.
*/
size_t Capacity() const {
size_t capacity = 0;
for (ChunkHeader* c = chunkHead_; c != 0; c = c->next)
capacity += c->capacity;
return capacity;
}
//! Computes the memory blocks allocated.
/*! \return total used bytes.
*/
size_t Size() const {
size_t size = 0;
for (ChunkHeader* c = chunkHead_; c != 0; c = c->next)
size += c->size;
return size;
}
//! Allocates a memory block. (concept Allocator)
void* Malloc(size_t size) {
if (!size)
return NULL;
size = RAPIDJSON_ALIGN(size);
if (chunkHead_ == 0 || chunkHead_->size + size > chunkHead_->capacity)
if (!AddChunk(chunk_capacity_ > size ? chunk_capacity_ : size))
return NULL;
void *buffer = reinterpret_cast<char *>(chunkHead_) + RAPIDJSON_ALIGN(sizeof(ChunkHeader)) + chunkHead_->size;
chunkHead_->size += size;
return buffer;
}
//! Resizes a memory block (concept Allocator)
void* Realloc(void* originalPtr, size_t originalSize, size_t newSize) {
if (originalPtr == 0)
return Malloc(newSize);
if (newSize == 0)
return NULL;
originalSize = RAPIDJSON_ALIGN(originalSize);
newSize = RAPIDJSON_ALIGN(newSize);
// Do not shrink if new size is smaller than original
if (originalSize >= newSize)
return originalPtr;
// Simply expand it if it is the last allocation and there is sufficient space
if (originalPtr == reinterpret_cast<char *>(chunkHead_) + RAPIDJSON_ALIGN(sizeof(ChunkHeader)) + chunkHead_->size - originalSize) {
size_t increment = static_cast<size_t>(newSize - originalSize);
if (chunkHead_->size + increment <= chunkHead_->capacity) {
chunkHead_->size += increment;
return originalPtr;
}
}
// Realloc process: allocate and copy memory, do not free original buffer.
if (void* newBuffer = Malloc(newSize)) {
if (originalSize)
std::memcpy(newBuffer, originalPtr, originalSize);
return newBuffer;
}
else
return NULL;
}
//! Frees a memory block (concept Allocator)
static void Free(void *ptr) { (void)ptr; } // Do nothing
private:
//! Copy constructor is not permitted.
MemoryPoolAllocator(const MemoryPoolAllocator& rhs) /* = delete */;
//! Copy assignment operator is not permitted.
MemoryPoolAllocator& operator=(const MemoryPoolAllocator& rhs) /* = delete */;
//! Creates a new chunk.
/*! \param capacity Capacity of the chunk in bytes.
\return true if success.
*/
bool AddChunk(size_t capacity) {
if (!baseAllocator_)
ownBaseAllocator_ = baseAllocator_ = RAPIDJSON_NEW(BaseAllocator());
if (ChunkHeader* chunk = reinterpret_cast<ChunkHeader*>(baseAllocator_->Malloc(RAPIDJSON_ALIGN(sizeof(ChunkHeader)) + capacity))) {
chunk->capacity = capacity;
chunk->size = 0;
chunk->next = chunkHead_;
chunkHead_ = chunk;
return true;
}
else
return false;
}
static const int kDefaultChunkCapacity = 64 * 1024; //!< Default chunk capacity.
//! Chunk header for perpending to each chunk.
/*! Chunks are stored as a singly linked list.
*/
struct ChunkHeader {
size_t capacity; //!< Capacity of the chunk in bytes (excluding the header itself).
size_t size; //!< Current size of allocated memory in bytes.
ChunkHeader *next; //!< Next chunk in the linked list.
};
ChunkHeader *chunkHead_; //!< Head of the chunk linked-list. Only the head chunk serves allocation.
size_t chunk_capacity_; //!< The minimum capacity of chunk when they are allocated.
void *userBuffer_; //!< User supplied buffer.
BaseAllocator* baseAllocator_; //!< base allocator for allocating memory chunks.
BaseAllocator* ownBaseAllocator_; //!< base allocator created by this object.
};
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_ENCODINGS_H_

2602
rapidjson/document.h Normal file
View File

File diff suppressed because it is too large Load Diff

299
rapidjson/encodedstream.h Normal file
View File

@ -0,0 +1,299 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_ENCODEDSTREAM_H_
#define RAPIDJSON_ENCODEDSTREAM_H_
#include "stream.h"
#include "memorystream.h"
#ifdef __GNUC__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(effc++)
#endif
#ifdef __clang__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(padded)
#endif
RAPIDJSON_NAMESPACE_BEGIN
//! Input byte stream wrapper with a statically bound encoding.
/*!
\tparam Encoding The interpretation of encoding of the stream. Either UTF8, UTF16LE, UTF16BE, UTF32LE, UTF32BE.
\tparam InputByteStream Type of input byte stream. For example, FileReadStream.
*/
template <typename Encoding, typename InputByteStream>
class EncodedInputStream {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
public:
typedef typename Encoding::Ch Ch;
EncodedInputStream(InputByteStream& is) : is_(is) {
current_ = Encoding::TakeBOM(is_);
}
Ch Peek() const { return current_; }
Ch Take() { Ch c = current_; current_ = Encoding::Take(is_); return c; }
size_t Tell() const { return is_.Tell(); }
// Not implemented
void Put(Ch) { RAPIDJSON_ASSERT(false); }
void Flush() { RAPIDJSON_ASSERT(false); }
Ch* PutBegin() { RAPIDJSON_ASSERT(false); return 0; }
size_t PutEnd(Ch*) { RAPIDJSON_ASSERT(false); return 0; }
private:
EncodedInputStream(const EncodedInputStream&);
EncodedInputStream& operator=(const EncodedInputStream&);
InputByteStream& is_;
Ch current_;
};
//! Specialized for UTF8 MemoryStream.
template <>
class EncodedInputStream<UTF8<>, MemoryStream> {
public:
typedef UTF8<>::Ch Ch;
EncodedInputStream(MemoryStream& is) : is_(is) {
if (static_cast<unsigned char>(is_.Peek()) == 0xEFu) is_.Take();
if (static_cast<unsigned char>(is_.Peek()) == 0xBBu) is_.Take();
if (static_cast<unsigned char>(is_.Peek()) == 0xBFu) is_.Take();
}
Ch Peek() const { return is_.Peek(); }
Ch Take() { return is_.Take(); }
size_t Tell() const { return is_.Tell(); }
// Not implemented
void Put(Ch) {}
void Flush() {}
Ch* PutBegin() { return 0; }
size_t PutEnd(Ch*) { return 0; }
MemoryStream& is_;
private:
EncodedInputStream(const EncodedInputStream&);
EncodedInputStream& operator=(const EncodedInputStream&);
};
//! Output byte stream wrapper with statically bound encoding.
/*!
\tparam Encoding The interpretation of encoding of the stream. Either UTF8, UTF16LE, UTF16BE, UTF32LE, UTF32BE.
\tparam OutputByteStream Type of input byte stream. For example, FileWriteStream.
*/
template <typename Encoding, typename OutputByteStream>
class EncodedOutputStream {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
public:
typedef typename Encoding::Ch Ch;
EncodedOutputStream(OutputByteStream& os, bool putBOM = true) : os_(os) {
if (putBOM)
Encoding::PutBOM(os_);
}
void Put(Ch c) { Encoding::Put(os_, c); }
void Flush() { os_.Flush(); }
// Not implemented
Ch Peek() const { RAPIDJSON_ASSERT(false); return 0;}
Ch Take() { RAPIDJSON_ASSERT(false); return 0;}
size_t Tell() const { RAPIDJSON_ASSERT(false); return 0; }
Ch* PutBegin() { RAPIDJSON_ASSERT(false); return 0; }
size_t PutEnd(Ch*) { RAPIDJSON_ASSERT(false); return 0; }
private:
EncodedOutputStream(const EncodedOutputStream&);
EncodedOutputStream& operator=(const EncodedOutputStream&);
OutputByteStream& os_;
};
#define RAPIDJSON_ENCODINGS_FUNC(x) UTF8<Ch>::x, UTF16LE<Ch>::x, UTF16BE<Ch>::x, UTF32LE<Ch>::x, UTF32BE<Ch>::x
//! Input stream wrapper with dynamically bound encoding and automatic encoding detection.
/*!
\tparam CharType Type of character for reading.
\tparam InputByteStream type of input byte stream to be wrapped.
*/
template <typename CharType, typename InputByteStream>
class AutoUTFInputStream {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
public:
typedef CharType Ch;
//! Constructor.
/*!
\param is input stream to be wrapped.
\param type UTF encoding type if it is not detected from the stream.
*/
AutoUTFInputStream(InputByteStream& is, UTFType type = kUTF8) : is_(&is), type_(type), hasBOM_(false) {
RAPIDJSON_ASSERT(type >= kUTF8 && type <= kUTF32BE);
DetectType();
static const TakeFunc f[] = { RAPIDJSON_ENCODINGS_FUNC(Take) };
takeFunc_ = f[type_];
current_ = takeFunc_(*is_);
}
UTFType GetType() const { return type_; }
bool HasBOM() const { return hasBOM_; }
Ch Peek() const { return current_; }
Ch Take() { Ch c = current_; current_ = takeFunc_(*is_); return c; }
size_t Tell() const { return is_->Tell(); }
// Not implemented
void Put(Ch) { RAPIDJSON_ASSERT(false); }
void Flush() { RAPIDJSON_ASSERT(false); }
Ch* PutBegin() { RAPIDJSON_ASSERT(false); return 0; }
size_t PutEnd(Ch*) { RAPIDJSON_ASSERT(false); return 0; }
private:
AutoUTFInputStream(const AutoUTFInputStream&);
AutoUTFInputStream& operator=(const AutoUTFInputStream&);
// Detect encoding type with BOM or RFC 4627
void DetectType() {
// BOM (Byte Order Mark):
// 00 00 FE FF UTF-32BE
// FF FE 00 00 UTF-32LE
// FE FF UTF-16BE
// FF FE UTF-16LE
// EF BB BF UTF-8
const unsigned char* c = reinterpret_cast<const unsigned char *>(is_->Peek4());
if (!c)
return;
unsigned bom = static_cast<unsigned>(c[0] | (c[1] << 8) | (c[2] << 16) | (c[3] << 24));
hasBOM_ = false;
if (bom == 0xFFFE0000) { type_ = kUTF32BE; hasBOM_ = true; is_->Take(); is_->Take(); is_->Take(); is_->Take(); }
else if (bom == 0x0000FEFF) { type_ = kUTF32LE; hasBOM_ = true; is_->Take(); is_->Take(); is_->Take(); is_->Take(); }
else if ((bom & 0xFFFF) == 0xFFFE) { type_ = kUTF16BE; hasBOM_ = true; is_->Take(); is_->Take(); }
else if ((bom & 0xFFFF) == 0xFEFF) { type_ = kUTF16LE; hasBOM_ = true; is_->Take(); is_->Take(); }
else if ((bom & 0xFFFFFF) == 0xBFBBEF) { type_ = kUTF8; hasBOM_ = true; is_->Take(); is_->Take(); is_->Take(); }
// RFC 4627: Section 3
// "Since the first two characters of a JSON text will always be ASCII
// characters [RFC0020], it is possible to determine whether an octet
// stream is UTF-8, UTF-16 (BE or LE), or UTF-32 (BE or LE) by looking
// at the pattern of nulls in the first four octets."
// 00 00 00 xx UTF-32BE
// 00 xx 00 xx UTF-16BE
// xx 00 00 00 UTF-32LE
// xx 00 xx 00 UTF-16LE
// xx xx xx xx UTF-8
if (!hasBOM_) {
unsigned pattern = (c[0] ? 1 : 0) | (c[1] ? 2 : 0) | (c[2] ? 4 : 0) | (c[3] ? 8 : 0);
switch (pattern) {
case 0x08: type_ = kUTF32BE; break;
case 0x0A: type_ = kUTF16BE; break;
case 0x01: type_ = kUTF32LE; break;
case 0x05: type_ = kUTF16LE; break;
case 0x0F: type_ = kUTF8; break;
default: break; // Use type defined by user.
}
}
// Runtime check whether the size of character type is sufficient. It only perform checks with assertion.
if (type_ == kUTF16LE || type_ == kUTF16BE) RAPIDJSON_ASSERT(sizeof(Ch) >= 2);
if (type_ == kUTF32LE || type_ == kUTF32BE) RAPIDJSON_ASSERT(sizeof(Ch) >= 4);
}
typedef Ch (*TakeFunc)(InputByteStream& is);
InputByteStream* is_;
UTFType type_;
Ch current_;
TakeFunc takeFunc_;
bool hasBOM_;
};
//! Output stream wrapper with dynamically bound encoding and automatic encoding detection.
/*!
\tparam CharType Type of character for writing.
\tparam OutputByteStream type of output byte stream to be wrapped.
*/
template <typename CharType, typename OutputByteStream>
class AutoUTFOutputStream {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
public:
typedef CharType Ch;
//! Constructor.
/*!
\param os output stream to be wrapped.
\param type UTF encoding type.
\param putBOM Whether to write BOM at the beginning of the stream.
*/
AutoUTFOutputStream(OutputByteStream& os, UTFType type, bool putBOM) : os_(&os), type_(type) {
RAPIDJSON_ASSERT(type >= kUTF8 && type <= kUTF32BE);
// Runtime check whether the size of character type is sufficient. It only perform checks with assertion.
if (type_ == kUTF16LE || type_ == kUTF16BE) RAPIDJSON_ASSERT(sizeof(Ch) >= 2);
if (type_ == kUTF32LE || type_ == kUTF32BE) RAPIDJSON_ASSERT(sizeof(Ch) >= 4);
static const PutFunc f[] = { RAPIDJSON_ENCODINGS_FUNC(Put) };
putFunc_ = f[type_];
if (putBOM)
PutBOM();
}
UTFType GetType() const { return type_; }
void Put(Ch c) { putFunc_(*os_, c); }
void Flush() { os_->Flush(); }
// Not implemented
Ch Peek() const { RAPIDJSON_ASSERT(false); return 0;}
Ch Take() { RAPIDJSON_ASSERT(false); return 0;}
size_t Tell() const { RAPIDJSON_ASSERT(false); return 0; }
Ch* PutBegin() { RAPIDJSON_ASSERT(false); return 0; }
size_t PutEnd(Ch*) { RAPIDJSON_ASSERT(false); return 0; }
private:
AutoUTFOutputStream(const AutoUTFOutputStream&);
AutoUTFOutputStream& operator=(const AutoUTFOutputStream&);
void PutBOM() {
typedef void (*PutBOMFunc)(OutputByteStream&);
static const PutBOMFunc f[] = { RAPIDJSON_ENCODINGS_FUNC(PutBOM) };
f[type_](*os_);
}
typedef void (*PutFunc)(OutputByteStream&, Ch);
OutputByteStream* os_;
UTFType type_;
PutFunc putFunc_;
};
#undef RAPIDJSON_ENCODINGS_FUNC
RAPIDJSON_NAMESPACE_END
#ifdef __clang__
RAPIDJSON_DIAG_POP
#endif
#ifdef __GNUC__
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_FILESTREAM_H_

716
rapidjson/encodings.h Normal file
View File

@ -0,0 +1,716 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_ENCODINGS_H_
#define RAPIDJSON_ENCODINGS_H_
#include "rapidjson.h"
#ifdef _MSC_VER
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(4244) // conversion from 'type1' to 'type2', possible loss of data
RAPIDJSON_DIAG_OFF(4702) // unreachable code
#elif defined(__GNUC__)
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(effc++)
RAPIDJSON_DIAG_OFF(overflow)
#endif
RAPIDJSON_NAMESPACE_BEGIN
///////////////////////////////////////////////////////////////////////////////
// Encoding
/*! \class rapidjson::Encoding
\brief Concept for encoding of Unicode characters.
\code
concept Encoding {
typename Ch; //! Type of character. A "character" is actually a code unit in unicode's definition.
enum { supportUnicode = 1 }; // or 0 if not supporting unicode
//! \brief Encode a Unicode codepoint to an output stream.
//! \param os Output stream.
//! \param codepoint An unicode codepoint, ranging from 0x0 to 0x10FFFF inclusively.
template<typename OutputStream>
static void Encode(OutputStream& os, unsigned codepoint);
//! \brief Decode a Unicode codepoint from an input stream.
//! \param is Input stream.
//! \param codepoint Output of the unicode codepoint.
//! \return true if a valid codepoint can be decoded from the stream.
template <typename InputStream>
static bool Decode(InputStream& is, unsigned* codepoint);
//! \brief Validate one Unicode codepoint from an encoded stream.
//! \param is Input stream to obtain codepoint.
//! \param os Output for copying one codepoint.
//! \return true if it is valid.
//! \note This function just validating and copying the codepoint without actually decode it.
template <typename InputStream, typename OutputStream>
static bool Validate(InputStream& is, OutputStream& os);
// The following functions are deal with byte streams.
//! Take a character from input byte stream, skip BOM if exist.
template <typename InputByteStream>
static CharType TakeBOM(InputByteStream& is);
//! Take a character from input byte stream.
template <typename InputByteStream>
static Ch Take(InputByteStream& is);
//! Put BOM to output byte stream.
template <typename OutputByteStream>
static void PutBOM(OutputByteStream& os);
//! Put a character to output byte stream.
template <typename OutputByteStream>
static void Put(OutputByteStream& os, Ch c);
};
\endcode
*/
///////////////////////////////////////////////////////////////////////////////
// UTF8
//! UTF-8 encoding.
/*! http://en.wikipedia.org/wiki/UTF-8
http://tools.ietf.org/html/rfc3629
\tparam CharType Code unit for storing 8-bit UTF-8 data. Default is char.
\note implements Encoding concept
*/
template<typename CharType = char>
struct UTF8 {
typedef CharType Ch;
enum { supportUnicode = 1 };
template<typename OutputStream>
static void Encode(OutputStream& os, unsigned codepoint) {
if (codepoint <= 0x7F)
os.Put(static_cast<Ch>(codepoint & 0xFF));
else if (codepoint <= 0x7FF) {
os.Put(static_cast<Ch>(0xC0 | ((codepoint >> 6) & 0xFF)));
os.Put(static_cast<Ch>(0x80 | ((codepoint & 0x3F))));
}
else if (codepoint <= 0xFFFF) {
os.Put(static_cast<Ch>(0xE0 | ((codepoint >> 12) & 0xFF)));
os.Put(static_cast<Ch>(0x80 | ((codepoint >> 6) & 0x3F)));
os.Put(static_cast<Ch>(0x80 | (codepoint & 0x3F)));
}
else {
RAPIDJSON_ASSERT(codepoint <= 0x10FFFF);
os.Put(static_cast<Ch>(0xF0 | ((codepoint >> 18) & 0xFF)));
os.Put(static_cast<Ch>(0x80 | ((codepoint >> 12) & 0x3F)));
os.Put(static_cast<Ch>(0x80 | ((codepoint >> 6) & 0x3F)));
os.Put(static_cast<Ch>(0x80 | (codepoint & 0x3F)));
}
}
template<typename OutputStream>
static void EncodeUnsafe(OutputStream& os, unsigned codepoint) {
if (codepoint <= 0x7F)
PutUnsafe(os, static_cast<Ch>(codepoint & 0xFF));
else if (codepoint <= 0x7FF) {
PutUnsafe(os, static_cast<Ch>(0xC0 | ((codepoint >> 6) & 0xFF)));
PutUnsafe(os, static_cast<Ch>(0x80 | ((codepoint & 0x3F))));
}
else if (codepoint <= 0xFFFF) {
PutUnsafe(os, static_cast<Ch>(0xE0 | ((codepoint >> 12) & 0xFF)));
PutUnsafe(os, static_cast<Ch>(0x80 | ((codepoint >> 6) & 0x3F)));
PutUnsafe(os, static_cast<Ch>(0x80 | (codepoint & 0x3F)));
}
else {
RAPIDJSON_ASSERT(codepoint <= 0x10FFFF);
PutUnsafe(os, static_cast<Ch>(0xF0 | ((codepoint >> 18) & 0xFF)));
PutUnsafe(os, static_cast<Ch>(0x80 | ((codepoint >> 12) & 0x3F)));
PutUnsafe(os, static_cast<Ch>(0x80 | ((codepoint >> 6) & 0x3F)));
PutUnsafe(os, static_cast<Ch>(0x80 | (codepoint & 0x3F)));
}
}
template <typename InputStream>
static bool Decode(InputStream& is, unsigned* codepoint) {
#define COPY() c = is.Take(); *codepoint = (*codepoint << 6) | (static_cast<unsigned char>(c) & 0x3Fu)
#define TRANS(mask) result &= ((GetRange(static_cast<unsigned char>(c)) & mask) != 0)
#define TAIL() COPY(); TRANS(0x70)
typename InputStream::Ch c = is.Take();
if (!(c & 0x80)) {
*codepoint = static_cast<unsigned char>(c);
return true;
}
unsigned char type = GetRange(static_cast<unsigned char>(c));
if (type >= 32) {
*codepoint = 0;
} else {
*codepoint = (0xFF >> type) & static_cast<unsigned char>(c);
}
bool result = true;
switch (type) {
case 2: TAIL(); return result;
case 3: TAIL(); TAIL(); return result;
case 4: COPY(); TRANS(0x50); TAIL(); return result;
case 5: COPY(); TRANS(0x10); TAIL(); TAIL(); return result;
case 6: TAIL(); TAIL(); TAIL(); return result;
case 10: COPY(); TRANS(0x20); TAIL(); return result;
case 11: COPY(); TRANS(0x60); TAIL(); TAIL(); return result;
default: return false;
}
#undef COPY
#undef TRANS
#undef TAIL
}
template <typename InputStream, typename OutputStream>
static bool Validate(InputStream& is, OutputStream& os) {
#define COPY() os.Put(c = is.Take())
#define TRANS(mask) result &= ((GetRange(static_cast<unsigned char>(c)) & mask) != 0)
#define TAIL() COPY(); TRANS(0x70)
Ch c;
COPY();
if (!(c & 0x80))
return true;
bool result = true;
switch (GetRange(static_cast<unsigned char>(c))) {
case 2: TAIL(); return result;
case 3: TAIL(); TAIL(); return result;
case 4: COPY(); TRANS(0x50); TAIL(); return result;
case 5: COPY(); TRANS(0x10); TAIL(); TAIL(); return result;
case 6: TAIL(); TAIL(); TAIL(); return result;
case 10: COPY(); TRANS(0x20); TAIL(); return result;
case 11: COPY(); TRANS(0x60); TAIL(); TAIL(); return result;
default: return false;
}
#undef COPY
#undef TRANS
#undef TAIL
}
static unsigned char GetRange(unsigned char c) {
// Referring to DFA of http://bjoern.hoehrmann.de/utf-8/decoder/dfa/
// With new mapping 1 -> 0x10, 7 -> 0x20, 9 -> 0x40, such that AND operation can test multiple types.
static const unsigned char type[] = {
0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,
0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,
0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,
0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,0x20,
8,8,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
10,3,3,3,3,3,3,3,3,3,3,3,3,4,3,3, 11,6,6,6,5,8,8,8,8,8,8,8,8,8,8,8,
};
return type[c];
}
template <typename InputByteStream>
static CharType TakeBOM(InputByteStream& is) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
typename InputByteStream::Ch c = Take(is);
if (static_cast<unsigned char>(c) != 0xEFu) return c;
c = is.Take();
if (static_cast<unsigned char>(c) != 0xBBu) return c;
c = is.Take();
if (static_cast<unsigned char>(c) != 0xBFu) return c;
c = is.Take();
return c;
}
template <typename InputByteStream>
static Ch Take(InputByteStream& is) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
return static_cast<Ch>(is.Take());
}
template <typename OutputByteStream>
static void PutBOM(OutputByteStream& os) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
os.Put(static_cast<typename OutputByteStream::Ch>(0xEFu));
os.Put(static_cast<typename OutputByteStream::Ch>(0xBBu));
os.Put(static_cast<typename OutputByteStream::Ch>(0xBFu));
}
template <typename OutputByteStream>
static void Put(OutputByteStream& os, Ch c) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
os.Put(static_cast<typename OutputByteStream::Ch>(c));
}
};
///////////////////////////////////////////////////////////////////////////////
// UTF16
//! UTF-16 encoding.
/*! http://en.wikipedia.org/wiki/UTF-16
http://tools.ietf.org/html/rfc2781
\tparam CharType Type for storing 16-bit UTF-16 data. Default is wchar_t. C++11 may use char16_t instead.
\note implements Encoding concept
\note For in-memory access, no need to concern endianness. The code units and code points are represented by CPU's endianness.
For streaming, use UTF16LE and UTF16BE, which handle endianness.
*/
template<typename CharType = wchar_t>
struct UTF16 {
typedef CharType Ch;
RAPIDJSON_STATIC_ASSERT(sizeof(Ch) >= 2);
enum { supportUnicode = 1 };
template<typename OutputStream>
static void Encode(OutputStream& os, unsigned codepoint) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputStream::Ch) >= 2);
if (codepoint <= 0xFFFF) {
RAPIDJSON_ASSERT(codepoint < 0xD800 || codepoint > 0xDFFF); // Code point itself cannot be surrogate pair
os.Put(static_cast<typename OutputStream::Ch>(codepoint));
}
else {
RAPIDJSON_ASSERT(codepoint <= 0x10FFFF);
unsigned v = codepoint - 0x10000;
os.Put(static_cast<typename OutputStream::Ch>((v >> 10) | 0xD800));
os.Put((v & 0x3FF) | 0xDC00);
}
}
template<typename OutputStream>
static void EncodeUnsafe(OutputStream& os, unsigned codepoint) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputStream::Ch) >= 2);
if (codepoint <= 0xFFFF) {
RAPIDJSON_ASSERT(codepoint < 0xD800 || codepoint > 0xDFFF); // Code point itself cannot be surrogate pair
PutUnsafe(os, static_cast<typename OutputStream::Ch>(codepoint));
}
else {
RAPIDJSON_ASSERT(codepoint <= 0x10FFFF);
unsigned v = codepoint - 0x10000;
PutUnsafe(os, static_cast<typename OutputStream::Ch>((v >> 10) | 0xD800));
PutUnsafe(os, (v & 0x3FF) | 0xDC00);
}
}
template <typename InputStream>
static bool Decode(InputStream& is, unsigned* codepoint) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputStream::Ch) >= 2);
typename InputStream::Ch c = is.Take();
if (c < 0xD800 || c > 0xDFFF) {
*codepoint = static_cast<unsigned>(c);
return true;
}
else if (c <= 0xDBFF) {
*codepoint = (static_cast<unsigned>(c) & 0x3FF) << 10;
c = is.Take();
*codepoint |= (static_cast<unsigned>(c) & 0x3FF);
*codepoint += 0x10000;
return c >= 0xDC00 && c <= 0xDFFF;
}
return false;
}
template <typename InputStream, typename OutputStream>
static bool Validate(InputStream& is, OutputStream& os) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputStream::Ch) >= 2);
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputStream::Ch) >= 2);
typename InputStream::Ch c;
os.Put(static_cast<typename OutputStream::Ch>(c = is.Take()));
if (c < 0xD800 || c > 0xDFFF)
return true;
else if (c <= 0xDBFF) {
os.Put(c = is.Take());
return c >= 0xDC00 && c <= 0xDFFF;
}
return false;
}
};
//! UTF-16 little endian encoding.
template<typename CharType = wchar_t>
struct UTF16LE : UTF16<CharType> {
template <typename InputByteStream>
static CharType TakeBOM(InputByteStream& is) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
CharType c = Take(is);
return static_cast<uint16_t>(c) == 0xFEFFu ? Take(is) : c;
}
template <typename InputByteStream>
static CharType Take(InputByteStream& is) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
unsigned c = static_cast<uint8_t>(is.Take());
c |= static_cast<unsigned>(static_cast<uint8_t>(is.Take())) << 8;
return static_cast<CharType>(c);
}
template <typename OutputByteStream>
static void PutBOM(OutputByteStream& os) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
os.Put(static_cast<typename OutputByteStream::Ch>(0xFFu));
os.Put(static_cast<typename OutputByteStream::Ch>(0xFEu));
}
template <typename OutputByteStream>
static void Put(OutputByteStream& os, CharType c) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
os.Put(static_cast<typename OutputByteStream::Ch>(static_cast<unsigned>(c) & 0xFFu));
os.Put(static_cast<typename OutputByteStream::Ch>((static_cast<unsigned>(c) >> 8) & 0xFFu));
}
};
//! UTF-16 big endian encoding.
template<typename CharType = wchar_t>
struct UTF16BE : UTF16<CharType> {
template <typename InputByteStream>
static CharType TakeBOM(InputByteStream& is) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
CharType c = Take(is);
return static_cast<uint16_t>(c) == 0xFEFFu ? Take(is) : c;
}
template <typename InputByteStream>
static CharType Take(InputByteStream& is) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
unsigned c = static_cast<unsigned>(static_cast<uint8_t>(is.Take())) << 8;
c |= static_cast<uint8_t>(is.Take());
return static_cast<CharType>(c);
}
template <typename OutputByteStream>
static void PutBOM(OutputByteStream& os) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
os.Put(static_cast<typename OutputByteStream::Ch>(0xFEu));
os.Put(static_cast<typename OutputByteStream::Ch>(0xFFu));
}
template <typename OutputByteStream>
static void Put(OutputByteStream& os, CharType c) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
os.Put(static_cast<typename OutputByteStream::Ch>((static_cast<unsigned>(c) >> 8) & 0xFFu));
os.Put(static_cast<typename OutputByteStream::Ch>(static_cast<unsigned>(c) & 0xFFu));
}
};
///////////////////////////////////////////////////////////////////////////////
// UTF32
//! UTF-32 encoding.
/*! http://en.wikipedia.org/wiki/UTF-32
\tparam CharType Type for storing 32-bit UTF-32 data. Default is unsigned. C++11 may use char32_t instead.
\note implements Encoding concept
\note For in-memory access, no need to concern endianness. The code units and code points are represented by CPU's endianness.
For streaming, use UTF32LE and UTF32BE, which handle endianness.
*/
template<typename CharType = unsigned>
struct UTF32 {
typedef CharType Ch;
RAPIDJSON_STATIC_ASSERT(sizeof(Ch) >= 4);
enum { supportUnicode = 1 };
template<typename OutputStream>
static void Encode(OutputStream& os, unsigned codepoint) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputStream::Ch) >= 4);
RAPIDJSON_ASSERT(codepoint <= 0x10FFFF);
os.Put(codepoint);
}
template<typename OutputStream>
static void EncodeUnsafe(OutputStream& os, unsigned codepoint) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputStream::Ch) >= 4);
RAPIDJSON_ASSERT(codepoint <= 0x10FFFF);
PutUnsafe(os, codepoint);
}
template <typename InputStream>
static bool Decode(InputStream& is, unsigned* codepoint) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputStream::Ch) >= 4);
Ch c = is.Take();
*codepoint = c;
return c <= 0x10FFFF;
}
template <typename InputStream, typename OutputStream>
static bool Validate(InputStream& is, OutputStream& os) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputStream::Ch) >= 4);
Ch c;
os.Put(c = is.Take());
return c <= 0x10FFFF;
}
};
//! UTF-32 little endian enocoding.
template<typename CharType = unsigned>
struct UTF32LE : UTF32<CharType> {
template <typename InputByteStream>
static CharType TakeBOM(InputByteStream& is) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
CharType c = Take(is);
return static_cast<uint32_t>(c) == 0x0000FEFFu ? Take(is) : c;
}
template <typename InputByteStream>
static CharType Take(InputByteStream& is) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
unsigned c = static_cast<uint8_t>(is.Take());
c |= static_cast<unsigned>(static_cast<uint8_t>(is.Take())) << 8;
c |= static_cast<unsigned>(static_cast<uint8_t>(is.Take())) << 16;
c |= static_cast<unsigned>(static_cast<uint8_t>(is.Take())) << 24;
return static_cast<CharType>(c);
}
template <typename OutputByteStream>
static void PutBOM(OutputByteStream& os) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
os.Put(static_cast<typename OutputByteStream::Ch>(0xFFu));
os.Put(static_cast<typename OutputByteStream::Ch>(0xFEu));
os.Put(static_cast<typename OutputByteStream::Ch>(0x00u));
os.Put(static_cast<typename OutputByteStream::Ch>(0x00u));
}
template <typename OutputByteStream>
static void Put(OutputByteStream& os, CharType c) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
os.Put(static_cast<typename OutputByteStream::Ch>(c & 0xFFu));
os.Put(static_cast<typename OutputByteStream::Ch>((c >> 8) & 0xFFu));
os.Put(static_cast<typename OutputByteStream::Ch>((c >> 16) & 0xFFu));
os.Put(static_cast<typename OutputByteStream::Ch>((c >> 24) & 0xFFu));
}
};
//! UTF-32 big endian encoding.
template<typename CharType = unsigned>
struct UTF32BE : UTF32<CharType> {
template <typename InputByteStream>
static CharType TakeBOM(InputByteStream& is) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
CharType c = Take(is);
return static_cast<uint32_t>(c) == 0x0000FEFFu ? Take(is) : c;
}
template <typename InputByteStream>
static CharType Take(InputByteStream& is) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
unsigned c = static_cast<unsigned>(static_cast<uint8_t>(is.Take())) << 24;
c |= static_cast<unsigned>(static_cast<uint8_t>(is.Take())) << 16;
c |= static_cast<unsigned>(static_cast<uint8_t>(is.Take())) << 8;
c |= static_cast<unsigned>(static_cast<uint8_t>(is.Take()));
return static_cast<CharType>(c);
}
template <typename OutputByteStream>
static void PutBOM(OutputByteStream& os) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
os.Put(static_cast<typename OutputByteStream::Ch>(0x00u));
os.Put(static_cast<typename OutputByteStream::Ch>(0x00u));
os.Put(static_cast<typename OutputByteStream::Ch>(0xFEu));
os.Put(static_cast<typename OutputByteStream::Ch>(0xFFu));
}
template <typename OutputByteStream>
static void Put(OutputByteStream& os, CharType c) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
os.Put(static_cast<typename OutputByteStream::Ch>((c >> 24) & 0xFFu));
os.Put(static_cast<typename OutputByteStream::Ch>((c >> 16) & 0xFFu));
os.Put(static_cast<typename OutputByteStream::Ch>((c >> 8) & 0xFFu));
os.Put(static_cast<typename OutputByteStream::Ch>(c & 0xFFu));
}
};
///////////////////////////////////////////////////////////////////////////////
// ASCII
//! ASCII encoding.
/*! http://en.wikipedia.org/wiki/ASCII
\tparam CharType Code unit for storing 7-bit ASCII data. Default is char.
\note implements Encoding concept
*/
template<typename CharType = char>
struct ASCII {
typedef CharType Ch;
enum { supportUnicode = 0 };
template<typename OutputStream>
static void Encode(OutputStream& os, unsigned codepoint) {
RAPIDJSON_ASSERT(codepoint <= 0x7F);
os.Put(static_cast<Ch>(codepoint & 0xFF));
}
template<typename OutputStream>
static void EncodeUnsafe(OutputStream& os, unsigned codepoint) {
RAPIDJSON_ASSERT(codepoint <= 0x7F);
PutUnsafe(os, static_cast<Ch>(codepoint & 0xFF));
}
template <typename InputStream>
static bool Decode(InputStream& is, unsigned* codepoint) {
uint8_t c = static_cast<uint8_t>(is.Take());
*codepoint = c;
return c <= 0X7F;
}
template <typename InputStream, typename OutputStream>
static bool Validate(InputStream& is, OutputStream& os) {
uint8_t c = static_cast<uint8_t>(is.Take());
os.Put(static_cast<typename OutputStream::Ch>(c));
return c <= 0x7F;
}
template <typename InputByteStream>
static CharType TakeBOM(InputByteStream& is) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
uint8_t c = static_cast<uint8_t>(Take(is));
return static_cast<Ch>(c);
}
template <typename InputByteStream>
static Ch Take(InputByteStream& is) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename InputByteStream::Ch) == 1);
return static_cast<Ch>(is.Take());
}
template <typename OutputByteStream>
static void PutBOM(OutputByteStream& os) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
(void)os;
}
template <typename OutputByteStream>
static void Put(OutputByteStream& os, Ch c) {
RAPIDJSON_STATIC_ASSERT(sizeof(typename OutputByteStream::Ch) == 1);
os.Put(static_cast<typename OutputByteStream::Ch>(c));
}
};
///////////////////////////////////////////////////////////////////////////////
// AutoUTF
//! Runtime-specified UTF encoding type of a stream.
enum UTFType {
kUTF8 = 0, //!< UTF-8.
kUTF16LE = 1, //!< UTF-16 little endian.
kUTF16BE = 2, //!< UTF-16 big endian.
kUTF32LE = 3, //!< UTF-32 little endian.
kUTF32BE = 4 //!< UTF-32 big endian.
};
//! Dynamically select encoding according to stream's runtime-specified UTF encoding type.
/*! \note This class can be used with AutoUTFInputtStream and AutoUTFOutputStream, which provides GetType().
*/
template<typename CharType>
struct AutoUTF {
typedef CharType Ch;
enum { supportUnicode = 1 };
#define RAPIDJSON_ENCODINGS_FUNC(x) UTF8<Ch>::x, UTF16LE<Ch>::x, UTF16BE<Ch>::x, UTF32LE<Ch>::x, UTF32BE<Ch>::x
template<typename OutputStream>
RAPIDJSON_FORCEINLINE static void Encode(OutputStream& os, unsigned codepoint) {
typedef void (*EncodeFunc)(OutputStream&, unsigned);
static const EncodeFunc f[] = { RAPIDJSON_ENCODINGS_FUNC(Encode) };
(*f[os.GetType()])(os, codepoint);
}
template<typename OutputStream>
RAPIDJSON_FORCEINLINE static void EncodeUnsafe(OutputStream& os, unsigned codepoint) {
typedef void (*EncodeFunc)(OutputStream&, unsigned);
static const EncodeFunc f[] = { RAPIDJSON_ENCODINGS_FUNC(EncodeUnsafe) };
(*f[os.GetType()])(os, codepoint);
}
template <typename InputStream>
RAPIDJSON_FORCEINLINE static bool Decode(InputStream& is, unsigned* codepoint) {
typedef bool (*DecodeFunc)(InputStream&, unsigned*);
static const DecodeFunc f[] = { RAPIDJSON_ENCODINGS_FUNC(Decode) };
return (*f[is.GetType()])(is, codepoint);
}
template <typename InputStream, typename OutputStream>
RAPIDJSON_FORCEINLINE static bool Validate(InputStream& is, OutputStream& os) {
typedef bool (*ValidateFunc)(InputStream&, OutputStream&);
static const ValidateFunc f[] = { RAPIDJSON_ENCODINGS_FUNC(Validate) };
return (*f[is.GetType()])(is, os);
}
#undef RAPIDJSON_ENCODINGS_FUNC
};
///////////////////////////////////////////////////////////////////////////////
// Transcoder
//! Encoding conversion.
template<typename SourceEncoding, typename TargetEncoding>
struct Transcoder {
//! Take one Unicode codepoint from source encoding, convert it to target encoding and put it to the output stream.
template<typename InputStream, typename OutputStream>
RAPIDJSON_FORCEINLINE static bool Transcode(InputStream& is, OutputStream& os) {
unsigned codepoint;
if (!SourceEncoding::Decode(is, &codepoint))
return false;
TargetEncoding::Encode(os, codepoint);
return true;
}
template<typename InputStream, typename OutputStream>
RAPIDJSON_FORCEINLINE static bool TranscodeUnsafe(InputStream& is, OutputStream& os) {
unsigned codepoint;
if (!SourceEncoding::Decode(is, &codepoint))
return false;
TargetEncoding::EncodeUnsafe(os, codepoint);
return true;
}
//! Validate one Unicode codepoint from an encoded stream.
template<typename InputStream, typename OutputStream>
RAPIDJSON_FORCEINLINE static bool Validate(InputStream& is, OutputStream& os) {
return Transcode(is, os); // Since source/target encoding is different, must transcode.
}
};
// Forward declaration.
template<typename Stream>
inline void PutUnsafe(Stream& stream, typename Stream::Ch c);
//! Specialization of Transcoder with same source and target encoding.
template<typename Encoding>
struct Transcoder<Encoding, Encoding> {
template<typename InputStream, typename OutputStream>
RAPIDJSON_FORCEINLINE static bool Transcode(InputStream& is, OutputStream& os) {
os.Put(is.Take()); // Just copy one code unit. This semantic is different from primary template class.
return true;
}
template<typename InputStream, typename OutputStream>
RAPIDJSON_FORCEINLINE static bool TranscodeUnsafe(InputStream& is, OutputStream& os) {
PutUnsafe(os, is.Take()); // Just copy one code unit. This semantic is different from primary template class.
return true;
}
template<typename InputStream, typename OutputStream>
RAPIDJSON_FORCEINLINE static bool Validate(InputStream& is, OutputStream& os) {
return Encoding::Validate(is, os); // source/target encoding are the same
}
};
RAPIDJSON_NAMESPACE_END
#if defined(__GNUC__) || defined(_MSC_VER)
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_ENCODINGS_H_

74
rapidjson/error/en.h Normal file
View File

@ -0,0 +1,74 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_ERROR_EN_H_
#define RAPIDJSON_ERROR_EN_H_
#include "error.h"
#ifdef __clang__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(switch-enum)
RAPIDJSON_DIAG_OFF(covered-switch-default)
#endif
RAPIDJSON_NAMESPACE_BEGIN
//! Maps error code of parsing into error message.
/*!
\ingroup RAPIDJSON_ERRORS
\param parseErrorCode Error code obtained in parsing.
\return the error message.
\note User can make a copy of this function for localization.
Using switch-case is safer for future modification of error codes.
*/
inline const RAPIDJSON_ERROR_CHARTYPE* GetParseError_En(ParseErrorCode parseErrorCode) {
switch (parseErrorCode) {
case kParseErrorNone: return RAPIDJSON_ERROR_STRING("No error.");
case kParseErrorDocumentEmpty: return RAPIDJSON_ERROR_STRING("The document is empty.");
case kParseErrorDocumentRootNotSingular: return RAPIDJSON_ERROR_STRING("The document root must not be followed by other values.");
case kParseErrorValueInvalid: return RAPIDJSON_ERROR_STRING("Invalid value.");
case kParseErrorObjectMissName: return RAPIDJSON_ERROR_STRING("Missing a name for object member.");
case kParseErrorObjectMissColon: return RAPIDJSON_ERROR_STRING("Missing a colon after a name of object member.");
case kParseErrorObjectMissCommaOrCurlyBracket: return RAPIDJSON_ERROR_STRING("Missing a comma or '}' after an object member.");
case kParseErrorArrayMissCommaOrSquareBracket: return RAPIDJSON_ERROR_STRING("Missing a comma or ']' after an array element.");
case kParseErrorStringUnicodeEscapeInvalidHex: return RAPIDJSON_ERROR_STRING("Incorrect hex digit after \\u escape in string.");
case kParseErrorStringUnicodeSurrogateInvalid: return RAPIDJSON_ERROR_STRING("The surrogate pair in string is invalid.");
case kParseErrorStringEscapeInvalid: return RAPIDJSON_ERROR_STRING("Invalid escape character in string.");
case kParseErrorStringMissQuotationMark: return RAPIDJSON_ERROR_STRING("Missing a closing quotation mark in string.");
case kParseErrorStringInvalidEncoding: return RAPIDJSON_ERROR_STRING("Invalid encoding in string.");
case kParseErrorNumberTooBig: return RAPIDJSON_ERROR_STRING("Number too big to be stored in double.");
case kParseErrorNumberMissFraction: return RAPIDJSON_ERROR_STRING("Miss fraction part in number.");
case kParseErrorNumberMissExponent: return RAPIDJSON_ERROR_STRING("Miss exponent in number.");
case kParseErrorTermination: return RAPIDJSON_ERROR_STRING("Terminate parsing due to Handler error.");
case kParseErrorUnspecificSyntaxError: return RAPIDJSON_ERROR_STRING("Unspecific syntax error.");
default: return RAPIDJSON_ERROR_STRING("Unknown error.");
}
}
RAPIDJSON_NAMESPACE_END
#ifdef __clang__
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_ERROR_EN_H_

155
rapidjson/error/error.h Normal file
View File

@ -0,0 +1,155 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_ERROR_ERROR_H_
#define RAPIDJSON_ERROR_ERROR_H_
#include "../rapidjson.h"
#ifdef __clang__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(padded)
#endif
/*! \file error.h */
/*! \defgroup RAPIDJSON_ERRORS RapidJSON error handling */
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_ERROR_CHARTYPE
//! Character type of error messages.
/*! \ingroup RAPIDJSON_ERRORS
The default character type is \c char.
On Windows, user can define this macro as \c TCHAR for supporting both
unicode/non-unicode settings.
*/
#ifndef RAPIDJSON_ERROR_CHARTYPE
#define RAPIDJSON_ERROR_CHARTYPE char
#endif
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_ERROR_STRING
//! Macro for converting string literial to \ref RAPIDJSON_ERROR_CHARTYPE[].
/*! \ingroup RAPIDJSON_ERRORS
By default this conversion macro does nothing.
On Windows, user can define this macro as \c _T(x) for supporting both
unicode/non-unicode settings.
*/
#ifndef RAPIDJSON_ERROR_STRING
#define RAPIDJSON_ERROR_STRING(x) x
#endif
RAPIDJSON_NAMESPACE_BEGIN
///////////////////////////////////////////////////////////////////////////////
// ParseErrorCode
//! Error code of parsing.
/*! \ingroup RAPIDJSON_ERRORS
\see GenericReader::Parse, GenericReader::GetParseErrorCode
*/
enum ParseErrorCode {
kParseErrorNone = 0, //!< No error.
kParseErrorDocumentEmpty, //!< The document is empty.
kParseErrorDocumentRootNotSingular, //!< The document root must not follow by other values.
kParseErrorValueInvalid, //!< Invalid value.
kParseErrorObjectMissName, //!< Missing a name for object member.
kParseErrorObjectMissColon, //!< Missing a colon after a name of object member.
kParseErrorObjectMissCommaOrCurlyBracket, //!< Missing a comma or '}' after an object member.
kParseErrorArrayMissCommaOrSquareBracket, //!< Missing a comma or ']' after an array element.
kParseErrorStringUnicodeEscapeInvalidHex, //!< Incorrect hex digit after \\u escape in string.
kParseErrorStringUnicodeSurrogateInvalid, //!< The surrogate pair in string is invalid.
kParseErrorStringEscapeInvalid, //!< Invalid escape character in string.
kParseErrorStringMissQuotationMark, //!< Missing a closing quotation mark in string.
kParseErrorStringInvalidEncoding, //!< Invalid encoding in string.
kParseErrorNumberTooBig, //!< Number too big to be stored in double.
kParseErrorNumberMissFraction, //!< Miss fraction part in number.
kParseErrorNumberMissExponent, //!< Miss exponent in number.
kParseErrorTermination, //!< Parsing was terminated.
kParseErrorUnspecificSyntaxError //!< Unspecific syntax error.
};
//! Result of parsing (wraps ParseErrorCode)
/*!
\ingroup RAPIDJSON_ERRORS
\code
Document doc;
ParseResult ok = doc.Parse("[42]");
if (!ok) {
fprintf(stderr, "JSON parse error: %s (%u)",
GetParseError_En(ok.Code()), ok.Offset());
exit(EXIT_FAILURE);
}
\endcode
\see GenericReader::Parse, GenericDocument::Parse
*/
struct ParseResult {
public:
//! Default constructor, no error.
ParseResult() : code_(kParseErrorNone), offset_(0) {}
//! Constructor to set an error.
ParseResult(ParseErrorCode code, size_t offset) : code_(code), offset_(offset) {}
//! Get the error code.
ParseErrorCode Code() const { return code_; }
//! Get the error offset, if \ref IsError(), 0 otherwise.
size_t Offset() const { return offset_; }
//! Conversion to \c bool, returns \c true, iff !\ref IsError().
operator bool() const { return !IsError(); }
//! Whether the result is an error.
bool IsError() const { return code_ != kParseErrorNone; }
bool operator==(const ParseResult& that) const { return code_ == that.code_; }
bool operator==(ParseErrorCode code) const { return code_ == code; }
friend bool operator==(ParseErrorCode code, const ParseResult & err) { return code == err.code_; }
//! Reset error code.
void Clear() { Set(kParseErrorNone); }
//! Update error code and offset.
void Set(ParseErrorCode code, size_t offset = 0) { code_ = code; offset_ = offset; }
private:
ParseErrorCode code_;
size_t offset_;
};
//! Function pointer type of GetParseError().
/*! \ingroup RAPIDJSON_ERRORS
This is the prototype for \c GetParseError_X(), where \c X is a locale.
User can dynamically change locale in runtime, e.g.:
\code
GetParseErrorFunc GetParseError = GetParseError_En; // or whatever
const RAPIDJSON_ERROR_CHARTYPE* s = GetParseError(document.GetParseErrorCode());
\endcode
*/
typedef const RAPIDJSON_ERROR_CHARTYPE* (*GetParseErrorFunc)(ParseErrorCode);
RAPIDJSON_NAMESPACE_END
#ifdef __clang__
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_ERROR_ERROR_H_

99
rapidjson/filereadstream.h Normal file
View File

@ -0,0 +1,99 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_FILEREADSTREAM_H_
#define RAPIDJSON_FILEREADSTREAM_H_
#include "stream.h"
#include <cstdio>
#ifdef __clang__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(padded)
RAPIDJSON_DIAG_OFF(unreachable-code)
RAPIDJSON_DIAG_OFF(missing-noreturn)
#endif
RAPIDJSON_NAMESPACE_BEGIN
//! File byte stream for input using fread().
/*!
\note implements Stream concept
*/
class FileReadStream {
public:
typedef char Ch; //!< Character type (byte).
//! Constructor.
/*!
\param fp File pointer opened for read.
\param buffer user-supplied buffer.
\param bufferSize size of buffer in bytes. Must >=4 bytes.
*/
FileReadStream(std::FILE* fp, char* buffer, size_t bufferSize) : fp_(fp), buffer_(buffer), bufferSize_(bufferSize), bufferLast_(0), current_(buffer_), readCount_(0), count_(0), eof_(false) {
RAPIDJSON_ASSERT(fp_ != 0);
RAPIDJSON_ASSERT(bufferSize >= 4);
Read();
}
Ch Peek() const { return *current_; }
Ch Take() { Ch c = *current_; Read(); return c; }
size_t Tell() const { return count_ + static_cast<size_t>(current_ - buffer_); }
// Not implemented
void Put(Ch) { RAPIDJSON_ASSERT(false); }
void Flush() { RAPIDJSON_ASSERT(false); }
Ch* PutBegin() { RAPIDJSON_ASSERT(false); return 0; }
size_t PutEnd(Ch*) { RAPIDJSON_ASSERT(false); return 0; }
// For encoding detection only.
const Ch* Peek4() const {
return (current_ + 4 <= bufferLast_) ? current_ : 0;
}
private:
void Read() {
if (current_ < bufferLast_)
++current_;
else if (!eof_) {
count_ += readCount_;
readCount_ = fread(buffer_, 1, bufferSize_, fp_);
bufferLast_ = buffer_ + readCount_ - 1;
current_ = buffer_;
if (readCount_ < bufferSize_) {
buffer_[readCount_] = '\0';
++bufferLast_;
eof_ = true;
}
}
}
std::FILE* fp_;
Ch *buffer_;
size_t bufferSize_;
Ch *bufferLast_;
Ch *current_;
size_t readCount_;
size_t count_; //!< Number of characters read
bool eof_;
};
RAPIDJSON_NAMESPACE_END
#ifdef __clang__
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_FILESTREAM_H_

104
rapidjson/filewritestream.h Normal file
View File

@ -0,0 +1,104 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_FILEWRITESTREAM_H_
#define RAPIDJSON_FILEWRITESTREAM_H_
#include "stream.h"
#include <cstdio>
#ifdef __clang__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(unreachable-code)
#endif
RAPIDJSON_NAMESPACE_BEGIN
//! Wrapper of C file stream for input using fread().
/*!
\note implements Stream concept
*/
class FileWriteStream {
public:
typedef char Ch; //!< Character type. Only support char.
FileWriteStream(std::FILE* fp, char* buffer, size_t bufferSize) : fp_(fp), buffer_(buffer), bufferEnd_(buffer + bufferSize), current_(buffer_) {
RAPIDJSON_ASSERT(fp_ != 0);
}
void Put(char c) {
if (current_ >= bufferEnd_)
Flush();
*current_++ = c;
}
void PutN(char c, size_t n) {
size_t avail = static_cast<size_t>(bufferEnd_ - current_);
while (n > avail) {
std::memset(current_, c, avail);
current_ += avail;
Flush();
n -= avail;
avail = static_cast<size_t>(bufferEnd_ - current_);
}
if (n > 0) {
std::memset(current_, c, n);
current_ += n;
}
}
void Flush() {
if (current_ != buffer_) {
size_t result = fwrite(buffer_, 1, static_cast<size_t>(current_ - buffer_), fp_);
if (result < static_cast<size_t>(current_ - buffer_)) {
// failure deliberately ignored at this time
// added to avoid warn_unused_result build errors
}
current_ = buffer_;
}
}
// Not implemented
char Peek() const { RAPIDJSON_ASSERT(false); return 0; }
char Take() { RAPIDJSON_ASSERT(false); return 0; }
size_t Tell() const { RAPIDJSON_ASSERT(false); return 0; }
char* PutBegin() { RAPIDJSON_ASSERT(false); return 0; }
size_t PutEnd(char*) { RAPIDJSON_ASSERT(false); return 0; }
private:
// Prohibit copy constructor & assignment operator.
FileWriteStream(const FileWriteStream&);
FileWriteStream& operator=(const FileWriteStream&);
std::FILE* fp_;
char *buffer_;
char *bufferEnd_;
char *current_;
};
//! Implement specialized version of PutN() with memset() for better performance.
template<>
inline void PutN(FileWriteStream& stream, char c, size_t n) {
stream.PutN(c, n);
}
RAPIDJSON_NAMESPACE_END
#ifdef __clang__
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_FILESTREAM_H_

151
rapidjson/fwd.h Normal file
View File

@ -0,0 +1,151 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_FWD_H_
#define RAPIDJSON_FWD_H_
#include "rapidjson.h"
RAPIDJSON_NAMESPACE_BEGIN
// encodings.h
template<typename CharType> struct UTF8;
template<typename CharType> struct UTF16;
template<typename CharType> struct UTF16BE;
template<typename CharType> struct UTF16LE;
template<typename CharType> struct UTF32;
template<typename CharType> struct UTF32BE;
template<typename CharType> struct UTF32LE;
template<typename CharType> struct ASCII;
template<typename CharType> struct AutoUTF;
template<typename SourceEncoding, typename TargetEncoding>
struct Transcoder;
// allocators.h
class CrtAllocator;
template <typename BaseAllocator>
class MemoryPoolAllocator;
// stream.h
template <typename Encoding>
struct GenericStringStream;
typedef GenericStringStream<UTF8<char> > StringStream;
template <typename Encoding>
struct GenericInsituStringStream;
typedef GenericInsituStringStream<UTF8<char> > InsituStringStream;
// stringbuffer.h
template <typename Encoding, typename Allocator>
class GenericStringBuffer;
typedef GenericStringBuffer<UTF8<char>, CrtAllocator> StringBuffer;
// filereadstream.h
class FileReadStream;
// filewritestream.h
class FileWriteStream;
// memorybuffer.h
template <typename Allocator>
struct GenericMemoryBuffer;
typedef GenericMemoryBuffer<CrtAllocator> MemoryBuffer;
// memorystream.h
struct MemoryStream;
// reader.h
template<typename Encoding, typename Derived>
struct BaseReaderHandler;
template <typename SourceEncoding, typename TargetEncoding, typename StackAllocator>
class GenericReader;
typedef GenericReader<UTF8<char>, UTF8<char>, CrtAllocator> Reader;
// writer.h
template<typename OutputStream, typename SourceEncoding, typename TargetEncoding, typename StackAllocator, unsigned writeFlags>
class Writer;
// prettywriter.h
template<typename OutputStream, typename SourceEncoding, typename TargetEncoding, typename StackAllocator, unsigned writeFlags>
class PrettyWriter;
// document.h
template <typename Encoding, typename Allocator>
struct GenericMember;
template <bool Const, typename Encoding, typename Allocator>
class GenericMemberIterator;
template<typename CharType>
struct GenericStringRef;
template <typename Encoding, typename Allocator>
class GenericValue;
typedef GenericValue<UTF8<char>, MemoryPoolAllocator<CrtAllocator> > Value;
template <typename Encoding, typename Allocator, typename StackAllocator>
class GenericDocument;
typedef GenericDocument<UTF8<char>, MemoryPoolAllocator<CrtAllocator>, CrtAllocator> Document;
// pointer.h
template <typename ValueType, typename Allocator>
class GenericPointer;
typedef GenericPointer<Value, CrtAllocator> Pointer;
// schema.h
template <typename SchemaDocumentType>
class IGenericRemoteSchemaDocumentProvider;
template <typename ValueT, typename Allocator>
class GenericSchemaDocument;
typedef GenericSchemaDocument<Value, CrtAllocator> SchemaDocument;
typedef IGenericRemoteSchemaDocumentProvider<SchemaDocument> IRemoteSchemaDocumentProvider;
template <
typename SchemaDocumentType,
typename OutputHandler,
typename StateAllocator>
class GenericSchemaValidator;
typedef GenericSchemaValidator<SchemaDocument, BaseReaderHandler<UTF8<char>, void>, CrtAllocator> SchemaValidator;
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_RAPIDJSONFWD_H_

290
rapidjson/internal/biginteger.h Normal file
View File

@ -0,0 +1,290 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_BIGINTEGER_H_
#define RAPIDJSON_BIGINTEGER_H_
#include "../rapidjson.h"
#if defined(_MSC_VER) && defined(_M_AMD64)
#include <intrin.h> // for _umul128
#pragma intrinsic(_umul128)
#endif
RAPIDJSON_NAMESPACE_BEGIN
namespace internal {
class BigInteger {
public:
typedef uint64_t Type;
BigInteger(const BigInteger& rhs) : count_(rhs.count_) {
std::memcpy(digits_, rhs.digits_, count_ * sizeof(Type));
}
explicit BigInteger(uint64_t u) : count_(1) {
digits_[0] = u;
}
BigInteger(const char* decimals, size_t length) : count_(1) {
RAPIDJSON_ASSERT(length > 0);
digits_[0] = 0;
size_t i = 0;
const size_t kMaxDigitPerIteration = 19; // 2^64 = 18446744073709551616 > 10^19
while (length >= kMaxDigitPerIteration) {
AppendDecimal64(decimals + i, decimals + i + kMaxDigitPerIteration);
length -= kMaxDigitPerIteration;
i += kMaxDigitPerIteration;
}
if (length > 0)
AppendDecimal64(decimals + i, decimals + i + length);
}
BigInteger& operator=(const BigInteger &rhs)
{
if (this != &rhs) {
count_ = rhs.count_;
std::memcpy(digits_, rhs.digits_, count_ * sizeof(Type));
}
return *this;
}
BigInteger& operator=(uint64_t u) {
digits_[0] = u;
count_ = 1;
return *this;
}
BigInteger& operator+=(uint64_t u) {
Type backup = digits_[0];
digits_[0] += u;
for (size_t i = 0; i < count_ - 1; i++) {
if (digits_[i] >= backup)
return *this; // no carry
backup = digits_[i + 1];
digits_[i + 1] += 1;
}
// Last carry
if (digits_[count_ - 1] < backup)
PushBack(1);
return *this;
}
BigInteger& operator*=(uint64_t u) {
if (u == 0) return *this = 0;
if (u == 1) return *this;
if (*this == 1) return *this = u;
uint64_t k = 0;
for (size_t i = 0; i < count_; i++) {
uint64_t hi;
digits_[i] = MulAdd64(digits_[i], u, k, &hi);
k = hi;
}
if (k > 0)
PushBack(k);
return *this;
}
BigInteger& operator*=(uint32_t u) {
if (u == 0) return *this = 0;
if (u == 1) return *this;
if (*this == 1) return *this = u;
uint64_t k = 0;
for (size_t i = 0; i < count_; i++) {
const uint64_t c = digits_[i] >> 32;
const uint64_t d = digits_[i] & 0xFFFFFFFF;
const uint64_t uc = u * c;
const uint64_t ud = u * d;
const uint64_t p0 = ud + k;
const uint64_t p1 = uc + (p0 >> 32);
digits_[i] = (p0 & 0xFFFFFFFF) | (p1 << 32);
k = p1 >> 32;
}
if (k > 0)
PushBack(k);
return *this;
}
BigInteger& operator<<=(size_t shift) {
if (IsZero() || shift == 0) return *this;
size_t offset = shift / kTypeBit;
size_t interShift = shift % kTypeBit;
RAPIDJSON_ASSERT(count_ + offset <= kCapacity);
if (interShift == 0) {
std::memmove(&digits_[count_ - 1 + offset], &digits_[count_ - 1], count_ * sizeof(Type));
count_ += offset;
}
else {
digits_[count_] = 0;
for (size_t i = count_; i > 0; i--)
digits_[i + offset] = (digits_[i] << interShift) | (digits_[i - 1] >> (kTypeBit - interShift));
digits_[offset] = digits_[0] << interShift;
count_ += offset;
if (digits_[count_])
count_++;
}
std::memset(digits_, 0, offset * sizeof(Type));
return *this;
}
bool operator==(const BigInteger& rhs) const {
return count_ == rhs.count_ && std::memcmp(digits_, rhs.digits_, count_ * sizeof(Type)) == 0;
}
bool operator==(const Type rhs) const {
return count_ == 1 && digits_[0] == rhs;
}
BigInteger& MultiplyPow5(unsigned exp) {
static const uint32_t kPow5[12] = {
5,
5 * 5,
5 * 5 * 5,
5 * 5 * 5 * 5,
5 * 5 * 5 * 5 * 5,
5 * 5 * 5 * 5 * 5 * 5,
5 * 5 * 5 * 5 * 5 * 5 * 5,
5 * 5 * 5 * 5 * 5 * 5 * 5 * 5,
5 * 5 * 5 * 5 * 5 * 5 * 5 * 5 * 5,
5 * 5 * 5 * 5 * 5 * 5 * 5 * 5 * 5 * 5,
5 * 5 * 5 * 5 * 5 * 5 * 5 * 5 * 5 * 5 * 5,
5 * 5 * 5 * 5 * 5 * 5 * 5 * 5 * 5 * 5 * 5 * 5
};
if (exp == 0) return *this;
for (; exp >= 27; exp -= 27) *this *= RAPIDJSON_UINT64_C2(0X6765C793, 0XFA10079D); // 5^27
for (; exp >= 13; exp -= 13) *this *= static_cast<uint32_t>(1220703125u); // 5^13
if (exp > 0) *this *= kPow5[exp - 1];
return *this;
}
// Compute absolute difference of this and rhs.
// Assume this != rhs
bool Difference(const BigInteger& rhs, BigInteger* out) const {
int cmp = Compare(rhs);
RAPIDJSON_ASSERT(cmp != 0);
const BigInteger *a, *b; // Makes a > b
bool ret;
if (cmp < 0) { a = &rhs; b = this; ret = true; }
else { a = this; b = &rhs; ret = false; }
Type borrow = 0;
for (size_t i = 0; i < a->count_; i++) {
Type d = a->digits_[i] - borrow;
if (i < b->count_)
d -= b->digits_[i];
borrow = (d > a->digits_[i]) ? 1 : 0;
out->digits_[i] = d;
if (d != 0)
out->count_ = i + 1;
}
return ret;
}
int Compare(const BigInteger& rhs) const {
if (count_ != rhs.count_)
return count_ < rhs.count_ ? -1 : 1;
for (size_t i = count_; i-- > 0;)
if (digits_[i] != rhs.digits_[i])
return digits_[i] < rhs.digits_[i] ? -1 : 1;
return 0;
}
size_t GetCount() const { return count_; }
Type GetDigit(size_t index) const { RAPIDJSON_ASSERT(index < count_); return digits_[index]; }
bool IsZero() const { return count_ == 1 && digits_[0] == 0; }
private:
void AppendDecimal64(const char* begin, const char* end) {
uint64_t u = ParseUint64(begin, end);
if (IsZero())
*this = u;
else {
unsigned exp = static_cast<unsigned>(end - begin);
(MultiplyPow5(exp) <<= exp) += u; // *this = *this * 10^exp + u
}
}
void PushBack(Type digit) {
RAPIDJSON_ASSERT(count_ < kCapacity);
digits_[count_++] = digit;
}
static uint64_t ParseUint64(const char* begin, const char* end) {
uint64_t r = 0;
for (const char* p = begin; p != end; ++p) {
RAPIDJSON_ASSERT(*p >= '0' && *p <= '9');
r = r * 10u + static_cast<unsigned>(*p - '0');
}
return r;
}
// Assume a * b + k < 2^128
static uint64_t MulAdd64(uint64_t a, uint64_t b, uint64_t k, uint64_t* outHigh) {
#if defined(_MSC_VER) && defined(_M_AMD64)
uint64_t low = _umul128(a, b, outHigh) + k;
if (low < k)
(*outHigh)++;
return low;
#elif (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 6)) && defined(__x86_64__)
__extension__ typedef unsigned __int128 uint128;
uint128 p = static_cast<uint128>(a) * static_cast<uint128>(b);
p += k;
*outHigh = static_cast<uint64_t>(p >> 64);
return static_cast<uint64_t>(p);
#else
const uint64_t a0 = a & 0xFFFFFFFF, a1 = a >> 32, b0 = b & 0xFFFFFFFF, b1 = b >> 32;
uint64_t x0 = a0 * b0, x1 = a0 * b1, x2 = a1 * b0, x3 = a1 * b1;
x1 += (x0 >> 32); // can't give carry
x1 += x2;
if (x1 < x2)
x3 += (static_cast<uint64_t>(1) << 32);
uint64_t lo = (x1 << 32) + (x0 & 0xFFFFFFFF);
uint64_t hi = x3 + (x1 >> 32);
lo += k;
if (lo < k)
hi++;
*outHigh = hi;
return lo;
#endif
}
static const size_t kBitCount = 3328; // 64bit * 54 > 10^1000
static const size_t kCapacity = kBitCount / sizeof(Type);
static const size_t kTypeBit = sizeof(Type) * 8;
Type digits_[kCapacity];
size_t count_;
};
} // namespace internal
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_BIGINTEGER_H_

258
rapidjson/internal/diyfp.h Normal file
View File

@ -0,0 +1,258 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
// This is a C++ header-only implementation of Grisu2 algorithm from the publication:
// Loitsch, Florian. "Printing floating-point numbers quickly and accurately with
// integers." ACM Sigplan Notices 45.6 (2010): 233-243.
#ifndef RAPIDJSON_DIYFP_H_
#define RAPIDJSON_DIYFP_H_
#include "../rapidjson.h"
#if defined(_MSC_VER) && defined(_M_AMD64)
#include <intrin.h>
#pragma intrinsic(_BitScanReverse64)
#pragma intrinsic(_umul128)
#endif
RAPIDJSON_NAMESPACE_BEGIN
namespace internal {
#ifdef __GNUC__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(effc++)
#endif
#ifdef __clang__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(padded)
#endif
struct DiyFp {
DiyFp() : f(), e() {}
DiyFp(uint64_t fp, int exp) : f(fp), e(exp) {}
explicit DiyFp(double d) {
union {
double d;
uint64_t u64;
} u = { d };
int biased_e = static_cast<int>((u.u64 & kDpExponentMask) >> kDpSignificandSize);
uint64_t significand = (u.u64 & kDpSignificandMask);
if (biased_e != 0) {
f = significand + kDpHiddenBit;
e = biased_e - kDpExponentBias;
}
else {
f = significand;
e = kDpMinExponent + 1;
}
}
DiyFp operator-(const DiyFp& rhs) const {
return DiyFp(f - rhs.f, e);
}
DiyFp operator*(const DiyFp& rhs) const {
#if defined(_MSC_VER) && defined(_M_AMD64)
uint64_t h;
uint64_t l = _umul128(f, rhs.f, &h);
if (l & (uint64_t(1) << 63)) // rounding
h++;
return DiyFp(h, e + rhs.e + 64);
#elif (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 6)) && defined(__x86_64__)
__extension__ typedef unsigned __int128 uint128;
uint128 p = static_cast<uint128>(f) * static_cast<uint128>(rhs.f);
uint64_t h = static_cast<uint64_t>(p >> 64);
uint64_t l = static_cast<uint64_t>(p);
if (l & (uint64_t(1) << 63)) // rounding
h++;
return DiyFp(h, e + rhs.e + 64);
#else
const uint64_t M32 = 0xFFFFFFFF;
const uint64_t a = f >> 32;
const uint64_t b = f & M32;
const uint64_t c = rhs.f >> 32;
const uint64_t d = rhs.f & M32;
const uint64_t ac = a * c;
const uint64_t bc = b * c;
const uint64_t ad = a * d;
const uint64_t bd = b * d;
uint64_t tmp = (bd >> 32) + (ad & M32) + (bc & M32);
tmp += 1U << 31; /// mult_round
return DiyFp(ac + (ad >> 32) + (bc >> 32) + (tmp >> 32), e + rhs.e + 64);
#endif
}
DiyFp Normalize() const {
#if defined(_MSC_VER) && defined(_M_AMD64)
unsigned long index;
_BitScanReverse64(&index, f);
return DiyFp(f << (63 - index), e - (63 - index));
#elif defined(__GNUC__) && __GNUC__ >= 4
int s = __builtin_clzll(f);
return DiyFp(f << s, e - s);
#else
DiyFp res = *this;
while (!(res.f & (static_cast<uint64_t>(1) << 63))) {
res.f <<= 1;
res.e--;
}
return res;
#endif
}
DiyFp NormalizeBoundary() const {
DiyFp res = *this;
while (!(res.f & (kDpHiddenBit << 1))) {
res.f <<= 1;
res.e--;
}
res.f <<= (kDiySignificandSize - kDpSignificandSize - 2);
res.e = res.e - (kDiySignificandSize - kDpSignificandSize - 2);
return res;
}
void NormalizedBoundaries(DiyFp* minus, DiyFp* plus) const {
DiyFp pl = DiyFp((f << 1) + 1, e - 1).NormalizeBoundary();
DiyFp mi = (f == kDpHiddenBit) ? DiyFp((f << 2) - 1, e - 2) : DiyFp((f << 1) - 1, e - 1);
mi.f <<= mi.e - pl.e;
mi.e = pl.e;
*plus = pl;
*minus = mi;
}
double ToDouble() const {
union {
double d;
uint64_t u64;
}u;
const uint64_t be = (e == kDpDenormalExponent && (f & kDpHiddenBit) == 0) ? 0 :
static_cast<uint64_t>(e + kDpExponentBias);
u.u64 = (f & kDpSignificandMask) | (be << kDpSignificandSize);
return u.d;
}
static const int kDiySignificandSize = 64;
static const int kDpSignificandSize = 52;
static const int kDpExponentBias = 0x3FF + kDpSignificandSize;
static const int kDpMaxExponent = 0x7FF - kDpExponentBias;
static const int kDpMinExponent = -kDpExponentBias;
static const int kDpDenormalExponent = -kDpExponentBias + 1;
static const uint64_t kDpExponentMask = RAPIDJSON_UINT64_C2(0x7FF00000, 0x00000000);
static const uint64_t kDpSignificandMask = RAPIDJSON_UINT64_C2(0x000FFFFF, 0xFFFFFFFF);
static const uint64_t kDpHiddenBit = RAPIDJSON_UINT64_C2(0x00100000, 0x00000000);
uint64_t f;
int e;
};
inline DiyFp GetCachedPowerByIndex(size_t index) {
// 10^-348, 10^-340, ..., 10^340
static const uint64_t kCachedPowers_F[] = {
RAPIDJSON_UINT64_C2(0xfa8fd5a0, 0x081c0288), RAPIDJSON_UINT64_C2(0xbaaee17f, 0xa23ebf76),
RAPIDJSON_UINT64_C2(0x8b16fb20, 0x3055ac76), RAPIDJSON_UINT64_C2(0xcf42894a, 0x5dce35ea),
RAPIDJSON_UINT64_C2(0x9a6bb0aa, 0x55653b2d), RAPIDJSON_UINT64_C2(0xe61acf03, 0x3d1a45df),
RAPIDJSON_UINT64_C2(0xab70fe17, 0xc79ac6ca), RAPIDJSON_UINT64_C2(0xff77b1fc, 0xbebcdc4f),
RAPIDJSON_UINT64_C2(0xbe5691ef, 0x416bd60c), RAPIDJSON_UINT64_C2(0x8dd01fad, 0x907ffc3c),
RAPIDJSON_UINT64_C2(0xd3515c28, 0x31559a83), RAPIDJSON_UINT64_C2(0x9d71ac8f, 0xada6c9b5),
RAPIDJSON_UINT64_C2(0xea9c2277, 0x23ee8bcb), RAPIDJSON_UINT64_C2(0xaecc4991, 0x4078536d),
RAPIDJSON_UINT64_C2(0x823c1279, 0x5db6ce57), RAPIDJSON_UINT64_C2(0xc2109436, 0x4dfb5637),
RAPIDJSON_UINT64_C2(0x9096ea6f, 0x3848984f), RAPIDJSON_UINT64_C2(0xd77485cb, 0x25823ac7),
RAPIDJSON_UINT64_C2(0xa086cfcd, 0x97bf97f4), RAPIDJSON_UINT64_C2(0xef340a98, 0x172aace5),
RAPIDJSON_UINT64_C2(0xb23867fb, 0x2a35b28e), RAPIDJSON_UINT64_C2(0x84c8d4df, 0xd2c63f3b),
RAPIDJSON_UINT64_C2(0xc5dd4427, 0x1ad3cdba), RAPIDJSON_UINT64_C2(0x936b9fce, 0xbb25c996),
RAPIDJSON_UINT64_C2(0xdbac6c24, 0x7d62a584), RAPIDJSON_UINT64_C2(0xa3ab6658, 0x0d5fdaf6),
RAPIDJSON_UINT64_C2(0xf3e2f893, 0xdec3f126), RAPIDJSON_UINT64_C2(0xb5b5ada8, 0xaaff80b8),
RAPIDJSON_UINT64_C2(0x87625f05, 0x6c7c4a8b), RAPIDJSON_UINT64_C2(0xc9bcff60, 0x34c13053),
RAPIDJSON_UINT64_C2(0x964e858c, 0x91ba2655), RAPIDJSON_UINT64_C2(0xdff97724, 0x70297ebd),
RAPIDJSON_UINT64_C2(0xa6dfbd9f, 0xb8e5b88f), RAPIDJSON_UINT64_C2(0xf8a95fcf, 0x88747d94),
RAPIDJSON_UINT64_C2(0xb9447093, 0x8fa89bcf), RAPIDJSON_UINT64_C2(0x8a08f0f8, 0xbf0f156b),
RAPIDJSON_UINT64_C2(0xcdb02555, 0x653131b6), RAPIDJSON_UINT64_C2(0x993fe2c6, 0xd07b7fac),
RAPIDJSON_UINT64_C2(0xe45c10c4, 0x2a2b3b06), RAPIDJSON_UINT64_C2(0xaa242499, 0x697392d3),
RAPIDJSON_UINT64_C2(0xfd87b5f2, 0x8300ca0e), RAPIDJSON_UINT64_C2(0xbce50864, 0x92111aeb),
RAPIDJSON_UINT64_C2(0x8cbccc09, 0x6f5088cc), RAPIDJSON_UINT64_C2(0xd1b71758, 0xe219652c),
RAPIDJSON_UINT64_C2(0x9c400000, 0x00000000), RAPIDJSON_UINT64_C2(0xe8d4a510, 0x00000000),
RAPIDJSON_UINT64_C2(0xad78ebc5, 0xac620000), RAPIDJSON_UINT64_C2(0x813f3978, 0xf8940984),
RAPIDJSON_UINT64_C2(0xc097ce7b, 0xc90715b3), RAPIDJSON_UINT64_C2(0x8f7e32ce, 0x7bea5c70),
RAPIDJSON_UINT64_C2(0xd5d238a4, 0xabe98068), RAPIDJSON_UINT64_C2(0x9f4f2726, 0x179a2245),
RAPIDJSON_UINT64_C2(0xed63a231, 0xd4c4fb27), RAPIDJSON_UINT64_C2(0xb0de6538, 0x8cc8ada8),
RAPIDJSON_UINT64_C2(0x83c7088e, 0x1aab65db), RAPIDJSON_UINT64_C2(0xc45d1df9, 0x42711d9a),
RAPIDJSON_UINT64_C2(0x924d692c, 0xa61be758), RAPIDJSON_UINT64_C2(0xda01ee64, 0x1a708dea),
RAPIDJSON_UINT64_C2(0xa26da399, 0x9aef774a), RAPIDJSON_UINT64_C2(0xf209787b, 0xb47d6b85),
RAPIDJSON_UINT64_C2(0xb454e4a1, 0x79dd1877), RAPIDJSON_UINT64_C2(0x865b8692, 0x5b9bc5c2),
RAPIDJSON_UINT64_C2(0xc83553c5, 0xc8965d3d), RAPIDJSON_UINT64_C2(0x952ab45c, 0xfa97a0b3),
RAPIDJSON_UINT64_C2(0xde469fbd, 0x99a05fe3), RAPIDJSON_UINT64_C2(0xa59bc234, 0xdb398c25),
RAPIDJSON_UINT64_C2(0xf6c69a72, 0xa3989f5c), RAPIDJSON_UINT64_C2(0xb7dcbf53, 0x54e9bece),
RAPIDJSON_UINT64_C2(0x88fcf317, 0xf22241e2), RAPIDJSON_UINT64_C2(0xcc20ce9b, 0xd35c78a5),
RAPIDJSON_UINT64_C2(0x98165af3, 0x7b2153df), RAPIDJSON_UINT64_C2(0xe2a0b5dc, 0x971f303a),
RAPIDJSON_UINT64_C2(0xa8d9d153, 0x5ce3b396), RAPIDJSON_UINT64_C2(0xfb9b7cd9, 0xa4a7443c),
RAPIDJSON_UINT64_C2(0xbb764c4c, 0xa7a44410), RAPIDJSON_UINT64_C2(0x8bab8eef, 0xb6409c1a),
RAPIDJSON_UINT64_C2(0xd01fef10, 0xa657842c), RAPIDJSON_UINT64_C2(0x9b10a4e5, 0xe9913129),
RAPIDJSON_UINT64_C2(0xe7109bfb, 0xa19c0c9d), RAPIDJSON_UINT64_C2(0xac2820d9, 0x623bf429),
RAPIDJSON_UINT64_C2(0x80444b5e, 0x7aa7cf85), RAPIDJSON_UINT64_C2(0xbf21e440, 0x03acdd2d),
RAPIDJSON_UINT64_C2(0x8e679c2f, 0x5e44ff8f), RAPIDJSON_UINT64_C2(0xd433179d, 0x9c8cb841),
RAPIDJSON_UINT64_C2(0x9e19db92, 0xb4e31ba9), RAPIDJSON_UINT64_C2(0xeb96bf6e, 0xbadf77d9),
RAPIDJSON_UINT64_C2(0xaf87023b, 0x9bf0ee6b)
};
static const int16_t kCachedPowers_E[] = {
-1220, -1193, -1166, -1140, -1113, -1087, -1060, -1034, -1007, -980,
-954, -927, -901, -874, -847, -821, -794, -768, -741, -715,
-688, -661, -635, -608, -582, -555, -529, -502, -475, -449,
-422, -396, -369, -343, -316, -289, -263, -236, -210, -183,
-157, -130, -103, -77, -50, -24, 3, 30, 56, 83,
109, 136, 162, 189, 216, 242, 269, 295, 322, 348,
375, 402, 428, 455, 481, 508, 534, 561, 588, 614,
641, 667, 694, 720, 747, 774, 800, 827, 853, 880,
907, 933, 960, 986, 1013, 1039, 1066
};
return DiyFp(kCachedPowers_F[index], kCachedPowers_E[index]);
}
inline DiyFp GetCachedPower(int e, int* K) {
//int k = static_cast<int>(ceil((-61 - e) * 0.30102999566398114)) + 374;
double dk = (-61 - e) * 0.30102999566398114 + 347; // dk must be positive, so can do ceiling in positive
int k = static_cast<int>(dk);
if (dk - k > 0.0)
k++;
unsigned index = static_cast<unsigned>((k >> 3) + 1);
*K = -(-348 + static_cast<int>(index << 3)); // decimal exponent no need lookup table
return GetCachedPowerByIndex(index);
}
inline DiyFp GetCachedPower10(int exp, int *outExp) {
unsigned index = (static_cast<unsigned>(exp) + 348u) / 8u;
*outExp = -348 + static_cast<int>(index) * 8;
return GetCachedPowerByIndex(index);
}
#ifdef __GNUC__
RAPIDJSON_DIAG_POP
#endif
#ifdef __clang__
RAPIDJSON_DIAG_POP
RAPIDJSON_DIAG_OFF(padded)
#endif
} // namespace internal
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_DIYFP_H_

245
rapidjson/internal/dtoa.h Normal file
View File

@ -0,0 +1,245 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
// This is a C++ header-only implementation of Grisu2 algorithm from the publication:
// Loitsch, Florian. "Printing floating-point numbers quickly and accurately with
// integers." ACM Sigplan Notices 45.6 (2010): 233-243.
#ifndef RAPIDJSON_DTOA_
#define RAPIDJSON_DTOA_
#include "itoa.h" // GetDigitsLut()
#include "diyfp.h"
#include "ieee754.h"
RAPIDJSON_NAMESPACE_BEGIN
namespace internal {
#ifdef __GNUC__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(effc++)
RAPIDJSON_DIAG_OFF(array-bounds) // some gcc versions generate wrong warnings https://gcc.gnu.org/bugzilla/show_bug.cgi?id=59124
#endif
inline void GrisuRound(char* buffer, int len, uint64_t delta, uint64_t rest, uint64_t ten_kappa, uint64_t wp_w) {
while (rest < wp_w && delta - rest >= ten_kappa &&
(rest + ten_kappa < wp_w || /// closer
wp_w - rest > rest + ten_kappa - wp_w)) {
buffer[len - 1]--;
rest += ten_kappa;
}
}
inline unsigned CountDecimalDigit32(uint32_t n) {
// Simple pure C++ implementation was faster than __builtin_clz version in this situation.
if (n < 10) return 1;
if (n < 100) return 2;
if (n < 1000) return 3;
if (n < 10000) return 4;
if (n < 100000) return 5;
if (n < 1000000) return 6;
if (n < 10000000) return 7;
if (n < 100000000) return 8;
// Will not reach 10 digits in DigitGen()
//if (n < 1000000000) return 9;
//return 10;
return 9;
}
inline void DigitGen(const DiyFp& W, const DiyFp& Mp, uint64_t delta, char* buffer, int* len, int* K) {
static const uint32_t kPow10[] = { 1, 10, 100, 1000, 10000, 100000, 1000000, 10000000, 100000000, 1000000000 };
const DiyFp one(uint64_t(1) << -Mp.e, Mp.e);
const DiyFp wp_w = Mp - W;
uint32_t p1 = static_cast<uint32_t>(Mp.f >> -one.e);
uint64_t p2 = Mp.f & (one.f - 1);
unsigned kappa = CountDecimalDigit32(p1); // kappa in [0, 9]
*len = 0;
while (kappa > 0) {
uint32_t d = 0;
switch (kappa) {
case 9: d = p1 / 100000000; p1 %= 100000000; break;
case 8: d = p1 / 10000000; p1 %= 10000000; break;
case 7: d = p1 / 1000000; p1 %= 1000000; break;
case 6: d = p1 / 100000; p1 %= 100000; break;
case 5: d = p1 / 10000; p1 %= 10000; break;
case 4: d = p1 / 1000; p1 %= 1000; break;
case 3: d = p1 / 100; p1 %= 100; break;
case 2: d = p1 / 10; p1 %= 10; break;
case 1: d = p1; p1 = 0; break;
default:;
}
if (d || *len)
buffer[(*len)++] = static_cast<char>('0' + static_cast<char>(d));
kappa--;
uint64_t tmp = (static_cast<uint64_t>(p1) << -one.e) + p2;
if (tmp <= delta) {
*K += kappa;
GrisuRound(buffer, *len, delta, tmp, static_cast<uint64_t>(kPow10[kappa]) << -one.e, wp_w.f);
return;
}
}
// kappa = 0
for (;;) {
p2 *= 10;
delta *= 10;
char d = static_cast<char>(p2 >> -one.e);
if (d || *len)
buffer[(*len)++] = static_cast<char>('0' + d);
p2 &= one.f - 1;
kappa--;
if (p2 < delta) {
*K += kappa;
int index = -static_cast<int>(kappa);
GrisuRound(buffer, *len, delta, p2, one.f, wp_w.f * (index < 9 ? kPow10[-static_cast<int>(kappa)] : 0));
return;
}
}
}
inline void Grisu2(double value, char* buffer, int* length, int* K) {
const DiyFp v(value);
DiyFp w_m, w_p;
v.NormalizedBoundaries(&w_m, &w_p);
const DiyFp c_mk = GetCachedPower(w_p.e, K);
const DiyFp W = v.Normalize() * c_mk;
DiyFp Wp = w_p * c_mk;
DiyFp Wm = w_m * c_mk;
Wm.f++;
Wp.f--;
DigitGen(W, Wp, Wp.f - Wm.f, buffer, length, K);
}
inline char* WriteExponent(int K, char* buffer) {
if (K < 0) {
*buffer++ = '-';
K = -K;
}
if (K >= 100) {
*buffer++ = static_cast<char>('0' + static_cast<char>(K / 100));
K %= 100;
const char* d = GetDigitsLut() + K * 2;
*buffer++ = d[0];
*buffer++ = d[1];
}
else if (K >= 10) {
const char* d = GetDigitsLut() + K * 2;
*buffer++ = d[0];
*buffer++ = d[1];
}
else
*buffer++ = static_cast<char>('0' + static_cast<char>(K));
return buffer;
}
inline char* Prettify(char* buffer, int length, int k, int maxDecimalPlaces) {
const int kk = length + k; // 10^(kk-1) <= v < 10^kk
if (0 <= k && kk <= 21) {
// 1234e7 -> 12340000000
for (int i = length; i < kk; i++)
buffer[i] = '0';
buffer[kk] = '.';
buffer[kk + 1] = '0';
return &buffer[kk + 2];
}
else if (0 < kk && kk <= 21) {
// 1234e-2 -> 12.34
std::memmove(&buffer[kk + 1], &buffer[kk], static_cast<size_t>(length - kk));
buffer[kk] = '.';
if (0 > k + maxDecimalPlaces) {
// When maxDecimalPlaces = 2, 1.2345 -> 1.23, 1.102 -> 1.1
// Remove extra trailing zeros (at least one) after truncation.
for (int i = kk + maxDecimalPlaces; i > kk + 1; i--)
if (buffer[i] != '0')
return &buffer[i + 1];
return &buffer[kk + 2]; // Reserve one zero
}
else
return &buffer[length + 1];
}
else if (-6 < kk && kk <= 0) {
// 1234e-6 -> 0.001234
const int offset = 2 - kk;
std::memmove(&buffer[offset], &buffer[0], static_cast<size_t>(length));
buffer[0] = '0';
buffer[1] = '.';
for (int i = 2; i < offset; i++)
buffer[i] = '0';
if (length - kk > maxDecimalPlaces) {
// When maxDecimalPlaces = 2, 0.123 -> 0.12, 0.102 -> 0.1
// Remove extra trailing zeros (at least one) after truncation.
for (int i = maxDecimalPlaces + 1; i > 2; i--)
if (buffer[i] != '0')
return &buffer[i + 1];
return &buffer[3]; // Reserve one zero
}
else
return &buffer[length + offset];
}
else if (kk < -maxDecimalPlaces) {
// Truncate to zero
buffer[0] = '0';
buffer[1] = '.';
buffer[2] = '0';
return &buffer[3];
}
else if (length == 1) {
// 1e30
buffer[1] = 'e';
return WriteExponent(kk - 1, &buffer[2]);
}
else {
// 1234e30 -> 1.234e33
std::memmove(&buffer[2], &buffer[1], static_cast<size_t>(length - 1));
buffer[1] = '.';
buffer[length + 1] = 'e';
return WriteExponent(kk - 1, &buffer[0 + length + 2]);
}
}
inline char* dtoa(double value, char* buffer, int maxDecimalPlaces = 324) {
RAPIDJSON_ASSERT(maxDecimalPlaces >= 1);
Double d(value);
if (d.IsZero()) {
if (d.Sign())
*buffer++ = '-'; // -0.0, Issue #289
buffer[0] = '0';
buffer[1] = '.';
buffer[2] = '0';
return &buffer[3];
}
else {
if (value < 0) {
*buffer++ = '-';
value = -value;
}
int length, K;
Grisu2(value, buffer, &length, &K);
return Prettify(buffer, length, K, maxDecimalPlaces);
}
}
#ifdef __GNUC__
RAPIDJSON_DIAG_POP
#endif
} // namespace internal
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_DTOA_

78
rapidjson/internal/ieee754.h Normal file
View File

@ -0,0 +1,78 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_IEEE754_
#define RAPIDJSON_IEEE754_
#include "../rapidjson.h"
RAPIDJSON_NAMESPACE_BEGIN
namespace internal {
class Double {
public:
Double() {}
Double(double d) : d_(d) {}
Double(uint64_t u) : u_(u) {}
double Value() const { return d_; }
uint64_t Uint64Value() const { return u_; }
double NextPositiveDouble() const {
RAPIDJSON_ASSERT(!Sign());
return Double(u_ + 1).Value();
}
bool Sign() const { return (u_ & kSignMask) != 0; }
uint64_t Significand() const { return u_ & kSignificandMask; }
int Exponent() const { return static_cast<int>(((u_ & kExponentMask) >> kSignificandSize) - kExponentBias); }
bool IsNan() const { return (u_ & kExponentMask) == kExponentMask && Significand() != 0; }
bool IsInf() const { return (u_ & kExponentMask) == kExponentMask && Significand() == 0; }
bool IsNanOrInf() const { return (u_ & kExponentMask) == kExponentMask; }
bool IsNormal() const { return (u_ & kExponentMask) != 0 || Significand() == 0; }
bool IsZero() const { return (u_ & (kExponentMask | kSignificandMask)) == 0; }
uint64_t IntegerSignificand() const { return IsNormal() ? Significand() | kHiddenBit : Significand(); }
int IntegerExponent() const { return (IsNormal() ? Exponent() : kDenormalExponent) - kSignificandSize; }
uint64_t ToBias() const { return (u_ & kSignMask) ? ~u_ + 1 : u_ | kSignMask; }
static unsigned EffectiveSignificandSize(int order) {
if (order >= -1021)
return 53;
else if (order <= -1074)
return 0;
else
return static_cast<unsigned>(order) + 1074;
}
private:
static const int kSignificandSize = 52;
static const int kExponentBias = 0x3FF;
static const int kDenormalExponent = 1 - kExponentBias;
static const uint64_t kSignMask = RAPIDJSON_UINT64_C2(0x80000000, 0x00000000);
static const uint64_t kExponentMask = RAPIDJSON_UINT64_C2(0x7FF00000, 0x00000000);
static const uint64_t kSignificandMask = RAPIDJSON_UINT64_C2(0x000FFFFF, 0xFFFFFFFF);
static const uint64_t kHiddenBit = RAPIDJSON_UINT64_C2(0x00100000, 0x00000000);
union {
double d_;
uint64_t u_;
};
};
} // namespace internal
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_IEEE754_

304
rapidjson/internal/itoa.h Normal file
View File

@ -0,0 +1,304 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_ITOA_
#define RAPIDJSON_ITOA_
#include "../rapidjson.h"
RAPIDJSON_NAMESPACE_BEGIN
namespace internal {
inline const char* GetDigitsLut() {
static const char cDigitsLut[200] = {
'0','0','0','1','0','2','0','3','0','4','0','5','0','6','0','7','0','8','0','9',
'1','0','1','1','1','2','1','3','1','4','1','5','1','6','1','7','1','8','1','9',
'2','0','2','1','2','2','2','3','2','4','2','5','2','6','2','7','2','8','2','9',
'3','0','3','1','3','2','3','3','3','4','3','5','3','6','3','7','3','8','3','9',
'4','0','4','1','4','2','4','3','4','4','4','5','4','6','4','7','4','8','4','9',
'5','0','5','1','5','2','5','3','5','4','5','5','5','6','5','7','5','8','5','9',
'6','0','6','1','6','2','6','3','6','4','6','5','6','6','6','7','6','8','6','9',
'7','0','7','1','7','2','7','3','7','4','7','5','7','6','7','7','7','8','7','9',
'8','0','8','1','8','2','8','3','8','4','8','5','8','6','8','7','8','8','8','9',
'9','0','9','1','9','2','9','3','9','4','9','5','9','6','9','7','9','8','9','9'
};
return cDigitsLut;
}
inline char* u32toa(uint32_t value, char* buffer) {
const char* cDigitsLut = GetDigitsLut();
if (value < 10000) {
const uint32_t d1 = (value / 100) << 1;
const uint32_t d2 = (value % 100) << 1;
if (value >= 1000)
*buffer++ = cDigitsLut[d1];
if (value >= 100)
*buffer++ = cDigitsLut[d1 + 1];
if (value >= 10)
*buffer++ = cDigitsLut[d2];
*buffer++ = cDigitsLut[d2 + 1];
}
else if (value < 100000000) {
// value = bbbbcccc
const uint32_t b = value / 10000;
const uint32_t c = value % 10000;
const uint32_t d1 = (b / 100) << 1;
const uint32_t d2 = (b % 100) << 1;
const uint32_t d3 = (c / 100) << 1;
const uint32_t d4 = (c % 100) << 1;
if (value >= 10000000)
*buffer++ = cDigitsLut[d1];
if (value >= 1000000)
*buffer++ = cDigitsLut[d1 + 1];
if (value >= 100000)
*buffer++ = cDigitsLut[d2];
*buffer++ = cDigitsLut[d2 + 1];
*buffer++ = cDigitsLut[d3];
*buffer++ = cDigitsLut[d3 + 1];
*buffer++ = cDigitsLut[d4];
*buffer++ = cDigitsLut[d4 + 1];
}
else {
// value = aabbbbcccc in decimal
const uint32_t a = value / 100000000; // 1 to 42
value %= 100000000;
if (a >= 10) {
const unsigned i = a << 1;
*buffer++ = cDigitsLut[i];
*buffer++ = cDigitsLut[i + 1];
}
else
*buffer++ = static_cast<char>('0' + static_cast<char>(a));
const uint32_t b = value / 10000; // 0 to 9999
const uint32_t c = value % 10000; // 0 to 9999
const uint32_t d1 = (b / 100) << 1;
const uint32_t d2 = (b % 100) << 1;
const uint32_t d3 = (c / 100) << 1;
const uint32_t d4 = (c % 100) << 1;
*buffer++ = cDigitsLut[d1];
*buffer++ = cDigitsLut[d1 + 1];
*buffer++ = cDigitsLut[d2];
*buffer++ = cDigitsLut[d2 + 1];
*buffer++ = cDigitsLut[d3];
*buffer++ = cDigitsLut[d3 + 1];
*buffer++ = cDigitsLut[d4];
*buffer++ = cDigitsLut[d4 + 1];
}
return buffer;
}
inline char* i32toa(int32_t value, char* buffer) {
uint32_t u = static_cast<uint32_t>(value);
if (value < 0) {
*buffer++ = '-';
u = ~u + 1;
}
return u32toa(u, buffer);
}
inline char* u64toa(uint64_t value, char* buffer) {
const char* cDigitsLut = GetDigitsLut();
const uint64_t kTen8 = 100000000;
const uint64_t kTen9 = kTen8 * 10;
const uint64_t kTen10 = kTen8 * 100;
const uint64_t kTen11 = kTen8 * 1000;
const uint64_t kTen12 = kTen8 * 10000;
const uint64_t kTen13 = kTen8 * 100000;
const uint64_t kTen14 = kTen8 * 1000000;
const uint64_t kTen15 = kTen8 * 10000000;
const uint64_t kTen16 = kTen8 * kTen8;
if (value < kTen8) {
uint32_t v = static_cast<uint32_t>(value);
if (v < 10000) {
const uint32_t d1 = (v / 100) << 1;
const uint32_t d2 = (v % 100) << 1;
if (v >= 1000)
*buffer++ = cDigitsLut[d1];
if (v >= 100)
*buffer++ = cDigitsLut[d1 + 1];
if (v >= 10)
*buffer++ = cDigitsLut[d2];
*buffer++ = cDigitsLut[d2 + 1];
}
else {
// value = bbbbcccc
const uint32_t b = v / 10000;
const uint32_t c = v % 10000;
const uint32_t d1 = (b / 100) << 1;
const uint32_t d2 = (b % 100) << 1;
const uint32_t d3 = (c / 100) << 1;
const uint32_t d4 = (c % 100) << 1;
if (value >= 10000000)
*buffer++ = cDigitsLut[d1];
if (value >= 1000000)
*buffer++ = cDigitsLut[d1 + 1];
if (value >= 100000)
*buffer++ = cDigitsLut[d2];
*buffer++ = cDigitsLut[d2 + 1];
*buffer++ = cDigitsLut[d3];
*buffer++ = cDigitsLut[d3 + 1];
*buffer++ = cDigitsLut[d4];
*buffer++ = cDigitsLut[d4 + 1];
}
}
else if (value < kTen16) {
const uint32_t v0 = static_cast<uint32_t>(value / kTen8);
const uint32_t v1 = static_cast<uint32_t>(value % kTen8);
const uint32_t b0 = v0 / 10000;
const uint32_t c0 = v0 % 10000;
const uint32_t d1 = (b0 / 100) << 1;
const uint32_t d2 = (b0 % 100) << 1;
const uint32_t d3 = (c0 / 100) << 1;
const uint32_t d4 = (c0 % 100) << 1;
const uint32_t b1 = v1 / 10000;
const uint32_t c1 = v1 % 10000;
const uint32_t d5 = (b1 / 100) << 1;
const uint32_t d6 = (b1 % 100) << 1;
const uint32_t d7 = (c1 / 100) << 1;
const uint32_t d8 = (c1 % 100) << 1;
if (value >= kTen15)
*buffer++ = cDigitsLut[d1];
if (value >= kTen14)
*buffer++ = cDigitsLut[d1 + 1];
if (value >= kTen13)
*buffer++ = cDigitsLut[d2];
if (value >= kTen12)
*buffer++ = cDigitsLut[d2 + 1];
if (value >= kTen11)
*buffer++ = cDigitsLut[d3];
if (value >= kTen10)
*buffer++ = cDigitsLut[d3 + 1];
if (value >= kTen9)
*buffer++ = cDigitsLut[d4];
if (value >= kTen8)
*buffer++ = cDigitsLut[d4 + 1];
*buffer++ = cDigitsLut[d5];
*buffer++ = cDigitsLut[d5 + 1];
*buffer++ = cDigitsLut[d6];
*buffer++ = cDigitsLut[d6 + 1];
*buffer++ = cDigitsLut[d7];
*buffer++ = cDigitsLut[d7 + 1];
*buffer++ = cDigitsLut[d8];
*buffer++ = cDigitsLut[d8 + 1];
}
else {
const uint32_t a = static_cast<uint32_t>(value / kTen16); // 1 to 1844
value %= kTen16;
if (a < 10)
*buffer++ = static_cast<char>('0' + static_cast<char>(a));
else if (a < 100) {
const uint32_t i = a << 1;
*buffer++ = cDigitsLut[i];
*buffer++ = cDigitsLut[i + 1];
}
else if (a < 1000) {
*buffer++ = static_cast<char>('0' + static_cast<char>(a / 100));
const uint32_t i = (a % 100) << 1;
*buffer++ = cDigitsLut[i];
*buffer++ = cDigitsLut[i + 1];
}
else {
const uint32_t i = (a / 100) << 1;
const uint32_t j = (a % 100) << 1;
*buffer++ = cDigitsLut[i];
*buffer++ = cDigitsLut[i + 1];
*buffer++ = cDigitsLut[j];
*buffer++ = cDigitsLut[j + 1];
}
const uint32_t v0 = static_cast<uint32_t>(value / kTen8);
const uint32_t v1 = static_cast<uint32_t>(value % kTen8);
const uint32_t b0 = v0 / 10000;
const uint32_t c0 = v0 % 10000;
const uint32_t d1 = (b0 / 100) << 1;
const uint32_t d2 = (b0 % 100) << 1;
const uint32_t d3 = (c0 / 100) << 1;
const uint32_t d4 = (c0 % 100) << 1;
const uint32_t b1 = v1 / 10000;
const uint32_t c1 = v1 % 10000;
const uint32_t d5 = (b1 / 100) << 1;
const uint32_t d6 = (b1 % 100) << 1;
const uint32_t d7 = (c1 / 100) << 1;
const uint32_t d8 = (c1 % 100) << 1;
*buffer++ = cDigitsLut[d1];
*buffer++ = cDigitsLut[d1 + 1];
*buffer++ = cDigitsLut[d2];
*buffer++ = cDigitsLut[d2 + 1];
*buffer++ = cDigitsLut[d3];
*buffer++ = cDigitsLut[d3 + 1];
*buffer++ = cDigitsLut[d4];
*buffer++ = cDigitsLut[d4 + 1];
*buffer++ = cDigitsLut[d5];
*buffer++ = cDigitsLut[d5 + 1];
*buffer++ = cDigitsLut[d6];
*buffer++ = cDigitsLut[d6 + 1];
*buffer++ = cDigitsLut[d7];
*buffer++ = cDigitsLut[d7 + 1];
*buffer++ = cDigitsLut[d8];
*buffer++ = cDigitsLut[d8 + 1];
}
return buffer;
}
inline char* i64toa(int64_t value, char* buffer) {
uint64_t u = static_cast<uint64_t>(value);
if (value < 0) {
*buffer++ = '-';
u = ~u + 1;
}
return u64toa(u, buffer);
}
} // namespace internal
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_ITOA_

181
rapidjson/internal/meta.h Normal file
View File

@ -0,0 +1,181 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_INTERNAL_META_H_
#define RAPIDJSON_INTERNAL_META_H_
#include "../rapidjson.h"
#ifdef __GNUC__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(effc++)
#endif
#if defined(_MSC_VER)
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(6334)
#endif
#if RAPIDJSON_HAS_CXX11_TYPETRAITS
#include <type_traits>
#endif
//@cond RAPIDJSON_INTERNAL
RAPIDJSON_NAMESPACE_BEGIN
namespace internal {
// Helper to wrap/convert arbitrary types to void, useful for arbitrary type matching
template <typename T> struct Void { typedef void Type; };
///////////////////////////////////////////////////////////////////////////////
// BoolType, TrueType, FalseType
//
template <bool Cond> struct BoolType {
static const bool Value = Cond;
typedef BoolType Type;
};
typedef BoolType<true> TrueType;
typedef BoolType<false> FalseType;
///////////////////////////////////////////////////////////////////////////////
// SelectIf, BoolExpr, NotExpr, AndExpr, OrExpr
//
template <bool C> struct SelectIfImpl { template <typename T1, typename T2> struct Apply { typedef T1 Type; }; };
template <> struct SelectIfImpl<false> { template <typename T1, typename T2> struct Apply { typedef T2 Type; }; };
template <bool C, typename T1, typename T2> struct SelectIfCond : SelectIfImpl<C>::template Apply<T1,T2> {};
template <typename C, typename T1, typename T2> struct SelectIf : SelectIfCond<C::Value, T1, T2> {};
template <bool Cond1, bool Cond2> struct AndExprCond : FalseType {};
template <> struct AndExprCond<true, true> : TrueType {};
template <bool Cond1, bool Cond2> struct OrExprCond : TrueType {};
template <> struct OrExprCond<false, false> : FalseType {};
template <typename C> struct BoolExpr : SelectIf<C,TrueType,FalseType>::Type {};
template <typename C> struct NotExpr : SelectIf<C,FalseType,TrueType>::Type {};
template <typename C1, typename C2> struct AndExpr : AndExprCond<C1::Value, C2::Value>::Type {};
template <typename C1, typename C2> struct OrExpr : OrExprCond<C1::Value, C2::Value>::Type {};
///////////////////////////////////////////////////////////////////////////////
// AddConst, MaybeAddConst, RemoveConst
template <typename T> struct AddConst { typedef const T Type; };
template <bool Constify, typename T> struct MaybeAddConst : SelectIfCond<Constify, const T, T> {};
template <typename T> struct RemoveConst { typedef T Type; };
template <typename T> struct RemoveConst<const T> { typedef T Type; };
///////////////////////////////////////////////////////////////////////////////
// IsSame, IsConst, IsMoreConst, IsPointer
//
template <typename T, typename U> struct IsSame : FalseType {};
template <typename T> struct IsSame<T, T> : TrueType {};
template <typename T> struct IsConst : FalseType {};
template <typename T> struct IsConst<const T> : TrueType {};
template <typename CT, typename T>
struct IsMoreConst
: AndExpr<IsSame<typename RemoveConst<CT>::Type, typename RemoveConst<T>::Type>,
BoolType<IsConst<CT>::Value >= IsConst<T>::Value> >::Type {};
template <typename T> struct IsPointer : FalseType {};
template <typename T> struct IsPointer<T*> : TrueType {};
///////////////////////////////////////////////////////////////////////////////
// IsBaseOf
//
#if RAPIDJSON_HAS_CXX11_TYPETRAITS
template <typename B, typename D> struct IsBaseOf
: BoolType< ::std::is_base_of<B,D>::value> {};
#else // simplified version adopted from Boost
template<typename B, typename D> struct IsBaseOfImpl {
RAPIDJSON_STATIC_ASSERT(sizeof(B) != 0);
RAPIDJSON_STATIC_ASSERT(sizeof(D) != 0);
typedef char (&Yes)[1];
typedef char (&No) [2];
template <typename T>
static Yes Check(const D*, T);
static No Check(const B*, int);
struct Host {
operator const B*() const;
operator const D*();
};
enum { Value = (sizeof(Check(Host(), 0)) == sizeof(Yes)) };
};
template <typename B, typename D> struct IsBaseOf
: OrExpr<IsSame<B, D>, BoolExpr<IsBaseOfImpl<B, D> > >::Type {};
#endif // RAPIDJSON_HAS_CXX11_TYPETRAITS
//////////////////////////////////////////////////////////////////////////
// EnableIf / DisableIf
//
template <bool Condition, typename T = void> struct EnableIfCond { typedef T Type; };
template <typename T> struct EnableIfCond<false, T> { /* empty */ };
template <bool Condition, typename T = void> struct DisableIfCond { typedef T Type; };
template <typename T> struct DisableIfCond<true, T> { /* empty */ };
template <typename Condition, typename T = void>
struct EnableIf : EnableIfCond<Condition::Value, T> {};
template <typename Condition, typename T = void>
struct DisableIf : DisableIfCond<Condition::Value, T> {};
// SFINAE helpers
struct SfinaeTag {};
template <typename T> struct RemoveSfinaeTag;
template <typename T> struct RemoveSfinaeTag<SfinaeTag&(*)(T)> { typedef T Type; };
#define RAPIDJSON_REMOVEFPTR_(type) \
typename ::RAPIDJSON_NAMESPACE::internal::RemoveSfinaeTag \
< ::RAPIDJSON_NAMESPACE::internal::SfinaeTag&(*) type>::Type
#define RAPIDJSON_ENABLEIF(cond) \
typename ::RAPIDJSON_NAMESPACE::internal::EnableIf \
<RAPIDJSON_REMOVEFPTR_(cond)>::Type * = NULL
#define RAPIDJSON_DISABLEIF(cond) \
typename ::RAPIDJSON_NAMESPACE::internal::DisableIf \
<RAPIDJSON_REMOVEFPTR_(cond)>::Type * = NULL
#define RAPIDJSON_ENABLEIF_RETURN(cond,returntype) \
typename ::RAPIDJSON_NAMESPACE::internal::EnableIf \
<RAPIDJSON_REMOVEFPTR_(cond), \
RAPIDJSON_REMOVEFPTR_(returntype)>::Type
#define RAPIDJSON_DISABLEIF_RETURN(cond,returntype) \
typename ::RAPIDJSON_NAMESPACE::internal::DisableIf \
<RAPIDJSON_REMOVEFPTR_(cond), \
RAPIDJSON_REMOVEFPTR_(returntype)>::Type
} // namespace internal
RAPIDJSON_NAMESPACE_END
//@endcond
#if defined(__GNUC__) || defined(_MSC_VER)
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_INTERNAL_META_H_

55
rapidjson/internal/pow10.h Normal file
View File

@ -0,0 +1,55 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_POW10_
#define RAPIDJSON_POW10_
#include "../rapidjson.h"
RAPIDJSON_NAMESPACE_BEGIN
namespace internal {
//! Computes integer powers of 10 in double (10.0^n).
/*! This function uses lookup table for fast and accurate results.
\param n non-negative exponent. Must <= 308.
\return 10.0^n
*/
inline double Pow10(int n) {
static const double e[] = { // 1e-0...1e308: 309 * 8 bytes = 2472 bytes
1e+0,
1e+1, 1e+2, 1e+3, 1e+4, 1e+5, 1e+6, 1e+7, 1e+8, 1e+9, 1e+10, 1e+11, 1e+12, 1e+13, 1e+14, 1e+15, 1e+16, 1e+17, 1e+18, 1e+19, 1e+20,
1e+21, 1e+22, 1e+23, 1e+24, 1e+25, 1e+26, 1e+27, 1e+28, 1e+29, 1e+30, 1e+31, 1e+32, 1e+33, 1e+34, 1e+35, 1e+36, 1e+37, 1e+38, 1e+39, 1e+40,
1e+41, 1e+42, 1e+43, 1e+44, 1e+45, 1e+46, 1e+47, 1e+48, 1e+49, 1e+50, 1e+51, 1e+52, 1e+53, 1e+54, 1e+55, 1e+56, 1e+57, 1e+58, 1e+59, 1e+60,
1e+61, 1e+62, 1e+63, 1e+64, 1e+65, 1e+66, 1e+67, 1e+68, 1e+69, 1e+70, 1e+71, 1e+72, 1e+73, 1e+74, 1e+75, 1e+76, 1e+77, 1e+78, 1e+79, 1e+80,
1e+81, 1e+82, 1e+83, 1e+84, 1e+85, 1e+86, 1e+87, 1e+88, 1e+89, 1e+90, 1e+91, 1e+92, 1e+93, 1e+94, 1e+95, 1e+96, 1e+97, 1e+98, 1e+99, 1e+100,
1e+101,1e+102,1e+103,1e+104,1e+105,1e+106,1e+107,1e+108,1e+109,1e+110,1e+111,1e+112,1e+113,1e+114,1e+115,1e+116,1e+117,1e+118,1e+119,1e+120,
1e+121,1e+122,1e+123,1e+124,1e+125,1e+126,1e+127,1e+128,1e+129,1e+130,1e+131,1e+132,1e+133,1e+134,1e+135,1e+136,1e+137,1e+138,1e+139,1e+140,
1e+141,1e+142,1e+143,1e+144,1e+145,1e+146,1e+147,1e+148,1e+149,1e+150,1e+151,1e+152,1e+153,1e+154,1e+155,1e+156,1e+157,1e+158,1e+159,1e+160,
1e+161,1e+162,1e+163,1e+164,1e+165,1e+166,1e+167,1e+168,1e+169,1e+170,1e+171,1e+172,1e+173,1e+174,1e+175,1e+176,1e+177,1e+178,1e+179,1e+180,
1e+181,1e+182,1e+183,1e+184,1e+185,1e+186,1e+187,1e+188,1e+189,1e+190,1e+191,1e+192,1e+193,1e+194,1e+195,1e+196,1e+197,1e+198,1e+199,1e+200,
1e+201,1e+202,1e+203,1e+204,1e+205,1e+206,1e+207,1e+208,1e+209,1e+210,1e+211,1e+212,1e+213,1e+214,1e+215,1e+216,1e+217,1e+218,1e+219,1e+220,
1e+221,1e+222,1e+223,1e+224,1e+225,1e+226,1e+227,1e+228,1e+229,1e+230,1e+231,1e+232,1e+233,1e+234,1e+235,1e+236,1e+237,1e+238,1e+239,1e+240,
1e+241,1e+242,1e+243,1e+244,1e+245,1e+246,1e+247,1e+248,1e+249,1e+250,1e+251,1e+252,1e+253,1e+254,1e+255,1e+256,1e+257,1e+258,1e+259,1e+260,
1e+261,1e+262,1e+263,1e+264,1e+265,1e+266,1e+267,1e+268,1e+269,1e+270,1e+271,1e+272,1e+273,1e+274,1e+275,1e+276,1e+277,1e+278,1e+279,1e+280,
1e+281,1e+282,1e+283,1e+284,1e+285,1e+286,1e+287,1e+288,1e+289,1e+290,1e+291,1e+292,1e+293,1e+294,1e+295,1e+296,1e+297,1e+298,1e+299,1e+300,
1e+301,1e+302,1e+303,1e+304,1e+305,1e+306,1e+307,1e+308
};
RAPIDJSON_ASSERT(n >= 0 && n <= 308);
return e[n];
}
} // namespace internal
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_POW10_

731
rapidjson/internal/regex.h Normal file
View File

@ -0,0 +1,731 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_INTERNAL_REGEX_H_
#define RAPIDJSON_INTERNAL_REGEX_H_
#include "../allocators.h"
#include "../stream.h"
#include "stack.h"
#ifdef __clang__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(padded)
RAPIDJSON_DIAG_OFF(switch-enum)
RAPIDJSON_DIAG_OFF(implicit-fallthrough)
#endif
#ifdef __GNUC__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(effc++)
#endif
#ifdef _MSC_VER
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(4512) // assignment operator could not be generated
#endif
#ifndef RAPIDJSON_REGEX_VERBOSE
#define RAPIDJSON_REGEX_VERBOSE 0
#endif
RAPIDJSON_NAMESPACE_BEGIN
namespace internal {
///////////////////////////////////////////////////////////////////////////////
// DecodedStream
template <typename SourceStream, typename Encoding>
class DecodedStream {
public:
DecodedStream(SourceStream& ss) : ss_(ss), codepoint_() { Decode(); }
unsigned Peek() { return codepoint_; }
unsigned Take() {
unsigned c = codepoint_;
if (c) // No further decoding when '\0'
Decode();
return c;
}
private:
void Decode() {
if (!Encoding::Decode(ss_, &codepoint_))
codepoint_ = 0;
}
SourceStream& ss_;
unsigned codepoint_;
};
///////////////////////////////////////////////////////////////////////////////
// GenericRegex
static const SizeType kRegexInvalidState = ~SizeType(0); //!< Represents an invalid index in GenericRegex::State::out, out1
static const SizeType kRegexInvalidRange = ~SizeType(0);
template <typename Encoding, typename Allocator>
class GenericRegexSearch;
//! Regular expression engine with subset of ECMAscript grammar.
/*!
Supported regular expression syntax:
- \c ab Concatenation
- \c a|b Alternation
- \c a? Zero or one
- \c a* Zero or more
- \c a+ One or more
- \c a{3} Exactly 3 times
- \c a{3,} At least 3 times
- \c a{3,5} 3 to 5 times
- \c (ab) Grouping
- \c ^a At the beginning
- \c a$ At the end
- \c . Any character
- \c [abc] Character classes
- \c [a-c] Character class range
- \c [a-z0-9_] Character class combination
- \c [^abc] Negated character classes
- \c [^a-c] Negated character class range
- \c [\b] Backspace (U+0008)
- \c \\| \\\\ ... Escape characters
- \c \\f Form feed (U+000C)
- \c \\n Line feed (U+000A)
- \c \\r Carriage return (U+000D)
- \c \\t Tab (U+0009)
- \c \\v Vertical tab (U+000B)
\note This is a Thompson NFA engine, implemented with reference to
Cox, Russ. "Regular Expression Matching Can Be Simple And Fast (but is slow in Java, Perl, PHP, Python, Ruby,...).",
https://swtch.com/~rsc/regexp/regexp1.html
*/
template <typename Encoding, typename Allocator = CrtAllocator>
class GenericRegex {
public:
typedef Encoding EncodingType;
typedef typename Encoding::Ch Ch;
template <typename, typename> friend class GenericRegexSearch;
GenericRegex(const Ch* source, Allocator* allocator = 0) :
states_(allocator, 256), ranges_(allocator, 256), root_(kRegexInvalidState), stateCount_(), rangeCount_(),
anchorBegin_(), anchorEnd_()
{
GenericStringStream<Encoding> ss(source);
DecodedStream<GenericStringStream<Encoding>, Encoding> ds(ss);
Parse(ds);
}
~GenericRegex() {}
bool IsValid() const {
return root_ != kRegexInvalidState;
}
private:
enum Operator {
kZeroOrOne,
kZeroOrMore,
kOneOrMore,
kConcatenation,
kAlternation,
kLeftParenthesis
};
static const unsigned kAnyCharacterClass = 0xFFFFFFFF; //!< For '.'
static const unsigned kRangeCharacterClass = 0xFFFFFFFE;
static const unsigned kRangeNegationFlag = 0x80000000;
struct Range {
unsigned start; //
unsigned end;
SizeType next;
};
struct State {
SizeType out; //!< Equals to kInvalid for matching state
SizeType out1; //!< Equals to non-kInvalid for split
SizeType rangeStart;
unsigned codepoint;
};
struct Frag {
Frag(SizeType s, SizeType o, SizeType m) : start(s), out(o), minIndex(m) {}
SizeType start;
SizeType out; //!< link-list of all output states
SizeType minIndex;
};
State& GetState(SizeType index) {
RAPIDJSON_ASSERT(index < stateCount_);
return states_.template Bottom<State>()[index];
}
const State& GetState(SizeType index) const {
RAPIDJSON_ASSERT(index < stateCount_);
return states_.template Bottom<State>()[index];
}
Range& GetRange(SizeType index) {
RAPIDJSON_ASSERT(index < rangeCount_);
return ranges_.template Bottom<Range>()[index];
}
const Range& GetRange(SizeType index) const {
RAPIDJSON_ASSERT(index < rangeCount_);
return ranges_.template Bottom<Range>()[index];
}
template <typename InputStream>
void Parse(DecodedStream<InputStream, Encoding>& ds) {
Allocator allocator;
Stack<Allocator> operandStack(&allocator, 256); // Frag
Stack<Allocator> operatorStack(&allocator, 256); // Operator
Stack<Allocator> atomCountStack(&allocator, 256); // unsigned (Atom per parenthesis)
*atomCountStack.template Push<unsigned>() = 0;
unsigned codepoint;
while (ds.Peek() != 0) {
switch (codepoint = ds.Take()) {
case '^':
anchorBegin_ = true;
break;
case '$':
anchorEnd_ = true;
break;
case '|':
while (!operatorStack.Empty() && *operatorStack.template Top<Operator>() < kAlternation)
if (!Eval(operandStack, *operatorStack.template Pop<Operator>(1)))
return;
*operatorStack.template Push<Operator>() = kAlternation;
*atomCountStack.template Top<unsigned>() = 0;
break;
case '(':
*operatorStack.template Push<Operator>() = kLeftParenthesis;
*atomCountStack.template Push<unsigned>() = 0;
break;
case ')':
while (!operatorStack.Empty() && *operatorStack.template Top<Operator>() != kLeftParenthesis)
if (!Eval(operandStack, *operatorStack.template Pop<Operator>(1)))
return;
if (operatorStack.Empty())
return;
operatorStack.template Pop<Operator>(1);
atomCountStack.template Pop<unsigned>(1);
ImplicitConcatenation(atomCountStack, operatorStack);
break;
case '?':
if (!Eval(operandStack, kZeroOrOne))
return;
break;
case '*':
if (!Eval(operandStack, kZeroOrMore))
return;
break;
case '+':
if (!Eval(operandStack, kOneOrMore))
return;
break;
case '{':
{
unsigned n, m;
if (!ParseUnsigned(ds, &n))
return;
if (ds.Peek() == ',') {
ds.Take();
if (ds.Peek() == '}')
m = kInfinityQuantifier;
else if (!ParseUnsigned(ds, &m) || m < n)
return;
}
else
m = n;
if (!EvalQuantifier(operandStack, n, m) || ds.Peek() != '}')
return;
ds.Take();
}
break;
case '.':
PushOperand(operandStack, kAnyCharacterClass);
ImplicitConcatenation(atomCountStack, operatorStack);
break;
case '[':
{
SizeType range;
if (!ParseRange(ds, &range))
return;
SizeType s = NewState(kRegexInvalidState, kRegexInvalidState, kRangeCharacterClass);
GetState(s).rangeStart = range;
*operandStack.template Push<Frag>() = Frag(s, s, s);
}
ImplicitConcatenation(atomCountStack, operatorStack);
break;
case '\\': // Escape character
if (!CharacterEscape(ds, &codepoint))
return; // Unsupported escape character
// fall through to default
default: // Pattern character
PushOperand(operandStack, codepoint);
ImplicitConcatenation(atomCountStack, operatorStack);
}
}
while (!operatorStack.Empty())
if (!Eval(operandStack, *operatorStack.template Pop<Operator>(1)))
return;
// Link the operand to matching state.
if (operandStack.GetSize() == sizeof(Frag)) {
Frag* e = operandStack.template Pop<Frag>(1);
Patch(e->out, NewState(kRegexInvalidState, kRegexInvalidState, 0));
root_ = e->start;
#if RAPIDJSON_REGEX_VERBOSE
printf("root: %d\n", root_);
for (SizeType i = 0; i < stateCount_ ; i++) {
State& s = GetState(i);
printf("[%2d] out: %2d out1: %2d c: '%c'\n", i, s.out, s.out1, (char)s.codepoint);
}
printf("\n");
#endif
}
}
SizeType NewState(SizeType out, SizeType out1, unsigned codepoint) {
State* s = states_.template Push<State>();
s->out = out;
s->out1 = out1;
s->codepoint = codepoint;
s->rangeStart = kRegexInvalidRange;
return stateCount_++;
}
void PushOperand(Stack<Allocator>& operandStack, unsigned codepoint) {
SizeType s = NewState(kRegexInvalidState, kRegexInvalidState, codepoint);
*operandStack.template Push<Frag>() = Frag(s, s, s);
}
void ImplicitConcatenation(Stack<Allocator>& atomCountStack, Stack<Allocator>& operatorStack) {
if (*atomCountStack.template Top<unsigned>())
*operatorStack.template Push<Operator>() = kConcatenation;
(*atomCountStack.template Top<unsigned>())++;
}
SizeType Append(SizeType l1, SizeType l2) {
SizeType old = l1;
while (GetState(l1).out != kRegexInvalidState)
l1 = GetState(l1).out;
GetState(l1).out = l2;
return old;
}
void Patch(SizeType l, SizeType s) {
for (SizeType next; l != kRegexInvalidState; l = next) {
next = GetState(l).out;
GetState(l).out = s;
}
}
bool Eval(Stack<Allocator>& operandStack, Operator op) {
switch (op) {
case kConcatenation:
RAPIDJSON_ASSERT(operandStack.GetSize() >= sizeof(Frag) * 2);
{
Frag e2 = *operandStack.template Pop<Frag>(1);
Frag e1 = *operandStack.template Pop<Frag>(1);
Patch(e1.out, e2.start);
*operandStack.template Push<Frag>() = Frag(e1.start, e2.out, Min(e1.minIndex, e2.minIndex));
}
return true;
case kAlternation:
if (operandStack.GetSize() >= sizeof(Frag) * 2) {
Frag e2 = *operandStack.template Pop<Frag>(1);
Frag e1 = *operandStack.template Pop<Frag>(1);
SizeType s = NewState(e1.start, e2.start, 0);
*operandStack.template Push<Frag>() = Frag(s, Append(e1.out, e2.out), Min(e1.minIndex, e2.minIndex));
return true;
}
return false;
case kZeroOrOne:
if (operandStack.GetSize() >= sizeof(Frag)) {
Frag e = *operandStack.template Pop<Frag>(1);
SizeType s = NewState(kRegexInvalidState, e.start, 0);
*operandStack.template Push<Frag>() = Frag(s, Append(e.out, s), e.minIndex);
return true;
}
return false;
case kZeroOrMore:
if (operandStack.GetSize() >= sizeof(Frag)) {
Frag e = *operandStack.template Pop<Frag>(1);
SizeType s = NewState(kRegexInvalidState, e.start, 0);
Patch(e.out, s);
*operandStack.template Push<Frag>() = Frag(s, s, e.minIndex);
return true;
}
return false;
default:
RAPIDJSON_ASSERT(op == kOneOrMore);
if (operandStack.GetSize() >= sizeof(Frag)) {
Frag e = *operandStack.template Pop<Frag>(1);
SizeType s = NewState(kRegexInvalidState, e.start, 0);
Patch(e.out, s);
*operandStack.template Push<Frag>() = Frag(e.start, s, e.minIndex);
return true;
}
return false;
}
}
bool EvalQuantifier(Stack<Allocator>& operandStack, unsigned n, unsigned m) {
RAPIDJSON_ASSERT(n <= m);
RAPIDJSON_ASSERT(operandStack.GetSize() >= sizeof(Frag));
if (n == 0) {
if (m == 0) // a{0} not support
return false;
else if (m == kInfinityQuantifier)
Eval(operandStack, kZeroOrMore); // a{0,} -> a*
else {
Eval(operandStack, kZeroOrOne); // a{0,5} -> a?
for (unsigned i = 0; i < m - 1; i++)
CloneTopOperand(operandStack); // a{0,5} -> a? a? a? a? a?
for (unsigned i = 0; i < m - 1; i++)
Eval(operandStack, kConcatenation); // a{0,5} -> a?a?a?a?a?
}
return true;
}
for (unsigned i = 0; i < n - 1; i++) // a{3} -> a a a
CloneTopOperand(operandStack);
if (m == kInfinityQuantifier)
Eval(operandStack, kOneOrMore); // a{3,} -> a a a+
else if (m > n) {
CloneTopOperand(operandStack); // a{3,5} -> a a a a
Eval(operandStack, kZeroOrOne); // a{3,5} -> a a a a?
for (unsigned i = n; i < m - 1; i++)
CloneTopOperand(operandStack); // a{3,5} -> a a a a? a?
for (unsigned i = n; i < m; i++)
Eval(operandStack, kConcatenation); // a{3,5} -> a a aa?a?
}
for (unsigned i = 0; i < n - 1; i++)
Eval(operandStack, kConcatenation); // a{3} -> aaa, a{3,} -> aaa+, a{3.5} -> aaaa?a?
return true;
}
static SizeType Min(SizeType a, SizeType b) { return a < b ? a : b; }
void CloneTopOperand(Stack<Allocator>& operandStack) {
const Frag src = *operandStack.template Top<Frag>(); // Copy constructor to prevent invalidation
SizeType count = stateCount_ - src.minIndex; // Assumes top operand contains states in [src->minIndex, stateCount_)
State* s = states_.template Push<State>(count);
memcpy(s, &GetState(src.minIndex), count * sizeof(State));
for (SizeType j = 0; j < count; j++) {
if (s[j].out != kRegexInvalidState)
s[j].out += count;
if (s[j].out1 != kRegexInvalidState)
s[j].out1 += count;
}
*operandStack.template Push<Frag>() = Frag(src.start + count, src.out + count, src.minIndex + count);
stateCount_ += count;
}
template <typename InputStream>
bool ParseUnsigned(DecodedStream<InputStream, Encoding>& ds, unsigned* u) {
unsigned r = 0;
if (ds.Peek() < '0' || ds.Peek() > '9')
return false;
while (ds.Peek() >= '0' && ds.Peek() <= '9') {
if (r >= 429496729 && ds.Peek() > '5') // 2^32 - 1 = 4294967295
return false; // overflow
r = r * 10 + (ds.Take() - '0');
}
*u = r;
return true;
}
template <typename InputStream>
bool ParseRange(DecodedStream<InputStream, Encoding>& ds, SizeType* range) {
bool isBegin = true;
bool negate = false;
int step = 0;
SizeType start = kRegexInvalidRange;
SizeType current = kRegexInvalidRange;
unsigned codepoint;
while ((codepoint = ds.Take()) != 0) {
if (isBegin) {
isBegin = false;
if (codepoint == '^') {
negate = true;
continue;
}
}
switch (codepoint) {
case ']':
if (start == kRegexInvalidRange)
return false; // Error: nothing inside []
if (step == 2) { // Add trailing '-'
SizeType r = NewRange('-');
RAPIDJSON_ASSERT(current != kRegexInvalidRange);
GetRange(current).next = r;
}
if (negate)
GetRange(start).start |= kRangeNegationFlag;
*range = start;
return true;
case '\\':
if (ds.Peek() == 'b') {
ds.Take();
codepoint = 0x0008; // Escape backspace character
}
else if (!CharacterEscape(ds, &codepoint))
return false;
// fall through to default
default:
switch (step) {
case 1:
if (codepoint == '-') {
step++;
break;
}
// fall through to step 0 for other characters
case 0:
{
SizeType r = NewRange(codepoint);
if (current != kRegexInvalidRange)
GetRange(current).next = r;
if (start == kRegexInvalidRange)
start = r;
current = r;
}
step = 1;
break;
default:
RAPIDJSON_ASSERT(step == 2);
GetRange(current).end = codepoint;
step = 0;
}
}
}
return false;
}
SizeType NewRange(unsigned codepoint) {
Range* r = ranges_.template Push<Range>();
r->start = r->end = codepoint;
r->next = kRegexInvalidRange;
return rangeCount_++;
}
template <typename InputStream>
bool CharacterEscape(DecodedStream<InputStream, Encoding>& ds, unsigned* escapedCodepoint) {
unsigned codepoint;
switch (codepoint = ds.Take()) {
case '^':
case '$':
case '|':
case '(':
case ')':
case '?':
case '*':
case '+':
case '.':
case '[':
case ']':
case '{':
case '}':
case '\\':
*escapedCodepoint = codepoint; return true;
case 'f': *escapedCodepoint = 0x000C; return true;
case 'n': *escapedCodepoint = 0x000A; return true;
case 'r': *escapedCodepoint = 0x000D; return true;
case 't': *escapedCodepoint = 0x0009; return true;
case 'v': *escapedCodepoint = 0x000B; return true;
default:
return false; // Unsupported escape character
}
}
Stack<Allocator> states_;
Stack<Allocator> ranges_;
SizeType root_;
SizeType stateCount_;
SizeType rangeCount_;
static const unsigned kInfinityQuantifier = ~0u;
// For SearchWithAnchoring()
bool anchorBegin_;
bool anchorEnd_;
};
template <typename RegexType, typename Allocator = CrtAllocator>
class GenericRegexSearch {
public:
typedef typename RegexType::EncodingType Encoding;
typedef typename Encoding::Ch Ch;
GenericRegexSearch(const RegexType& regex, Allocator* allocator = 0) :
regex_(regex), allocator_(allocator), ownAllocator_(0),
state0_(allocator, 0), state1_(allocator, 0), stateSet_()
{
RAPIDJSON_ASSERT(regex_.IsValid());
if (!allocator_)
ownAllocator_ = allocator_ = RAPIDJSON_NEW(Allocator());
stateSet_ = static_cast<unsigned*>(allocator_->Malloc(GetStateSetSize()));
state0_.template Reserve<SizeType>(regex_.stateCount_);
state1_.template Reserve<SizeType>(regex_.stateCount_);
}
~GenericRegexSearch() {
Allocator::Free(stateSet_);
RAPIDJSON_DELETE(ownAllocator_);
}
template <typename InputStream>
bool Match(InputStream& is) {
return SearchWithAnchoring(is, true, true);
}
bool Match(const Ch* s) {
GenericStringStream<Encoding> is(s);
return Match(is);
}
template <typename InputStream>
bool Search(InputStream& is) {
return SearchWithAnchoring(is, regex_.anchorBegin_, regex_.anchorEnd_);
}
bool Search(const Ch* s) {
GenericStringStream<Encoding> is(s);
return Search(is);
}
private:
typedef typename RegexType::State State;
typedef typename RegexType::Range Range;
template <typename InputStream>
bool SearchWithAnchoring(InputStream& is, bool anchorBegin, bool anchorEnd) {
DecodedStream<InputStream, Encoding> ds(is);
state0_.Clear();
Stack<Allocator> *current = &state0_, *next = &state1_;
const size_t stateSetSize = GetStateSetSize();
std::memset(stateSet_, 0, stateSetSize);
bool matched = AddState(*current, regex_.root_);
unsigned codepoint;
while (!current->Empty() && (codepoint = ds.Take()) != 0) {
std::memset(stateSet_, 0, stateSetSize);
next->Clear();
matched = false;
for (const SizeType* s = current->template Bottom<SizeType>(); s != current->template End<SizeType>(); ++s) {
const State& sr = regex_.GetState(*s);
if (sr.codepoint == codepoint ||
sr.codepoint == RegexType::kAnyCharacterClass ||
(sr.codepoint == RegexType::kRangeCharacterClass && MatchRange(sr.rangeStart, codepoint)))
{
matched = AddState(*next, sr.out) || matched;
if (!anchorEnd && matched)
return true;
}
if (!anchorBegin)
AddState(*next, regex_.root_);
}
internal::Swap(current, next);
}
return matched;
}
size_t GetStateSetSize() const {
return (regex_.stateCount_ + 31) / 32 * 4;
}
// Return whether the added states is a match state
bool AddState(Stack<Allocator>& l, SizeType index) {
RAPIDJSON_ASSERT(index != kRegexInvalidState);
const State& s = regex_.GetState(index);
if (s.out1 != kRegexInvalidState) { // Split
bool matched = AddState(l, s.out);
return AddState(l, s.out1) || matched;
}
else if (!(stateSet_[index >> 5] & (1 << (index & 31)))) {
stateSet_[index >> 5] |= (1 << (index & 31));
*l.template PushUnsafe<SizeType>() = index;
}
return s.out == kRegexInvalidState; // by using PushUnsafe() above, we can ensure s is not validated due to reallocation.
}
bool MatchRange(SizeType rangeIndex, unsigned codepoint) const {
bool yes = (regex_.GetRange(rangeIndex).start & RegexType::kRangeNegationFlag) == 0;
while (rangeIndex != kRegexInvalidRange) {
const Range& r = regex_.GetRange(rangeIndex);
if (codepoint >= (r.start & ~RegexType::kRangeNegationFlag) && codepoint <= r.end)
return yes;
rangeIndex = r.next;
}
return !yes;
}
const RegexType& regex_;
Allocator* allocator_;
Allocator* ownAllocator_;
Stack<Allocator> state0_;
Stack<Allocator> state1_;
uint32_t* stateSet_;
};
typedef GenericRegex<UTF8<> > Regex;
typedef GenericRegexSearch<Regex> RegexSearch;
} // namespace internal
RAPIDJSON_NAMESPACE_END
#ifdef __clang__
RAPIDJSON_DIAG_POP
#endif
#ifdef _MSC_VER
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_INTERNAL_REGEX_H_

230
rapidjson/internal/stack.h Normal file
View File

@ -0,0 +1,230 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_INTERNAL_STACK_H_
#define RAPIDJSON_INTERNAL_STACK_H_
#include "../allocators.h"
#include "swap.h"
#if defined(__clang__)
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(c++98-compat)
#endif
RAPIDJSON_NAMESPACE_BEGIN
namespace internal {
///////////////////////////////////////////////////////////////////////////////
// Stack
//! A type-unsafe stack for storing different types of data.
/*! \tparam Allocator Allocator for allocating stack memory.
*/
template <typename Allocator>
class Stack {
public:
// Optimization note: Do not allocate memory for stack_ in constructor.
// Do it lazily when first Push() -> Expand() -> Resize().
Stack(Allocator* allocator, size_t stackCapacity) : allocator_(allocator), ownAllocator_(0), stack_(0), stackTop_(0), stackEnd_(0), initialCapacity_(stackCapacity) {
}
#if RAPIDJSON_HAS_CXX11_RVALUE_REFS
Stack(Stack&& rhs)
: allocator_(rhs.allocator_),
ownAllocator_(rhs.ownAllocator_),
stack_(rhs.stack_),
stackTop_(rhs.stackTop_),
stackEnd_(rhs.stackEnd_),
initialCapacity_(rhs.initialCapacity_)
{
rhs.allocator_ = 0;
rhs.ownAllocator_ = 0;
rhs.stack_ = 0;
rhs.stackTop_ = 0;
rhs.stackEnd_ = 0;
rhs.initialCapacity_ = 0;
}
#endif
~Stack() {
Destroy();
}
#if RAPIDJSON_HAS_CXX11_RVALUE_REFS
Stack& operator=(Stack&& rhs) {
if (&rhs != this)
{
Destroy();
allocator_ = rhs.allocator_;
ownAllocator_ = rhs.ownAllocator_;
stack_ = rhs.stack_;
stackTop_ = rhs.stackTop_;
stackEnd_ = rhs.stackEnd_;
initialCapacity_ = rhs.initialCapacity_;
rhs.allocator_ = 0;
rhs.ownAllocator_ = 0;
rhs.stack_ = 0;
rhs.stackTop_ = 0;
rhs.stackEnd_ = 0;
rhs.initialCapacity_ = 0;
}
return *this;
}
#endif
void Swap(Stack& rhs) RAPIDJSON_NOEXCEPT {
internal::Swap(allocator_, rhs.allocator_);
internal::Swap(ownAllocator_, rhs.ownAllocator_);
internal::Swap(stack_, rhs.stack_);
internal::Swap(stackTop_, rhs.stackTop_);
internal::Swap(stackEnd_, rhs.stackEnd_);
internal::Swap(initialCapacity_, rhs.initialCapacity_);
}
void Clear() { stackTop_ = stack_; }
void ShrinkToFit() {
if (Empty()) {
// If the stack is empty, completely deallocate the memory.
Allocator::Free(stack_);
stack_ = 0;
stackTop_ = 0;
stackEnd_ = 0;
}
else
Resize(GetSize());
}
// Optimization note: try to minimize the size of this function for force inline.
// Expansion is run very infrequently, so it is moved to another (probably non-inline) function.
template<typename T>
RAPIDJSON_FORCEINLINE void Reserve(size_t count = 1) {
// Expand the stack if needed
if (RAPIDJSON_UNLIKELY(stackTop_ + sizeof(T) * count > stackEnd_))
Expand<T>(count);
}
template<typename T>
RAPIDJSON_FORCEINLINE T* Push(size_t count = 1) {
Reserve<T>(count);
return PushUnsafe<T>(count);
}
template<typename T>
RAPIDJSON_FORCEINLINE T* PushUnsafe(size_t count = 1) {
RAPIDJSON_ASSERT(stackTop_ + sizeof(T) * count <= stackEnd_);
T* ret = reinterpret_cast<T*>(stackTop_);
stackTop_ += sizeof(T) * count;
return ret;
}
template<typename T>
T* Pop(size_t count) {
RAPIDJSON_ASSERT(GetSize() >= count * sizeof(T));
stackTop_ -= count * sizeof(T);
return reinterpret_cast<T*>(stackTop_);
}
template<typename T>
T* Top() {
RAPIDJSON_ASSERT(GetSize() >= sizeof(T));
return reinterpret_cast<T*>(stackTop_ - sizeof(T));
}
template<typename T>
const T* Top() const {
RAPIDJSON_ASSERT(GetSize() >= sizeof(T));
return reinterpret_cast<T*>(stackTop_ - sizeof(T));
}
template<typename T>
T* End() { return reinterpret_cast<T*>(stackTop_); }
template<typename T>
const T* End() const { return reinterpret_cast<T*>(stackTop_); }
template<typename T>
T* Bottom() { return reinterpret_cast<T*>(stack_); }
template<typename T>
const T* Bottom() const { return reinterpret_cast<T*>(stack_); }
bool HasAllocator() const {
return allocator_ != 0;
}
Allocator& GetAllocator() {
RAPIDJSON_ASSERT(allocator_);
return *allocator_;
}
bool Empty() const { return stackTop_ == stack_; }
size_t GetSize() const { return static_cast<size_t>(stackTop_ - stack_); }
size_t GetCapacity() const { return static_cast<size_t>(stackEnd_ - stack_); }
private:
template<typename T>
void Expand(size_t count) {
// Only expand the capacity if the current stack exists. Otherwise just create a stack with initial capacity.
size_t newCapacity;
if (stack_ == 0) {
if (!allocator_)
ownAllocator_ = allocator_ = RAPIDJSON_NEW(Allocator());
newCapacity = initialCapacity_;
} else {
newCapacity = GetCapacity();
newCapacity += (newCapacity + 1) / 2;
}
size_t newSize = GetSize() + sizeof(T) * count;
if (newCapacity < newSize)
newCapacity = newSize;
Resize(newCapacity);
}
void Resize(size_t newCapacity) {
const size_t size = GetSize(); // Backup the current size
stack_ = static_cast<char*>(allocator_->Realloc(stack_, GetCapacity(), newCapacity));
stackTop_ = stack_ + size;
stackEnd_ = stack_ + newCapacity;
}
void Destroy() {
Allocator::Free(stack_);
RAPIDJSON_DELETE(ownAllocator_); // Only delete if it is owned by the stack
}
// Prohibit copy constructor & assignment operator.
Stack(const Stack&);
Stack& operator=(const Stack&);
Allocator* allocator_;
Allocator* ownAllocator_;
char *stack_;
char *stackTop_;
char *stackEnd_;
size_t initialCapacity_;
};
} // namespace internal
RAPIDJSON_NAMESPACE_END
#if defined(__clang__)
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_STACK_H_

58
rapidjson/internal/strfunc.h Normal file
View File

@ -0,0 +1,58 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_INTERNAL_STRFUNC_H_
#define RAPIDJSON_INTERNAL_STRFUNC_H_
#include "../stream.h"
RAPIDJSON_NAMESPACE_BEGIN
namespace internal {
//! Custom strlen() which works on different character types.
/*! \tparam Ch Character type (e.g. char, wchar_t, short)
\param s Null-terminated input string.
\return Number of characters in the string.
\note This has the same semantics as strlen(), the return value is not number of Unicode codepoints.
*/
template <typename Ch>
inline SizeType StrLen(const Ch* s) {
RAPIDJSON_ASSERT(s != 0);
const Ch* p = s;
while (*p) ++p;
return SizeType(p - s);
}
//! Returns number of code points in a encoded string.
template<typename Encoding>
bool CountStringCodePoint(const typename Encoding::Ch* s, SizeType length, SizeType* outCount) {
RAPIDJSON_ASSERT(s != 0);
RAPIDJSON_ASSERT(outCount != 0);
GenericStringStream<Encoding> is(s);
const typename Encoding::Ch* end = s + length;
SizeType count = 0;
while (is.src_ < end) {
unsigned codepoint;
if (!Encoding::Decode(is, &codepoint))
return false;
count++;
}
*outCount = count;
return true;
}
} // namespace internal
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_INTERNAL_STRFUNC_H_

269
rapidjson/internal/strtod.h Normal file
View File

@ -0,0 +1,269 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_STRTOD_
#define RAPIDJSON_STRTOD_
#include "ieee754.h"
#include "biginteger.h"
#include "diyfp.h"
#include "pow10.h"
RAPIDJSON_NAMESPACE_BEGIN
namespace internal {
inline double FastPath(double significand, int exp) {
if (exp < -308)
return 0.0;
else if (exp >= 0)
return significand * internal::Pow10(exp);
else
return significand / internal::Pow10(-exp);
}
inline double StrtodNormalPrecision(double d, int p) {
if (p < -308) {
// Prevent expSum < -308, making Pow10(p) = 0
d = FastPath(d, -308);
d = FastPath(d, p + 308);
}
else
d = FastPath(d, p);
return d;
}
template <typename T>
inline T Min3(T a, T b, T c) {
T m = a;
if (m > b) m = b;
if (m > c) m = c;
return m;
}
inline int CheckWithinHalfULP(double b, const BigInteger& d, int dExp) {
const Double db(b);
const uint64_t bInt = db.IntegerSignificand();
const int bExp = db.IntegerExponent();
const int hExp = bExp - 1;
int dS_Exp2 = 0, dS_Exp5 = 0, bS_Exp2 = 0, bS_Exp5 = 0, hS_Exp2 = 0, hS_Exp5 = 0;
// Adjust for decimal exponent
if (dExp >= 0) {
dS_Exp2 += dExp;
dS_Exp5 += dExp;
}
else {
bS_Exp2 -= dExp;
bS_Exp5 -= dExp;
hS_Exp2 -= dExp;
hS_Exp5 -= dExp;
}
// Adjust for binary exponent
if (bExp >= 0)
bS_Exp2 += bExp;
else {
dS_Exp2 -= bExp;
hS_Exp2 -= bExp;
}
// Adjust for half ulp exponent
if (hExp >= 0)
hS_Exp2 += hExp;
else {
dS_Exp2 -= hExp;
bS_Exp2 -= hExp;
}
// Remove common power of two factor from all three scaled values
int common_Exp2 = Min3(dS_Exp2, bS_Exp2, hS_Exp2);
dS_Exp2 -= common_Exp2;
bS_Exp2 -= common_Exp2;
hS_Exp2 -= common_Exp2;
BigInteger dS = d;
dS.MultiplyPow5(static_cast<unsigned>(dS_Exp5)) <<= static_cast<unsigned>(dS_Exp2);
BigInteger bS(bInt);
bS.MultiplyPow5(static_cast<unsigned>(bS_Exp5)) <<= static_cast<unsigned>(bS_Exp2);
BigInteger hS(1);
hS.MultiplyPow5(static_cast<unsigned>(hS_Exp5)) <<= static_cast<unsigned>(hS_Exp2);
BigInteger delta(0);
dS.Difference(bS, &delta);
return delta.Compare(hS);
}
inline bool StrtodFast(double d, int p, double* result) {
// Use fast path for string-to-double conversion if possible
// see http://www.exploringbinary.com/fast-path-decimal-to-floating-point-conversion/
if (p > 22 && p < 22 + 16) {
// Fast Path Cases In Disguise
d *= internal::Pow10(p - 22);
p = 22;
}
if (p >= -22 && p <= 22 && d <= 9007199254740991.0) { // 2^53 - 1
*result = FastPath(d, p);
return true;
}
else
return false;
}
// Compute an approximation and see if it is within 1/2 ULP
inline bool StrtodDiyFp(const char* decimals, size_t length, size_t decimalPosition, int exp, double* result) {
uint64_t significand = 0;
size_t i = 0; // 2^64 - 1 = 18446744073709551615, 1844674407370955161 = 0x1999999999999999
for (; i < length; i++) {
if (significand > RAPIDJSON_UINT64_C2(0x19999999, 0x99999999) ||
(significand == RAPIDJSON_UINT64_C2(0x19999999, 0x99999999) && decimals[i] > '5'))
break;
significand = significand * 10u + static_cast<unsigned>(decimals[i] - '0');
}
if (i < length && decimals[i] >= '5') // Rounding
significand++;
size_t remaining = length - i;
const unsigned kUlpShift = 3;
const unsigned kUlp = 1 << kUlpShift;
int64_t error = (remaining == 0) ? 0 : kUlp / 2;
DiyFp v(significand, 0);
v = v.Normalize();
error <<= -v.e;
const int dExp = static_cast<int>(decimalPosition) - static_cast<int>(i) + exp;
int actualExp;
DiyFp cachedPower = GetCachedPower10(dExp, &actualExp);
if (actualExp != dExp) {
static const DiyFp kPow10[] = {
DiyFp(RAPIDJSON_UINT64_C2(0xa0000000, 00000000), -60), // 10^1
DiyFp(RAPIDJSON_UINT64_C2(0xc8000000, 00000000), -57), // 10^2
DiyFp(RAPIDJSON_UINT64_C2(0xfa000000, 00000000), -54), // 10^3
DiyFp(RAPIDJSON_UINT64_C2(0x9c400000, 00000000), -50), // 10^4
DiyFp(RAPIDJSON_UINT64_C2(0xc3500000, 00000000), -47), // 10^5
DiyFp(RAPIDJSON_UINT64_C2(0xf4240000, 00000000), -44), // 10^6
DiyFp(RAPIDJSON_UINT64_C2(0x98968000, 00000000), -40) // 10^7
};
int adjustment = dExp - actualExp - 1;
RAPIDJSON_ASSERT(adjustment >= 0 && adjustment < 7);
v = v * kPow10[adjustment];
if (length + static_cast<unsigned>(adjustment)> 19u) // has more digits than decimal digits in 64-bit
error += kUlp / 2;
}
v = v * cachedPower;
error += kUlp + (error == 0 ? 0 : 1);
const int oldExp = v.e;
v = v.Normalize();
error <<= oldExp - v.e;
const unsigned effectiveSignificandSize = Double::EffectiveSignificandSize(64 + v.e);
unsigned precisionSize = 64 - effectiveSignificandSize;
if (precisionSize + kUlpShift >= 64) {
unsigned scaleExp = (precisionSize + kUlpShift) - 63;
v.f >>= scaleExp;
v.e += scaleExp;
error = (error >> scaleExp) + 1 + static_cast<int>(kUlp);
precisionSize -= scaleExp;
}
DiyFp rounded(v.f >> precisionSize, v.e + static_cast<int>(precisionSize));
const uint64_t precisionBits = (v.f & ((uint64_t(1) << precisionSize) - 1)) * kUlp;
const uint64_t halfWay = (uint64_t(1) << (precisionSize - 1)) * kUlp;
if (precisionBits >= halfWay + static_cast<unsigned>(error)) {
rounded.f++;
if (rounded.f & (DiyFp::kDpHiddenBit << 1)) { // rounding overflows mantissa (issue #340)
rounded.f >>= 1;
rounded.e++;
}
}
*result = rounded.ToDouble();
return halfWay - static_cast<unsigned>(error) >= precisionBits || precisionBits >= halfWay + static_cast<unsigned>(error);
}
inline double StrtodBigInteger(double approx, const char* decimals, size_t length, size_t decimalPosition, int exp) {
const BigInteger dInt(decimals, length);
const int dExp = static_cast<int>(decimalPosition) - static_cast<int>(length) + exp;
Double a(approx);
int cmp = CheckWithinHalfULP(a.Value(), dInt, dExp);
if (cmp < 0)
return a.Value(); // within half ULP
else if (cmp == 0) {
// Round towards even
if (a.Significand() & 1)
return a.NextPositiveDouble();
else
return a.Value();
}
else // adjustment
return a.NextPositiveDouble();
}
inline double StrtodFullPrecision(double d, int p, const char* decimals, size_t length, size_t decimalPosition, int exp) {
RAPIDJSON_ASSERT(d >= 0.0);
RAPIDJSON_ASSERT(length >= 1);
double result;
if (StrtodFast(d, p, &result))
return result;
// Trim leading zeros
while (*decimals == '0' && length > 1) {
length--;
decimals++;
decimalPosition--;
}
// Trim trailing zeros
while (decimals[length - 1] == '0' && length > 1) {
length--;
decimalPosition--;
exp++;
}
// Trim right-most digits
const int kMaxDecimalDigit = 780;
if (static_cast<int>(length) > kMaxDecimalDigit) {
int delta = (static_cast<int>(length) - kMaxDecimalDigit);
exp += delta;
decimalPosition -= static_cast<unsigned>(delta);
length = kMaxDecimalDigit;
}
// If too small, underflow to zero
if (int(length) + exp < -324)
return 0.0;
if (StrtodDiyFp(decimals, length, decimalPosition, exp, &result))
return result;
// Use approximation from StrtodDiyFp and make adjustment with BigInteger comparison
return StrtodBigInteger(result, decimals, length, decimalPosition, exp);
}
} // namespace internal
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_STRTOD_

46
rapidjson/internal/swap.h Normal file
View File

@ -0,0 +1,46 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_INTERNAL_SWAP_H_
#define RAPIDJSON_INTERNAL_SWAP_H_
#include "../rapidjson.h"
#if defined(__clang__)
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(c++98-compat)
#endif
RAPIDJSON_NAMESPACE_BEGIN
namespace internal {
//! Custom swap() to avoid dependency on C++ <algorithm> header
/*! \tparam T Type of the arguments to swap, should be instantiated with primitive C++ types only.
\note This has the same semantics as std::swap().
*/
template <typename T>
inline void Swap(T& a, T& b) RAPIDJSON_NOEXCEPT {
T tmp = a;
a = b;
b = tmp;
}
} // namespace internal
RAPIDJSON_NAMESPACE_END
#if defined(__clang__)
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_INTERNAL_SWAP_H_

115
rapidjson/istreamwrapper.h Normal file
View File

@ -0,0 +1,115 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_ISTREAMWRAPPER_H_
#define RAPIDJSON_ISTREAMWRAPPER_H_
#include "stream.h"
#include <iosfwd>
#ifdef __clang__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(padded)
#endif
#ifdef _MSC_VER
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(4351) // new behavior: elements of array 'array' will be default initialized
#endif
RAPIDJSON_NAMESPACE_BEGIN
//! Wrapper of \c std::basic_istream into RapidJSON's Stream concept.
/*!
The classes can be wrapped including but not limited to:
- \c std::istringstream
- \c std::stringstream
- \c std::wistringstream
- \c std::wstringstream
- \c std::ifstream
- \c std::fstream
- \c std::wifstream
- \c std::wfstream
\tparam StreamType Class derived from \c std::basic_istream.
*/
template <typename StreamType>
class BasicIStreamWrapper {
public:
typedef typename StreamType::char_type Ch;
BasicIStreamWrapper(StreamType& stream) : stream_(stream), count_(), peekBuffer_() {}
Ch Peek() const {
typename StreamType::int_type c = stream_.peek();
return RAPIDJSON_LIKELY(c != StreamType::traits_type::eof()) ? static_cast<Ch>(c) : '\0';
}
Ch Take() {
typename StreamType::int_type c = stream_.get();
if (RAPIDJSON_LIKELY(c != StreamType::traits_type::eof())) {
count_++;
return static_cast<Ch>(c);
}
else
return '\0';
}
// tellg() may return -1 when failed. So we count by ourself.
size_t Tell() const { return count_; }
Ch* PutBegin() { RAPIDJSON_ASSERT(false); return 0; }
void Put(Ch) { RAPIDJSON_ASSERT(false); }
void Flush() { RAPIDJSON_ASSERT(false); }
size_t PutEnd(Ch*) { RAPIDJSON_ASSERT(false); return 0; }
// For encoding detection only.
const Ch* Peek4() const {
RAPIDJSON_ASSERT(sizeof(Ch) == 1); // Only usable for byte stream.
int i;
bool hasError = false;
for (i = 0; i < 4; ++i) {
typename StreamType::int_type c = stream_.get();
if (c == StreamType::traits_type::eof()) {
hasError = true;
stream_.clear();
break;
}
peekBuffer_[i] = static_cast<Ch>(c);
}
for (--i; i >= 0; --i)
stream_.putback(peekBuffer_[i]);
return !hasError ? peekBuffer_ : 0;
}
private:
BasicIStreamWrapper(const BasicIStreamWrapper&);
BasicIStreamWrapper& operator=(const BasicIStreamWrapper&);
StreamType& stream_;
size_t count_; //!< Number of characters read. Note:
mutable Ch peekBuffer_[4];
};
typedef BasicIStreamWrapper<std::istream> IStreamWrapper;
typedef BasicIStreamWrapper<std::wistream> WIStreamWrapper;
#if defined(__clang__) || defined(_MSC_VER)
RAPIDJSON_DIAG_POP
#endif
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_ISTREAMWRAPPER_H_

70
rapidjson/memorybuffer.h Normal file
View File

@ -0,0 +1,70 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_MEMORYBUFFER_H_
#define RAPIDJSON_MEMORYBUFFER_H_
#include "stream.h"
#include "internal/stack.h"
RAPIDJSON_NAMESPACE_BEGIN
//! Represents an in-memory output byte stream.
/*!
This class is mainly for being wrapped by EncodedOutputStream or AutoUTFOutputStream.
It is similar to FileWriteBuffer but the destination is an in-memory buffer instead of a file.
Differences between MemoryBuffer and StringBuffer:
1. StringBuffer has Encoding but MemoryBuffer is only a byte buffer.
2. StringBuffer::GetString() returns a null-terminated string. MemoryBuffer::GetBuffer() returns a buffer without terminator.
\tparam Allocator type for allocating memory buffer.
\note implements Stream concept
*/
template <typename Allocator = CrtAllocator>
struct GenericMemoryBuffer {
typedef char Ch; // byte
GenericMemoryBuffer(Allocator* allocator = 0, size_t capacity = kDefaultCapacity) : stack_(allocator, capacity) {}
void Put(Ch c) { *stack_.template Push<Ch>() = c; }
void Flush() {}
void Clear() { stack_.Clear(); }
void ShrinkToFit() { stack_.ShrinkToFit(); }
Ch* Push(size_t count) { return stack_.template Push<Ch>(count); }
void Pop(size_t count) { stack_.template Pop<Ch>(count); }
const Ch* GetBuffer() const {
return stack_.template Bottom<Ch>();
}
size_t GetSize() const { return stack_.GetSize(); }
static const size_t kDefaultCapacity = 256;
mutable internal::Stack<Allocator> stack_;
};
typedef GenericMemoryBuffer<> MemoryBuffer;
//! Implement specialized version of PutN() with memset() for better performance.
template<>
inline void PutN(MemoryBuffer& memoryBuffer, char c, size_t n) {
std::memset(memoryBuffer.stack_.Push<char>(n), c, n * sizeof(c));
}
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_MEMORYBUFFER_H_

71
rapidjson/memorystream.h Normal file
View File

@ -0,0 +1,71 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_MEMORYSTREAM_H_
#define RAPIDJSON_MEMORYSTREAM_H_
#include "stream.h"
#ifdef __clang__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(unreachable-code)
RAPIDJSON_DIAG_OFF(missing-noreturn)
#endif
RAPIDJSON_NAMESPACE_BEGIN
//! Represents an in-memory input byte stream.
/*!
This class is mainly for being wrapped by EncodedInputStream or AutoUTFInputStream.
It is similar to FileReadBuffer but the source is an in-memory buffer instead of a file.
Differences between MemoryStream and StringStream:
1. StringStream has encoding but MemoryStream is a byte stream.
2. MemoryStream needs size of the source buffer and the buffer don't need to be null terminated. StringStream assume null-terminated string as source.
3. MemoryStream supports Peek4() for encoding detection. StringStream is specified with an encoding so it should not have Peek4().
\note implements Stream concept
*/
struct MemoryStream {
typedef char Ch; // byte
MemoryStream(const Ch *src, size_t size) : src_(src), begin_(src), end_(src + size), size_(size) {}
Ch Peek() const { return RAPIDJSON_UNLIKELY(src_ == end_) ? '\0' : *src_; }
Ch Take() { return RAPIDJSON_UNLIKELY(src_ == end_) ? '\0' : *src_++; }
size_t Tell() const { return static_cast<size_t>(src_ - begin_); }
Ch* PutBegin() { RAPIDJSON_ASSERT(false); return 0; }
void Put(Ch) { RAPIDJSON_ASSERT(false); }
void Flush() { RAPIDJSON_ASSERT(false); }
size_t PutEnd(Ch*) { RAPIDJSON_ASSERT(false); return 0; }
// For encoding detection only.
const Ch* Peek4() const {
return Tell() + 4 <= size_ ? src_ : 0;
}
const Ch* src_; //!< Current read position.
const Ch* begin_; //!< Original head of the string.
const Ch* end_; //!< End of stream.
size_t size_; //!< Size of the stream.
};
RAPIDJSON_NAMESPACE_END
#ifdef __clang__
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_MEMORYBUFFER_H_

316
rapidjson/msinttypes/inttypes.h Normal file
View File

@ -0,0 +1,316 @@
// ISO C9x compliant inttypes.h for Microsoft Visual Studio
// Based on ISO/IEC 9899:TC2 Committee draft (May 6, 2005) WG14/N1124
//
// Copyright (c) 2006-2013 Alexander Chemeris
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are met:
//
// 1. Redistributions of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.
//
// 2. Redistributions in binary form must reproduce the above copyright
// notice, this list of conditions and the following disclaimer in the
// documentation and/or other materials provided with the distribution.
//
// 3. Neither the name of the product nor the names of its contributors may
// be used to endorse or promote products derived from this software
// without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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.
//
///////////////////////////////////////////////////////////////////////////////
// The above software in this distribution may have been modified by
// THL A29 Limited ("Tencent Modifications").
// All Tencent Modifications are Copyright (C) 2015 THL A29 Limited.
#ifndef _MSC_VER // [
#error "Use this header only with Microsoft Visual C++ compilers!"
#endif // _MSC_VER ]
#ifndef _MSC_INTTYPES_H_ // [
#define _MSC_INTTYPES_H_
#if _MSC_VER > 1000
#pragma once
#endif
#include "stdint.h"
// miloyip: VC supports inttypes.h since VC2013
#if _MSC_VER >= 1800
#include <inttypes.h>
#else
// 7.8 Format conversion of integer types
typedef struct {
intmax_t quot;
intmax_t rem;
} imaxdiv_t;
// 7.8.1 Macros for format specifiers
#if !defined(__cplusplus) || defined(__STDC_FORMAT_MACROS) // [ See footnote 185 at page 198
// The fprintf macros for signed integers are:
#define PRId8 "d"
#define PRIi8 "i"
#define PRIdLEAST8 "d"
#define PRIiLEAST8 "i"
#define PRIdFAST8 "d"
#define PRIiFAST8 "i"
#define PRId16 "hd"
#define PRIi16 "hi"
#define PRIdLEAST16 "hd"
#define PRIiLEAST16 "hi"
#define PRIdFAST16 "hd"
#define PRIiFAST16 "hi"
#define PRId32 "I32d"
#define PRIi32 "I32i"
#define PRIdLEAST32 "I32d"
#define PRIiLEAST32 "I32i"
#define PRIdFAST32 "I32d"
#define PRIiFAST32 "I32i"
#define PRId64 "I64d"
#define PRIi64 "I64i"
#define PRIdLEAST64 "I64d"
#define PRIiLEAST64 "I64i"
#define PRIdFAST64 "I64d"
#define PRIiFAST64 "I64i"
#define PRIdMAX "I64d"
#define PRIiMAX "I64i"
#define PRIdPTR "Id"
#define PRIiPTR "Ii"
// The fprintf macros for unsigned integers are:
#define PRIo8 "o"
#define PRIu8 "u"
#define PRIx8 "x"
#define PRIX8 "X"
#define PRIoLEAST8 "o"
#define PRIuLEAST8 "u"
#define PRIxLEAST8 "x"
#define PRIXLEAST8 "X"
#define PRIoFAST8 "o"
#define PRIuFAST8 "u"
#define PRIxFAST8 "x"
#define PRIXFAST8 "X"
#define PRIo16 "ho"
#define PRIu16 "hu"
#define PRIx16 "hx"
#define PRIX16 "hX"
#define PRIoLEAST16 "ho"
#define PRIuLEAST16 "hu"
#define PRIxLEAST16 "hx"
#define PRIXLEAST16 "hX"
#define PRIoFAST16 "ho"
#define PRIuFAST16 "hu"
#define PRIxFAST16 "hx"
#define PRIXFAST16 "hX"
#define PRIo32 "I32o"
#define PRIu32 "I32u"
#define PRIx32 "I32x"
#define PRIX32 "I32X"
#define PRIoLEAST32 "I32o"
#define PRIuLEAST32 "I32u"
#define PRIxLEAST32 "I32x"
#define PRIXLEAST32 "I32X"
#define PRIoFAST32 "I32o"
#define PRIuFAST32 "I32u"
#define PRIxFAST32 "I32x"
#define PRIXFAST32 "I32X"
#define PRIo64 "I64o"
#define PRIu64 "I64u"
#define PRIx64 "I64x"
#define PRIX64 "I64X"
#define PRIoLEAST64 "I64o"
#define PRIuLEAST64 "I64u"
#define PRIxLEAST64 "I64x"
#define PRIXLEAST64 "I64X"
#define PRIoFAST64 "I64o"
#define PRIuFAST64 "I64u"
#define PRIxFAST64 "I64x"
#define PRIXFAST64 "I64X"
#define PRIoMAX "I64o"
#define PRIuMAX "I64u"
#define PRIxMAX "I64x"
#define PRIXMAX "I64X"
#define PRIoPTR "Io"
#define PRIuPTR "Iu"
#define PRIxPTR "Ix"
#define PRIXPTR "IX"
// The fscanf macros for signed integers are:
#define SCNd8 "d"
#define SCNi8 "i"
#define SCNdLEAST8 "d"
#define SCNiLEAST8 "i"
#define SCNdFAST8 "d"
#define SCNiFAST8 "i"
#define SCNd16 "hd"
#define SCNi16 "hi"
#define SCNdLEAST16 "hd"
#define SCNiLEAST16 "hi"
#define SCNdFAST16 "hd"
#define SCNiFAST16 "hi"
#define SCNd32 "ld"
#define SCNi32 "li"
#define SCNdLEAST32 "ld"
#define SCNiLEAST32 "li"
#define SCNdFAST32 "ld"
#define SCNiFAST32 "li"
#define SCNd64 "I64d"
#define SCNi64 "I64i"
#define SCNdLEAST64 "I64d"
#define SCNiLEAST64 "I64i"
#define SCNdFAST64 "I64d"
#define SCNiFAST64 "I64i"
#define SCNdMAX "I64d"
#define SCNiMAX "I64i"
#ifdef _WIN64 // [
# define SCNdPTR "I64d"
# define SCNiPTR "I64i"
#else // _WIN64 ][
# define SCNdPTR "ld"
# define SCNiPTR "li"
#endif // _WIN64 ]
// The fscanf macros for unsigned integers are:
#define SCNo8 "o"
#define SCNu8 "u"
#define SCNx8 "x"
#define SCNX8 "X"
#define SCNoLEAST8 "o"
#define SCNuLEAST8 "u"
#define SCNxLEAST8 "x"
#define SCNXLEAST8 "X"
#define SCNoFAST8 "o"
#define SCNuFAST8 "u"
#define SCNxFAST8 "x"
#define SCNXFAST8 "X"
#define SCNo16 "ho"
#define SCNu16 "hu"
#define SCNx16 "hx"
#define SCNX16 "hX"
#define SCNoLEAST16 "ho"
#define SCNuLEAST16 "hu"
#define SCNxLEAST16 "hx"
#define SCNXLEAST16 "hX"
#define SCNoFAST16 "ho"
#define SCNuFAST16 "hu"
#define SCNxFAST16 "hx"
#define SCNXFAST16 "hX"
#define SCNo32 "lo"
#define SCNu32 "lu"
#define SCNx32 "lx"
#define SCNX32 "lX"
#define SCNoLEAST32 "lo"
#define SCNuLEAST32 "lu"
#define SCNxLEAST32 "lx"
#define SCNXLEAST32 "lX"
#define SCNoFAST32 "lo"
#define SCNuFAST32 "lu"
#define SCNxFAST32 "lx"
#define SCNXFAST32 "lX"
#define SCNo64 "I64o"
#define SCNu64 "I64u"
#define SCNx64 "I64x"
#define SCNX64 "I64X"
#define SCNoLEAST64 "I64o"
#define SCNuLEAST64 "I64u"
#define SCNxLEAST64 "I64x"
#define SCNXLEAST64 "I64X"
#define SCNoFAST64 "I64o"
#define SCNuFAST64 "I64u"
#define SCNxFAST64 "I64x"
#define SCNXFAST64 "I64X"
#define SCNoMAX "I64o"
#define SCNuMAX "I64u"
#define SCNxMAX "I64x"
#define SCNXMAX "I64X"
#ifdef _WIN64 // [
# define SCNoPTR "I64o"
# define SCNuPTR "I64u"
# define SCNxPTR "I64x"
# define SCNXPTR "I64X"
#else // _WIN64 ][
# define SCNoPTR "lo"
# define SCNuPTR "lu"
# define SCNxPTR "lx"
# define SCNXPTR "lX"
#endif // _WIN64 ]
#endif // __STDC_FORMAT_MACROS ]
// 7.8.2 Functions for greatest-width integer types
// 7.8.2.1 The imaxabs function
#define imaxabs _abs64
// 7.8.2.2 The imaxdiv function
// This is modified version of div() function from Microsoft's div.c found
// in %MSVC.NET%\crt\src\div.c
#ifdef STATIC_IMAXDIV // [
static
#else // STATIC_IMAXDIV ][
_inline
#endif // STATIC_IMAXDIV ]
imaxdiv_t __cdecl imaxdiv(intmax_t numer, intmax_t denom)
{
imaxdiv_t result;
result.quot = numer / denom;
result.rem = numer % denom;
if (numer < 0 && result.rem > 0) {
// did division wrong; must fix up
++result.quot;
result.rem -= denom;
}
return result;
}
// 7.8.2.3 The strtoimax and strtoumax functions
#define strtoimax _strtoi64
#define strtoumax _strtoui64
// 7.8.2.4 The wcstoimax and wcstoumax functions
#define wcstoimax _wcstoi64
#define wcstoumax _wcstoui64
#endif // _MSC_VER >= 1800
#endif // _MSC_INTTYPES_H_ ]

300
rapidjson/msinttypes/stdint.h Normal file
View File

@ -0,0 +1,300 @@
// ISO C9x compliant stdint.h for Microsoft Visual Studio
// Based on ISO/IEC 9899:TC2 Committee draft (May 6, 2005) WG14/N1124
//
// Copyright (c) 2006-2013 Alexander Chemeris
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are met:
//
// 1. Redistributions of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.
//
// 2. Redistributions in binary form must reproduce the above copyright
// notice, this list of conditions and the following disclaimer in the
// documentation and/or other materials provided with the distribution.
//
// 3. Neither the name of the product nor the names of its contributors may
// be used to endorse or promote products derived from this software
// without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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.
//
///////////////////////////////////////////////////////////////////////////////
// The above software in this distribution may have been modified by
// THL A29 Limited ("Tencent Modifications").
// All Tencent Modifications are Copyright (C) 2015 THL A29 Limited.
#ifndef _MSC_VER // [
#error "Use this header only with Microsoft Visual C++ compilers!"
#endif // _MSC_VER ]
#ifndef _MSC_STDINT_H_ // [
#define _MSC_STDINT_H_
#if _MSC_VER > 1000
#pragma once
#endif
// miloyip: Originally Visual Studio 2010 uses its own stdint.h. However it generates warning with INT64_C(), so change to use this file for vs2010.
#if _MSC_VER >= 1600 // [
#include <stdint.h>
#if !defined(__cplusplus) || defined(__STDC_CONSTANT_MACROS) // [ See footnote 224 at page 260
#undef INT8_C
#undef INT16_C
#undef INT32_C
#undef INT64_C
#undef UINT8_C
#undef UINT16_C
#undef UINT32_C
#undef UINT64_C
// 7.18.4.1 Macros for minimum-width integer constants
#define INT8_C(val) val##i8
#define INT16_C(val) val##i16
#define INT32_C(val) val##i32
#define INT64_C(val) val##i64
#define UINT8_C(val) val##ui8
#define UINT16_C(val) val##ui16
#define UINT32_C(val) val##ui32
#define UINT64_C(val) val##ui64
// 7.18.4.2 Macros for greatest-width integer constants
// These #ifndef's are needed to prevent collisions with <boost/cstdint.hpp>.
// Check out Issue 9 for the details.
#ifndef INTMAX_C // [
# define INTMAX_C INT64_C
#endif // INTMAX_C ]
#ifndef UINTMAX_C // [
# define UINTMAX_C UINT64_C
#endif // UINTMAX_C ]
#endif // __STDC_CONSTANT_MACROS ]
#else // ] _MSC_VER >= 1700 [
#include <limits.h>
// For Visual Studio 6 in C++ mode and for many Visual Studio versions when
// compiling for ARM we have to wrap <wchar.h> include with 'extern "C++" {}'
// or compiler would give many errors like this:
// error C2733: second C linkage of overloaded function 'wmemchr' not allowed
#if defined(__cplusplus) && !defined(_M_ARM)
extern "C" {
#endif
# include <wchar.h>
#if defined(__cplusplus) && !defined(_M_ARM)
}
#endif
// Define _W64 macros to mark types changing their size, like intptr_t.
#ifndef _W64
# if !defined(__midl) && (defined(_X86_) || defined(_M_IX86)) && _MSC_VER >= 1300
# define _W64 __w64
# else
# define _W64
# endif
#endif
// 7.18.1 Integer types
// 7.18.1.1 Exact-width integer types
// Visual Studio 6 and Embedded Visual C++ 4 doesn't
// realize that, e.g. char has the same size as __int8
// so we give up on __intX for them.
#if (_MSC_VER < 1300)
typedef signed char int8_t;
typedef signed short int16_t;
typedef signed int int32_t;
typedef unsigned char uint8_t;
typedef unsigned short uint16_t;
typedef unsigned int uint32_t;
#else
typedef signed __int8 int8_t;
typedef signed __int16 int16_t;
typedef signed __int32 int32_t;
typedef unsigned __int8 uint8_t;
typedef unsigned __int16 uint16_t;
typedef unsigned __int32 uint32_t;
#endif
typedef signed __int64 int64_t;
typedef unsigned __int64 uint64_t;
// 7.18.1.2 Minimum-width integer types
typedef int8_t int_least8_t;
typedef int16_t int_least16_t;
typedef int32_t int_least32_t;
typedef int64_t int_least64_t;
typedef uint8_t uint_least8_t;
typedef uint16_t uint_least16_t;
typedef uint32_t uint_least32_t;
typedef uint64_t uint_least64_t;
// 7.18.1.3 Fastest minimum-width integer types
typedef int8_t int_fast8_t;
typedef int16_t int_fast16_t;
typedef int32_t int_fast32_t;
typedef int64_t int_fast64_t;
typedef uint8_t uint_fast8_t;
typedef uint16_t uint_fast16_t;
typedef uint32_t uint_fast32_t;
typedef uint64_t uint_fast64_t;
// 7.18.1.4 Integer types capable of holding object pointers
#ifdef _WIN64 // [
typedef signed __int64 intptr_t;
typedef unsigned __int64 uintptr_t;
#else // _WIN64 ][
typedef _W64 signed int intptr_t;
typedef _W64 unsigned int uintptr_t;
#endif // _WIN64 ]
// 7.18.1.5 Greatest-width integer types
typedef int64_t intmax_t;
typedef uint64_t uintmax_t;
// 7.18.2 Limits of specified-width integer types
#if !defined(__cplusplus) || defined(__STDC_LIMIT_MACROS) // [ See footnote 220 at page 257 and footnote 221 at page 259
// 7.18.2.1 Limits of exact-width integer types
#define INT8_MIN ((int8_t)_I8_MIN)
#define INT8_MAX _I8_MAX
#define INT16_MIN ((int16_t)_I16_MIN)
#define INT16_MAX _I16_MAX
#define INT32_MIN ((int32_t)_I32_MIN)
#define INT32_MAX _I32_MAX
#define INT64_MIN ((int64_t)_I64_MIN)
#define INT64_MAX _I64_MAX
#define UINT8_MAX _UI8_MAX
#define UINT16_MAX _UI16_MAX
#define UINT32_MAX _UI32_MAX
#define UINT64_MAX _UI64_MAX
// 7.18.2.2 Limits of minimum-width integer types
#define INT_LEAST8_MIN INT8_MIN
#define INT_LEAST8_MAX INT8_MAX
#define INT_LEAST16_MIN INT16_MIN
#define INT_LEAST16_MAX INT16_MAX
#define INT_LEAST32_MIN INT32_MIN
#define INT_LEAST32_MAX INT32_MAX
#define INT_LEAST64_MIN INT64_MIN
#define INT_LEAST64_MAX INT64_MAX
#define UINT_LEAST8_MAX UINT8_MAX
#define UINT_LEAST16_MAX UINT16_MAX
#define UINT_LEAST32_MAX UINT32_MAX
#define UINT_LEAST64_MAX UINT64_MAX
// 7.18.2.3 Limits of fastest minimum-width integer types
#define INT_FAST8_MIN INT8_MIN
#define INT_FAST8_MAX INT8_MAX
#define INT_FAST16_MIN INT16_MIN
#define INT_FAST16_MAX INT16_MAX
#define INT_FAST32_MIN INT32_MIN
#define INT_FAST32_MAX INT32_MAX
#define INT_FAST64_MIN INT64_MIN
#define INT_FAST64_MAX INT64_MAX
#define UINT_FAST8_MAX UINT8_MAX
#define UINT_FAST16_MAX UINT16_MAX
#define UINT_FAST32_MAX UINT32_MAX
#define UINT_FAST64_MAX UINT64_MAX
// 7.18.2.4 Limits of integer types capable of holding object pointers
#ifdef _WIN64 // [
# define INTPTR_MIN INT64_MIN
# define INTPTR_MAX INT64_MAX
# define UINTPTR_MAX UINT64_MAX
#else // _WIN64 ][
# define INTPTR_MIN INT32_MIN
# define INTPTR_MAX INT32_MAX
# define UINTPTR_MAX UINT32_MAX
#endif // _WIN64 ]
// 7.18.2.5 Limits of greatest-width integer types
#define INTMAX_MIN INT64_MIN
#define INTMAX_MAX INT64_MAX
#define UINTMAX_MAX UINT64_MAX
// 7.18.3 Limits of other integer types
#ifdef _WIN64 // [
# define PTRDIFF_MIN _I64_MIN
# define PTRDIFF_MAX _I64_MAX
#else // _WIN64 ][
# define PTRDIFF_MIN _I32_MIN
# define PTRDIFF_MAX _I32_MAX
#endif // _WIN64 ]
#define SIG_ATOMIC_MIN INT_MIN
#define SIG_ATOMIC_MAX INT_MAX
#ifndef SIZE_MAX // [
# ifdef _WIN64 // [
# define SIZE_MAX _UI64_MAX
# else // _WIN64 ][
# define SIZE_MAX _UI32_MAX
# endif // _WIN64 ]
#endif // SIZE_MAX ]
// WCHAR_MIN and WCHAR_MAX are also defined in <wchar.h>
#ifndef WCHAR_MIN // [
# define WCHAR_MIN 0
#endif // WCHAR_MIN ]
#ifndef WCHAR_MAX // [
# define WCHAR_MAX _UI16_MAX
#endif // WCHAR_MAX ]
#define WINT_MIN 0
#define WINT_MAX _UI16_MAX
#endif // __STDC_LIMIT_MACROS ]
// 7.18.4 Limits of other integer types
#if !defined(__cplusplus) || defined(__STDC_CONSTANT_MACROS) // [ See footnote 224 at page 260
// 7.18.4.1 Macros for minimum-width integer constants
#define INT8_C(val) val##i8
#define INT16_C(val) val##i16
#define INT32_C(val) val##i32
#define INT64_C(val) val##i64
#define UINT8_C(val) val##ui8
#define UINT16_C(val) val##ui16
#define UINT32_C(val) val##ui32
#define UINT64_C(val) val##ui64
// 7.18.4.2 Macros for greatest-width integer constants
// These #ifndef's are needed to prevent collisions with <boost/cstdint.hpp>.
// Check out Issue 9 for the details.
#ifndef INTMAX_C // [
# define INTMAX_C INT64_C
#endif // INTMAX_C ]
#ifndef UINTMAX_C // [
# define UINTMAX_C UINT64_C
#endif // UINTMAX_C ]
#endif // __STDC_CONSTANT_MACROS ]
#endif // _MSC_VER >= 1600 ]
#endif // _MSC_STDINT_H_ ]

81
rapidjson/ostreamwrapper.h Normal file
View File

@ -0,0 +1,81 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_OSTREAMWRAPPER_H_
#define RAPIDJSON_OSTREAMWRAPPER_H_
#include "stream.h"
#include <iosfwd>
#ifdef __clang__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(padded)
#endif
RAPIDJSON_NAMESPACE_BEGIN
//! Wrapper of \c std::basic_ostream into RapidJSON's Stream concept.
/*!
The classes can be wrapped including but not limited to:
- \c std::ostringstream
- \c std::stringstream
- \c std::wpstringstream
- \c std::wstringstream
- \c std::ifstream
- \c std::fstream
- \c std::wofstream
- \c std::wfstream
\tparam StreamType Class derived from \c std::basic_ostream.
*/
template <typename StreamType>
class BasicOStreamWrapper {
public:
typedef typename StreamType::char_type Ch;
BasicOStreamWrapper(StreamType& stream) : stream_(stream) {}
void Put(Ch c) {
stream_.put(c);
}
void Flush() {
stream_.flush();
}
// Not implemented
char Peek() const { RAPIDJSON_ASSERT(false); return 0; }
char Take() { RAPIDJSON_ASSERT(false); return 0; }
size_t Tell() const { RAPIDJSON_ASSERT(false); return 0; }
char* PutBegin() { RAPIDJSON_ASSERT(false); return 0; }
size_t PutEnd(char*) { RAPIDJSON_ASSERT(false); return 0; }
private:
BasicOStreamWrapper(const BasicOStreamWrapper&);
BasicOStreamWrapper& operator=(const BasicOStreamWrapper&);
StreamType& stream_;
};
typedef BasicOStreamWrapper<std::ostream> OStreamWrapper;
typedef BasicOStreamWrapper<std::wostream> WOStreamWrapper;
#ifdef __clang__
RAPIDJSON_DIAG_POP
#endif
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_OSTREAMWRAPPER_H_

1358
rapidjson/pointer.h Normal file
View File

File diff suppressed because it is too large Load Diff

275
rapidjson/prettywriter.h Normal file
View File

@ -0,0 +1,275 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_PRETTYWRITER_H_
#define RAPIDJSON_PRETTYWRITER_H_
#include "writer.h"
#ifdef __GNUC__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(effc++)
#endif
#if defined(__clang__)
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(c++98-compat)
#endif
RAPIDJSON_NAMESPACE_BEGIN
//! Combination of PrettyWriter format flags.
/*! \see PrettyWriter::SetFormatOptions
*/
enum PrettyFormatOptions {
kFormatDefault = 0, //!< Default pretty formatting.
kFormatSingleLineArray = 1 //!< Format arrays on a single line.
};
//! Writer with indentation and spacing.
/*!
\tparam OutputStream Type of ouptut os.
\tparam SourceEncoding Encoding of source string.
\tparam TargetEncoding Encoding of output stream.
\tparam StackAllocator Type of allocator for allocating memory of stack.
*/
template<typename OutputStream, typename SourceEncoding = UTF8<>, typename TargetEncoding = UTF8<>, typename StackAllocator = CrtAllocator, unsigned writeFlags = kWriteDefaultFlags>
class PrettyWriter : public Writer<OutputStream, SourceEncoding, TargetEncoding, StackAllocator, writeFlags> {
public:
typedef Writer<OutputStream, SourceEncoding, TargetEncoding, StackAllocator> Base;
typedef typename Base::Ch Ch;
//! Constructor
/*! \param os Output stream.
\param allocator User supplied allocator. If it is null, it will create a private one.
\param levelDepth Initial capacity of stack.
*/
explicit PrettyWriter(OutputStream& os, StackAllocator* allocator = 0, size_t levelDepth = Base::kDefaultLevelDepth) :
Base(os, allocator, levelDepth), indentChar_(' '), indentCharCount_(4), formatOptions_(kFormatDefault) {}
explicit PrettyWriter(StackAllocator* allocator = 0, size_t levelDepth = Base::kDefaultLevelDepth) :
Base(allocator, levelDepth), indentChar_(' '), indentCharCount_(4) {}
#if RAPIDJSON_HAS_CXX11_RVALUE_REFS
PrettyWriter(PrettyWriter&& rhs) :
Base(std::forward<PrettyWriter>(rhs)), indentChar_(rhs.indentChar_), indentCharCount_(rhs.indentCharCount_), formatOptions_(rhs.formatOptions_) {}
#endif
//! Set custom indentation.
/*! \param indentChar Character for indentation. Must be whitespace character (' ', '\\t', '\\n', '\\r').
\param indentCharCount Number of indent characters for each indentation level.
\note The default indentation is 4 spaces.
*/
PrettyWriter& SetIndent(Ch indentChar, unsigned indentCharCount) {
RAPIDJSON_ASSERT(indentChar == ' ' || indentChar == '\t' || indentChar == '\n' || indentChar == '\r');
indentChar_ = indentChar;
indentCharCount_ = indentCharCount;
return *this;
}
//! Set pretty writer formatting options.
/*! \param options Formatting options.
*/
PrettyWriter& SetFormatOptions(PrettyFormatOptions options) {
formatOptions_ = options;
return *this;
}
/*! @name Implementation of Handler
\see Handler
*/
//@{
bool Null() { PrettyPrefix(kNullType); return Base::WriteNull(); }
bool Bool(bool b) { PrettyPrefix(b ? kTrueType : kFalseType); return Base::WriteBool(b); }
bool Int(int i) { PrettyPrefix(kNumberType); return Base::WriteInt(i); }
bool Uint(unsigned u) { PrettyPrefix(kNumberType); return Base::WriteUint(u); }
bool Int64(int64_t i64) { PrettyPrefix(kNumberType); return Base::WriteInt64(i64); }
bool Uint64(uint64_t u64) { PrettyPrefix(kNumberType); return Base::WriteUint64(u64); }
bool Double(double d) { PrettyPrefix(kNumberType); return Base::WriteDouble(d); }
bool RawNumber(const Ch* str, SizeType length, bool copy = false) {
RAPIDJSON_ASSERT(str != 0);
(void)copy;
PrettyPrefix(kNumberType);
return Base::WriteString(str, length);
}
bool String(const Ch* str, SizeType length, bool copy = false) {
RAPIDJSON_ASSERT(str != 0);
(void)copy;
PrettyPrefix(kStringType);
return Base::WriteString(str, length);
}
#if RAPIDJSON_HAS_STDSTRING
bool String(const std::basic_string<Ch>& str) {
return String(str.data(), SizeType(str.size()));
}
#endif
bool StartObject() {
PrettyPrefix(kObjectType);
new (Base::level_stack_.template Push<typename Base::Level>()) typename Base::Level(false);
return Base::WriteStartObject();
}
bool Key(const Ch* str, SizeType length, bool copy = false) { return String(str, length, copy); }
#if RAPIDJSON_HAS_STDSTRING
bool Key(const std::basic_string<Ch>& str) {
return Key(str.data(), SizeType(str.size()));
}
#endif
bool EndObject(SizeType memberCount = 0) {
(void)memberCount;
RAPIDJSON_ASSERT(Base::level_stack_.GetSize() >= sizeof(typename Base::Level));
RAPIDJSON_ASSERT(!Base::level_stack_.template Top<typename Base::Level>()->inArray);
bool empty = Base::level_stack_.template Pop<typename Base::Level>(1)->valueCount == 0;
if (!empty) {
Base::os_->Put('\n');
WriteIndent();
}
bool ret = Base::WriteEndObject();
(void)ret;
RAPIDJSON_ASSERT(ret == true);
if (Base::level_stack_.Empty()) // end of json text
Base::os_->Flush();
return true;
}
bool StartArray() {
PrettyPrefix(kArrayType);
new (Base::level_stack_.template Push<typename Base::Level>()) typename Base::Level(true);
return Base::WriteStartArray();
}
bool EndArray(SizeType memberCount = 0) {
(void)memberCount;
RAPIDJSON_ASSERT(Base::level_stack_.GetSize() >= sizeof(typename Base::Level));
RAPIDJSON_ASSERT(Base::level_stack_.template Top<typename Base::Level>()->inArray);
bool empty = Base::level_stack_.template Pop<typename Base::Level>(1)->valueCount == 0;
if (!empty && !(formatOptions_ & kFormatSingleLineArray)) {
Base::os_->Put('\n');
WriteIndent();
}
bool ret = Base::WriteEndArray();
(void)ret;
RAPIDJSON_ASSERT(ret == true);
if (Base::level_stack_.Empty()) // end of json text
Base::os_->Flush();
return true;
}
//@}
/*! @name Convenience extensions */
//@{
//! Simpler but slower overload.
bool String(const Ch* str) { return String(str, internal::StrLen(str)); }
bool Key(const Ch* str) { return Key(str, internal::StrLen(str)); }
//@}
//! Write a raw JSON value.
/*!
For user to write a stringified JSON as a value.
\param json A well-formed JSON value. It should not contain null character within [0, length - 1] range.
\param length Length of the json.
\param type Type of the root of json.
\note When using PrettyWriter::RawValue(), the result json may not be indented correctly.
*/
bool RawValue(const Ch* json, size_t length, Type type) {
RAPIDJSON_ASSERT(json != 0);
PrettyPrefix(type);
return Base::WriteRawValue(json, length);
}
protected:
void PrettyPrefix(Type type) {
(void)type;
if (Base::level_stack_.GetSize() != 0) { // this value is not at root
typename Base::Level* level = Base::level_stack_.template Top<typename Base::Level>();
if (level->inArray) {
if (level->valueCount > 0) {
Base::os_->Put(','); // add comma if it is not the first element in array
if (formatOptions_ & kFormatSingleLineArray)
Base::os_->Put(' ');
}
if (!(formatOptions_ & kFormatSingleLineArray)) {
Base::os_->Put('\n');
WriteIndent();
}
}
else { // in object
if (level->valueCount > 0) {
if (level->valueCount % 2 == 0) {
Base::os_->Put(',');
Base::os_->Put('\n');
}
else {
Base::os_->Put(':');
Base::os_->Put(' ');
}
}
else
Base::os_->Put('\n');
if (level->valueCount % 2 == 0)
WriteIndent();
}
if (!level->inArray && level->valueCount % 2 == 0)
RAPIDJSON_ASSERT(type == kStringType); // if it's in object, then even number should be a name
level->valueCount++;
}
else {
RAPIDJSON_ASSERT(!Base::hasRoot_); // Should only has one and only one root.
Base::hasRoot_ = true;
}
}
void WriteIndent() {
size_t count = (Base::level_stack_.GetSize() / sizeof(typename Base::Level)) * indentCharCount_;
PutN(*Base::os_, static_cast<typename TargetEncoding::Ch>(indentChar_), count);
}
Ch indentChar_;
unsigned indentCharCount_;
PrettyFormatOptions formatOptions_;
private:
// Prohibit copy constructor & assignment operator.
PrettyWriter(const PrettyWriter&);
PrettyWriter& operator=(const PrettyWriter&);
};
RAPIDJSON_NAMESPACE_END
#if defined(__clang__)
RAPIDJSON_DIAG_POP
#endif
#ifdef __GNUC__
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_RAPIDJSON_H_

615
rapidjson/rapidjson.h Normal file
View File

@ -0,0 +1,615 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_RAPIDJSON_H_
#define RAPIDJSON_RAPIDJSON_H_
/*!\file rapidjson.h
\brief common definitions and configuration
\see RAPIDJSON_CONFIG
*/
/*! \defgroup RAPIDJSON_CONFIG RapidJSON configuration
\brief Configuration macros for library features
Some RapidJSON features are configurable to adapt the library to a wide
variety of platforms, environments and usage scenarios. Most of the
features can be configured in terms of overriden or predefined
preprocessor macros at compile-time.
Some additional customization is available in the \ref RAPIDJSON_ERRORS APIs.
\note These macros should be given on the compiler command-line
(where applicable) to avoid inconsistent values when compiling
different translation units of a single application.
*/
#include <cstdlib> // malloc(), realloc(), free(), size_t
#include <cstring> // memset(), memcpy(), memmove(), memcmp()
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_VERSION_STRING
//
// ALWAYS synchronize the following 3 macros with corresponding variables in /CMakeLists.txt.
//
//!@cond RAPIDJSON_HIDDEN_FROM_DOXYGEN
// token stringification
#define RAPIDJSON_STRINGIFY(x) RAPIDJSON_DO_STRINGIFY(x)
#define RAPIDJSON_DO_STRINGIFY(x) #x
//!@endcond
/*! \def RAPIDJSON_MAJOR_VERSION
\ingroup RAPIDJSON_CONFIG
\brief Major version of RapidJSON in integer.
*/
/*! \def RAPIDJSON_MINOR_VERSION
\ingroup RAPIDJSON_CONFIG
\brief Minor version of RapidJSON in integer.
*/
/*! \def RAPIDJSON_PATCH_VERSION
\ingroup RAPIDJSON_CONFIG
\brief Patch version of RapidJSON in integer.
*/
/*! \def RAPIDJSON_VERSION_STRING
\ingroup RAPIDJSON_CONFIG
\brief Version of RapidJSON in "<major>.<minor>.<patch>" string format.
*/
#define RAPIDJSON_MAJOR_VERSION 1
#define RAPIDJSON_MINOR_VERSION 1
#define RAPIDJSON_PATCH_VERSION 0
#define RAPIDJSON_VERSION_STRING \
RAPIDJSON_STRINGIFY(RAPIDJSON_MAJOR_VERSION.RAPIDJSON_MINOR_VERSION.RAPIDJSON_PATCH_VERSION)
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_NAMESPACE_(BEGIN|END)
/*! \def RAPIDJSON_NAMESPACE
\ingroup RAPIDJSON_CONFIG
\brief provide custom rapidjson namespace
In order to avoid symbol clashes and/or "One Definition Rule" errors
between multiple inclusions of (different versions of) RapidJSON in
a single binary, users can customize the name of the main RapidJSON
namespace.
In case of a single nesting level, defining \c RAPIDJSON_NAMESPACE
to a custom name (e.g. \c MyRapidJSON) is sufficient. If multiple
levels are needed, both \ref RAPIDJSON_NAMESPACE_BEGIN and \ref
RAPIDJSON_NAMESPACE_END need to be defined as well:
\code
// in some .cpp file
#define RAPIDJSON_NAMESPACE my::rapidjson
#define RAPIDJSON_NAMESPACE_BEGIN namespace my { namespace rapidjson {
#define RAPIDJSON_NAMESPACE_END } }
#include "rapidjson/..."
\endcode
\see rapidjson
*/
/*! \def RAPIDJSON_NAMESPACE_BEGIN
\ingroup RAPIDJSON_CONFIG
\brief provide custom rapidjson namespace (opening expression)
\see RAPIDJSON_NAMESPACE
*/
/*! \def RAPIDJSON_NAMESPACE_END
\ingroup RAPIDJSON_CONFIG
\brief provide custom rapidjson namespace (closing expression)
\see RAPIDJSON_NAMESPACE
*/
#ifndef RAPIDJSON_NAMESPACE
#define RAPIDJSON_NAMESPACE rapidjson
#endif
#ifndef RAPIDJSON_NAMESPACE_BEGIN
#define RAPIDJSON_NAMESPACE_BEGIN namespace RAPIDJSON_NAMESPACE {
#endif
#ifndef RAPIDJSON_NAMESPACE_END
#define RAPIDJSON_NAMESPACE_END }
#endif
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_HAS_STDSTRING
#ifndef RAPIDJSON_HAS_STDSTRING
#ifdef RAPIDJSON_DOXYGEN_RUNNING
#define RAPIDJSON_HAS_STDSTRING 1 // force generation of documentation
#else
#define RAPIDJSON_HAS_STDSTRING 0 // no std::string support by default
#endif
/*! \def RAPIDJSON_HAS_STDSTRING
\ingroup RAPIDJSON_CONFIG
\brief Enable RapidJSON support for \c std::string
By defining this preprocessor symbol to \c 1, several convenience functions for using
\ref rapidjson::GenericValue with \c std::string are enabled, especially
for construction and comparison.
\hideinitializer
*/
#endif // !defined(RAPIDJSON_HAS_STDSTRING)
#if RAPIDJSON_HAS_STDSTRING
#include <string>
#endif // RAPIDJSON_HAS_STDSTRING
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_NO_INT64DEFINE
/*! \def RAPIDJSON_NO_INT64DEFINE
\ingroup RAPIDJSON_CONFIG
\brief Use external 64-bit integer types.
RapidJSON requires the 64-bit integer types \c int64_t and \c uint64_t types
to be available at global scope.
If users have their own definition, define RAPIDJSON_NO_INT64DEFINE to
prevent RapidJSON from defining its own types.
*/
#ifndef RAPIDJSON_NO_INT64DEFINE
//!@cond RAPIDJSON_HIDDEN_FROM_DOXYGEN
#if defined(_MSC_VER) && (_MSC_VER < 1800) // Visual Studio 2013
#include "msinttypes/stdint.h"
#include "msinttypes/inttypes.h"
#else
// Other compilers should have this.
#include <stdint.h>
#include <inttypes.h>
#endif
//!@endcond
#ifdef RAPIDJSON_DOXYGEN_RUNNING
#define RAPIDJSON_NO_INT64DEFINE
#endif
#endif // RAPIDJSON_NO_INT64TYPEDEF
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_FORCEINLINE
#ifndef RAPIDJSON_FORCEINLINE
//!@cond RAPIDJSON_HIDDEN_FROM_DOXYGEN
#if defined(_MSC_VER) && defined(NDEBUG)
#define RAPIDJSON_FORCEINLINE __forceinline
#elif defined(__GNUC__) && __GNUC__ >= 4 && defined(NDEBUG)
#define RAPIDJSON_FORCEINLINE __attribute__((always_inline))
#else
#define RAPIDJSON_FORCEINLINE
#endif
//!@endcond
#endif // RAPIDJSON_FORCEINLINE
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_ENDIAN
#define RAPIDJSON_LITTLEENDIAN 0 //!< Little endian machine
#define RAPIDJSON_BIGENDIAN 1 //!< Big endian machine
//! Endianness of the machine.
/*!
\def RAPIDJSON_ENDIAN
\ingroup RAPIDJSON_CONFIG
GCC 4.6 provided macro for detecting endianness of the target machine. But other
compilers may not have this. User can define RAPIDJSON_ENDIAN to either
\ref RAPIDJSON_LITTLEENDIAN or \ref RAPIDJSON_BIGENDIAN.
Default detection implemented with reference to
\li https://gcc.gnu.org/onlinedocs/gcc-4.6.0/cpp/Common-Predefined-Macros.html
\li http://www.boost.org/doc/libs/1_42_0/boost/detail/endian.hpp
*/
#ifndef RAPIDJSON_ENDIAN
// Detect with GCC 4.6's macro
# ifdef __BYTE_ORDER__
# if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
# define RAPIDJSON_ENDIAN RAPIDJSON_LITTLEENDIAN
# elif __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
# define RAPIDJSON_ENDIAN RAPIDJSON_BIGENDIAN
# else
# error Unknown machine endianess detected. User needs to define RAPIDJSON_ENDIAN.
# endif // __BYTE_ORDER__
// Detect with GLIBC's endian.h
# elif defined(__GLIBC__)
# include <endian.h>
# if (__BYTE_ORDER == __LITTLE_ENDIAN)
# define RAPIDJSON_ENDIAN RAPIDJSON_LITTLEENDIAN
# elif (__BYTE_ORDER == __BIG_ENDIAN)
# define RAPIDJSON_ENDIAN RAPIDJSON_BIGENDIAN
# else
# error Unknown machine endianess detected. User needs to define RAPIDJSON_ENDIAN.
# endif // __GLIBC__
// Detect with _LITTLE_ENDIAN and _BIG_ENDIAN macro
# elif defined(_LITTLE_ENDIAN) && !defined(_BIG_ENDIAN)
# define RAPIDJSON_ENDIAN RAPIDJSON_LITTLEENDIAN
# elif defined(_BIG_ENDIAN) && !defined(_LITTLE_ENDIAN)
# define RAPIDJSON_ENDIAN RAPIDJSON_BIGENDIAN
// Detect with architecture macros
# elif defined(__sparc) || defined(__sparc__) || defined(_POWER) || defined(__powerpc__) || defined(__ppc__) || defined(__hpux) || defined(__hppa) || defined(_MIPSEB) || defined(_POWER) || defined(__s390__)
# define RAPIDJSON_ENDIAN RAPIDJSON_BIGENDIAN
# elif defined(__i386__) || defined(__alpha__) || defined(__ia64) || defined(__ia64__) || defined(_M_IX86) || defined(_M_IA64) || defined(_M_ALPHA) || defined(__amd64) || defined(__amd64__) || defined(_M_AMD64) || defined(__x86_64) || defined(__x86_64__) || defined(_M_X64) || defined(__bfin__)
# define RAPIDJSON_ENDIAN RAPIDJSON_LITTLEENDIAN
# elif defined(_MSC_VER) && defined(_M_ARM)
# define RAPIDJSON_ENDIAN RAPIDJSON_LITTLEENDIAN
# elif defined(RAPIDJSON_DOXYGEN_RUNNING)
# define RAPIDJSON_ENDIAN
# else
# error Unknown machine endianess detected. User needs to define RAPIDJSON_ENDIAN.
# endif
#endif // RAPIDJSON_ENDIAN
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_64BIT
//! Whether using 64-bit architecture
#ifndef RAPIDJSON_64BIT
#if defined(__LP64__) || (defined(__x86_64__) && defined(__ILP32__)) || defined(_WIN64) || defined(__EMSCRIPTEN__)
#define RAPIDJSON_64BIT 1
#else
#define RAPIDJSON_64BIT 0
#endif
#endif // RAPIDJSON_64BIT
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_ALIGN
//! Data alignment of the machine.
/*! \ingroup RAPIDJSON_CONFIG
\param x pointer to align
Some machines require strict data alignment. Currently the default uses 4 bytes
alignment on 32-bit platforms and 8 bytes alignment for 64-bit platforms.
User can customize by defining the RAPIDJSON_ALIGN function macro.
*/
#ifndef RAPIDJSON_ALIGN
#if RAPIDJSON_64BIT == 1
#define RAPIDJSON_ALIGN(x) (((x) + static_cast<uint64_t>(7u)) & ~static_cast<uint64_t>(7u))
#else
#define RAPIDJSON_ALIGN(x) (((x) + 3u) & ~3u)
#endif
#endif
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_UINT64_C2
//! Construct a 64-bit literal by a pair of 32-bit integer.
/*!
64-bit literal with or without ULL suffix is prone to compiler warnings.
UINT64_C() is C macro which cause compilation problems.
Use this macro to define 64-bit constants by a pair of 32-bit integer.
*/
#ifndef RAPIDJSON_UINT64_C2
#define RAPIDJSON_UINT64_C2(high32, low32) ((static_cast<uint64_t>(high32) << 32) | static_cast<uint64_t>(low32))
#endif
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_48BITPOINTER_OPTIMIZATION
//! Use only lower 48-bit address for some pointers.
/*!
\ingroup RAPIDJSON_CONFIG
This optimization uses the fact that current X86-64 architecture only implement lower 48-bit virtual address.
The higher 16-bit can be used for storing other data.
\c GenericValue uses this optimization to reduce its size form 24 bytes to 16 bytes in 64-bit architecture.
*/
#ifndef RAPIDJSON_48BITPOINTER_OPTIMIZATION
#if defined(__amd64__) || defined(__amd64) || defined(__x86_64__) || defined(__x86_64) || defined(_M_X64) || defined(_M_AMD64)
#define RAPIDJSON_48BITPOINTER_OPTIMIZATION 1
#else
#define RAPIDJSON_48BITPOINTER_OPTIMIZATION 0
#endif
#endif // RAPIDJSON_48BITPOINTER_OPTIMIZATION
#if RAPIDJSON_48BITPOINTER_OPTIMIZATION == 1
#if RAPIDJSON_64BIT != 1
#error RAPIDJSON_48BITPOINTER_OPTIMIZATION can only be set to 1 when RAPIDJSON_64BIT=1
#endif
#define RAPIDJSON_SETPOINTER(type, p, x) (p = reinterpret_cast<type *>((reinterpret_cast<uintptr_t>(p) & static_cast<uintptr_t>(RAPIDJSON_UINT64_C2(0xFFFF0000, 0x00000000))) | reinterpret_cast<uintptr_t>(reinterpret_cast<const void*>(x))))
#define RAPIDJSON_GETPOINTER(type, p) (reinterpret_cast<type *>(reinterpret_cast<uintptr_t>(p) & static_cast<uintptr_t>(RAPIDJSON_UINT64_C2(0x0000FFFF, 0xFFFFFFFF))))
#else
#define RAPIDJSON_SETPOINTER(type, p, x) (p = (x))
#define RAPIDJSON_GETPOINTER(type, p) (p)
#endif
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_SSE2/RAPIDJSON_SSE42/RAPIDJSON_SIMD
/*! \def RAPIDJSON_SIMD
\ingroup RAPIDJSON_CONFIG
\brief Enable SSE2/SSE4.2 optimization.
RapidJSON supports optimized implementations for some parsing operations
based on the SSE2 or SSE4.2 SIMD extensions on modern Intel-compatible
processors.
To enable these optimizations, two different symbols can be defined;
\code
// Enable SSE2 optimization.
#define RAPIDJSON_SSE2
// Enable SSE4.2 optimization.
#define RAPIDJSON_SSE42
\endcode
\c RAPIDJSON_SSE42 takes precedence, if both are defined.
If any of these symbols is defined, RapidJSON defines the macro
\c RAPIDJSON_SIMD to indicate the availability of the optimized code.
*/
#if defined(RAPIDJSON_SSE2) || defined(RAPIDJSON_SSE42) \
|| defined(RAPIDJSON_DOXYGEN_RUNNING)
#define RAPIDJSON_SIMD
#endif
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_NO_SIZETYPEDEFINE
#ifndef RAPIDJSON_NO_SIZETYPEDEFINE
/*! \def RAPIDJSON_NO_SIZETYPEDEFINE
\ingroup RAPIDJSON_CONFIG
\brief User-provided \c SizeType definition.
In order to avoid using 32-bit size types for indexing strings and arrays,
define this preprocessor symbol and provide the type rapidjson::SizeType
before including RapidJSON:
\code
#define RAPIDJSON_NO_SIZETYPEDEFINE
namespace rapidjson { typedef ::std::size_t SizeType; }
#include "rapidjson/..."
\endcode
\see rapidjson::SizeType
*/
#ifdef RAPIDJSON_DOXYGEN_RUNNING
#define RAPIDJSON_NO_SIZETYPEDEFINE
#endif
RAPIDJSON_NAMESPACE_BEGIN
//! Size type (for string lengths, array sizes, etc.)
/*! RapidJSON uses 32-bit array/string indices even on 64-bit platforms,
instead of using \c size_t. Users may override the SizeType by defining
\ref RAPIDJSON_NO_SIZETYPEDEFINE.
*/
typedef unsigned SizeType;
RAPIDJSON_NAMESPACE_END
#endif
// always import std::size_t to rapidjson namespace
RAPIDJSON_NAMESPACE_BEGIN
using std::size_t;
RAPIDJSON_NAMESPACE_END
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_ASSERT
//! Assertion.
/*! \ingroup RAPIDJSON_CONFIG
By default, rapidjson uses C \c assert() for internal assertions.
User can override it by defining RAPIDJSON_ASSERT(x) macro.
\note Parsing errors are handled and can be customized by the
\ref RAPIDJSON_ERRORS APIs.
*/
#ifndef RAPIDJSON_ASSERT
#include <cassert>
#define RAPIDJSON_ASSERT(x) assert(x)
#endif // RAPIDJSON_ASSERT
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_STATIC_ASSERT
// Adopt from boost
#ifndef RAPIDJSON_STATIC_ASSERT
#ifndef __clang__
//!@cond RAPIDJSON_HIDDEN_FROM_DOXYGEN
#endif
RAPIDJSON_NAMESPACE_BEGIN
template <bool x> struct STATIC_ASSERTION_FAILURE;
template <> struct STATIC_ASSERTION_FAILURE<true> { enum { value = 1 }; };
template<int x> struct StaticAssertTest {};
RAPIDJSON_NAMESPACE_END
#define RAPIDJSON_JOIN(X, Y) RAPIDJSON_DO_JOIN(X, Y)
#define RAPIDJSON_DO_JOIN(X, Y) RAPIDJSON_DO_JOIN2(X, Y)
#define RAPIDJSON_DO_JOIN2(X, Y) X##Y
#if defined(__GNUC__)
#define RAPIDJSON_STATIC_ASSERT_UNUSED_ATTRIBUTE __attribute__((unused))
#else
#define RAPIDJSON_STATIC_ASSERT_UNUSED_ATTRIBUTE
#endif
#ifndef __clang__
//!@endcond
#endif
/*! \def RAPIDJSON_STATIC_ASSERT
\brief (Internal) macro to check for conditions at compile-time
\param x compile-time condition
\hideinitializer
*/
#define RAPIDJSON_STATIC_ASSERT(x) \
typedef ::RAPIDJSON_NAMESPACE::StaticAssertTest< \
sizeof(::RAPIDJSON_NAMESPACE::STATIC_ASSERTION_FAILURE<bool(x) >)> \
RAPIDJSON_JOIN(StaticAssertTypedef, __LINE__) RAPIDJSON_STATIC_ASSERT_UNUSED_ATTRIBUTE
#endif
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_LIKELY, RAPIDJSON_UNLIKELY
//! Compiler branching hint for expression with high probability to be true.
/*!
\ingroup RAPIDJSON_CONFIG
\param x Boolean expression likely to be true.
*/
#ifndef RAPIDJSON_LIKELY
#if defined(__GNUC__) || defined(__clang__)
#define RAPIDJSON_LIKELY(x) __builtin_expect(!!(x), 1)
#else
#define RAPIDJSON_LIKELY(x) (x)
#endif
#endif
//! Compiler branching hint for expression with low probability to be true.
/*!
\ingroup RAPIDJSON_CONFIG
\param x Boolean expression unlikely to be true.
*/
#ifndef RAPIDJSON_UNLIKELY
#if defined(__GNUC__) || defined(__clang__)
#define RAPIDJSON_UNLIKELY(x) __builtin_expect(!!(x), 0)
#else
#define RAPIDJSON_UNLIKELY(x) (x)
#endif
#endif
///////////////////////////////////////////////////////////////////////////////
// Helpers
//!@cond RAPIDJSON_HIDDEN_FROM_DOXYGEN
#define RAPIDJSON_MULTILINEMACRO_BEGIN do {
#define RAPIDJSON_MULTILINEMACRO_END \
} while((void)0, 0)
// adopted from Boost
#define RAPIDJSON_VERSION_CODE(x,y,z) \
(((x)*100000) + ((y)*100) + (z))
///////////////////////////////////////////////////////////////////////////////
// RAPIDJSON_DIAG_PUSH/POP, RAPIDJSON_DIAG_OFF
#if defined(__GNUC__)
#define RAPIDJSON_GNUC \
RAPIDJSON_VERSION_CODE(__GNUC__,__GNUC_MINOR__,__GNUC_PATCHLEVEL__)
#endif
#if defined(__clang__) || (defined(RAPIDJSON_GNUC) && RAPIDJSON_GNUC >= RAPIDJSON_VERSION_CODE(4,2,0))
#define RAPIDJSON_PRAGMA(x) _Pragma(RAPIDJSON_STRINGIFY(x))
#define RAPIDJSON_DIAG_PRAGMA(x) RAPIDJSON_PRAGMA(GCC diagnostic x)
#define RAPIDJSON_DIAG_OFF(x) \
RAPIDJSON_DIAG_PRAGMA(ignored RAPIDJSON_STRINGIFY(RAPIDJSON_JOIN(-W,x)))
// push/pop support in Clang and GCC>=4.6
#if defined(__clang__) || (defined(RAPIDJSON_GNUC) && RAPIDJSON_GNUC >= RAPIDJSON_VERSION_CODE(4,6,0))
#define RAPIDJSON_DIAG_PUSH RAPIDJSON_DIAG_PRAGMA(push)
#define RAPIDJSON_DIAG_POP RAPIDJSON_DIAG_PRAGMA(pop)
#else // GCC >= 4.2, < 4.6
#define RAPIDJSON_DIAG_PUSH /* ignored */
#define RAPIDJSON_DIAG_POP /* ignored */
#endif
#elif defined(_MSC_VER)
// pragma (MSVC specific)
#define RAPIDJSON_PRAGMA(x) __pragma(x)
#define RAPIDJSON_DIAG_PRAGMA(x) RAPIDJSON_PRAGMA(warning(x))
#define RAPIDJSON_DIAG_OFF(x) RAPIDJSON_DIAG_PRAGMA(disable: x)
#define RAPIDJSON_DIAG_PUSH RAPIDJSON_DIAG_PRAGMA(push)
#define RAPIDJSON_DIAG_POP RAPIDJSON_DIAG_PRAGMA(pop)
#else
#define RAPIDJSON_DIAG_OFF(x) /* ignored */
#define RAPIDJSON_DIAG_PUSH /* ignored */
#define RAPIDJSON_DIAG_POP /* ignored */
#endif // RAPIDJSON_DIAG_*
///////////////////////////////////////////////////////////////////////////////
// C++11 features
#ifndef RAPIDJSON_HAS_CXX11_RVALUE_REFS
#if defined(__clang__)
#if __has_feature(cxx_rvalue_references) && \
(defined(_LIBCPP_VERSION) || defined(__GLIBCXX__) && __GLIBCXX__ >= 20080306)
#define RAPIDJSON_HAS_CXX11_RVALUE_REFS 1
#else
#define RAPIDJSON_HAS_CXX11_RVALUE_REFS 0
#endif
#elif (defined(RAPIDJSON_GNUC) && (RAPIDJSON_GNUC >= RAPIDJSON_VERSION_CODE(4,3,0)) && defined(__GXX_EXPERIMENTAL_CXX0X__)) || \
(defined(_MSC_VER) && _MSC_VER >= 1600)
#define RAPIDJSON_HAS_CXX11_RVALUE_REFS 1
#else
#define RAPIDJSON_HAS_CXX11_RVALUE_REFS 0
#endif
#endif // RAPIDJSON_HAS_CXX11_RVALUE_REFS
#ifndef RAPIDJSON_HAS_CXX11_NOEXCEPT
#if defined(__clang__)
#define RAPIDJSON_HAS_CXX11_NOEXCEPT __has_feature(cxx_noexcept)
#elif (defined(RAPIDJSON_GNUC) && (RAPIDJSON_GNUC >= RAPIDJSON_VERSION_CODE(4,6,0)) && defined(__GXX_EXPERIMENTAL_CXX0X__))
// (defined(_MSC_VER) && _MSC_VER >= ????) // not yet supported
#define RAPIDJSON_HAS_CXX11_NOEXCEPT 1
#else
#define RAPIDJSON_HAS_CXX11_NOEXCEPT 0
#endif
#endif
#if RAPIDJSON_HAS_CXX11_NOEXCEPT
#define RAPIDJSON_NOEXCEPT noexcept
#else
#define RAPIDJSON_NOEXCEPT /* noexcept */
#endif // RAPIDJSON_HAS_CXX11_NOEXCEPT
// no automatic detection, yet
#ifndef RAPIDJSON_HAS_CXX11_TYPETRAITS
#define RAPIDJSON_HAS_CXX11_TYPETRAITS 0
#endif
#ifndef RAPIDJSON_HAS_CXX11_RANGE_FOR
#if defined(__clang__)
#define RAPIDJSON_HAS_CXX11_RANGE_FOR __has_feature(cxx_range_for)
#elif (defined(RAPIDJSON_GNUC) && (RAPIDJSON_GNUC >= RAPIDJSON_VERSION_CODE(4,3,0)) && defined(__GXX_EXPERIMENTAL_CXX0X__)) || \
(defined(_MSC_VER) && _MSC_VER >= 1700)
#define RAPIDJSON_HAS_CXX11_RANGE_FOR 1
#else
#define RAPIDJSON_HAS_CXX11_RANGE_FOR 0
#endif
#endif // RAPIDJSON_HAS_CXX11_RANGE_FOR
//!@endcond
///////////////////////////////////////////////////////////////////////////////
// new/delete
#ifndef RAPIDJSON_NEW
///! customization point for global \c new
#define RAPIDJSON_NEW(x) new x
#endif
#ifndef RAPIDJSON_DELETE
///! customization point for global \c delete
#define RAPIDJSON_DELETE(x) delete x
#endif
///////////////////////////////////////////////////////////////////////////////
// Type
/*! \namespace rapidjson
\brief main RapidJSON namespace
\see RAPIDJSON_NAMESPACE
*/
RAPIDJSON_NAMESPACE_BEGIN
//! Type of JSON value
enum Type {
kNullType = 0, //!< null
kFalseType = 1, //!< false
kTrueType = 2, //!< true
kObjectType = 3, //!< object
kArrayType = 4, //!< array
kStringType = 5, //!< string
kNumberType = 6 //!< number
};
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_RAPIDJSON_H_

1862
rapidjson/reader.h Normal file
View File

File diff suppressed because it is too large Load Diff

2024
rapidjson/schema.h Normal file
View File

File diff suppressed because it is too large Load Diff

179
rapidjson/stream.h Normal file
View File

@ -0,0 +1,179 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#include "rapidjson.h"
#ifndef RAPIDJSON_STREAM_H_
#define RAPIDJSON_STREAM_H_
#include "encodings.h"
RAPIDJSON_NAMESPACE_BEGIN
///////////////////////////////////////////////////////////////////////////////
// Stream
/*! \class rapidjson::Stream
\brief Concept for reading and writing characters.
For read-only stream, no need to implement PutBegin(), Put(), Flush() and PutEnd().
For write-only stream, only need to implement Put() and Flush().
\code
concept Stream {
typename Ch; //!< Character type of the stream.
//! Read the current character from stream without moving the read cursor.
Ch Peek() const;
//! Read the current character from stream and moving the read cursor to next character.
Ch Take();
//! Get the current read cursor.
//! \return Number of characters read from start.
size_t Tell();
//! Begin writing operation at the current read pointer.
//! \return The begin writer pointer.
Ch* PutBegin();
//! Write a character.
void Put(Ch c);
//! Flush the buffer.
void Flush();
//! End the writing operation.
//! \param begin The begin write pointer returned by PutBegin().
//! \return Number of characters written.
size_t PutEnd(Ch* begin);
}
\endcode
*/
//! Provides additional information for stream.
/*!
By using traits pattern, this type provides a default configuration for stream.
For custom stream, this type can be specialized for other configuration.
See TEST(Reader, CustomStringStream) in readertest.cpp for example.
*/
template<typename Stream>
struct StreamTraits {
//! Whether to make local copy of stream for optimization during parsing.
/*!
By default, for safety, streams do not use local copy optimization.
Stream that can be copied fast should specialize this, like StreamTraits<StringStream>.
*/
enum { copyOptimization = 0 };
};
//! Reserve n characters for writing to a stream.
template<typename Stream>
inline void PutReserve(Stream& stream, size_t count) {
(void)stream;
(void)count;
}
//! Write character to a stream, presuming buffer is reserved.
template<typename Stream>
inline void PutUnsafe(Stream& stream, typename Stream::Ch c) {
stream.Put(c);
}
//! Put N copies of a character to a stream.
template<typename Stream, typename Ch>
inline void PutN(Stream& stream, Ch c, size_t n) {
PutReserve(stream, n);
for (size_t i = 0; i < n; i++)
PutUnsafe(stream, c);
}
///////////////////////////////////////////////////////////////////////////////
// StringStream
//! Read-only string stream.
/*! \note implements Stream concept
*/
template <typename Encoding>
struct GenericStringStream {
typedef typename Encoding::Ch Ch;
GenericStringStream(const Ch *src) : src_(src), head_(src) {}
Ch Peek() const { return *src_; }
Ch Take() { return *src_++; }
size_t Tell() const { return static_cast<size_t>(src_ - head_); }
Ch* PutBegin() { RAPIDJSON_ASSERT(false); return 0; }
void Put(Ch) { RAPIDJSON_ASSERT(false); }
void Flush() { RAPIDJSON_ASSERT(false); }
size_t PutEnd(Ch*) { RAPIDJSON_ASSERT(false); return 0; }
const Ch* src_; //!< Current read position.
const Ch* head_; //!< Original head of the string.
};
template <typename Encoding>
struct StreamTraits<GenericStringStream<Encoding> > {
enum { copyOptimization = 1 };
};
//! String stream with UTF8 encoding.
typedef GenericStringStream<UTF8<> > StringStream;
///////////////////////////////////////////////////////////////////////////////
// InsituStringStream
//! A read-write string stream.
/*! This string stream is particularly designed for in-situ parsing.
\note implements Stream concept
*/
template <typename Encoding>
struct GenericInsituStringStream {
typedef typename Encoding::Ch Ch;
GenericInsituStringStream(Ch *src) : src_(src), dst_(0), head_(src) {}
// Read
Ch Peek() { return *src_; }
Ch Take() { return *src_++; }
size_t Tell() { return static_cast<size_t>(src_ - head_); }
// Write
void Put(Ch c) { RAPIDJSON_ASSERT(dst_ != 0); *dst_++ = c; }
Ch* PutBegin() { return dst_ = src_; }
size_t PutEnd(Ch* begin) { return static_cast<size_t>(dst_ - begin); }
void Flush() {}
Ch* Push(size_t count) { Ch* begin = dst_; dst_ += count; return begin; }
void Pop(size_t count) { dst_ -= count; }
Ch* src_;
Ch* dst_;
Ch* head_;
};
template <typename Encoding>
struct StreamTraits<GenericInsituStringStream<Encoding> > {
enum { copyOptimization = 1 };
};
//! Insitu string stream with UTF8 encoding.
typedef GenericInsituStringStream<UTF8<> > InsituStringStream;
RAPIDJSON_NAMESPACE_END
#endif // RAPIDJSON_STREAM_H_

121
rapidjson/stringbuffer.h Normal file
View File

@ -0,0 +1,121 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_STRINGBUFFER_H_
#define RAPIDJSON_STRINGBUFFER_H_
#include "stream.h"
#include "internal/stack.h"
#if RAPIDJSON_HAS_CXX11_RVALUE_REFS
#include <utility> // std::move
#endif
#include "internal/stack.h"
#if defined(__clang__)
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(c++98-compat)
#endif
RAPIDJSON_NAMESPACE_BEGIN
//! Represents an in-memory output stream.
/*!
\tparam Encoding Encoding of the stream.
\tparam Allocator type for allocating memory buffer.
\note implements Stream concept
*/
template <typename Encoding, typename Allocator = CrtAllocator>
class GenericStringBuffer {
public:
typedef typename Encoding::Ch Ch;
GenericStringBuffer(Allocator* allocator = 0, size_t capacity = kDefaultCapacity) : stack_(allocator, capacity) {}
#if RAPIDJSON_HAS_CXX11_RVALUE_REFS
GenericStringBuffer(GenericStringBuffer&& rhs) : stack_(std::move(rhs.stack_)) {}
GenericStringBuffer& operator=(GenericStringBuffer&& rhs) {
if (&rhs != this)
stack_ = std::move(rhs.stack_);
return *this;
}
#endif
void Put(Ch c) { *stack_.template Push<Ch>() = c; }
void PutUnsafe(Ch c) { *stack_.template PushUnsafe<Ch>() = c; }
void Flush() {}
void Clear() { stack_.Clear(); }
void ShrinkToFit() {
// Push and pop a null terminator. This is safe.
*stack_.template Push<Ch>() = '\0';
stack_.ShrinkToFit();
stack_.template Pop<Ch>(1);
}
void Reserve(size_t count) { stack_.template Reserve<Ch>(count); }
Ch* Push(size_t count) { return stack_.template Push<Ch>(count); }
Ch* PushUnsafe(size_t count) { return stack_.template PushUnsafe<Ch>(count); }
void Pop(size_t count) { stack_.template Pop<Ch>(count); }
const Ch* GetString() const {
// Push and pop a null terminator. This is safe.
*stack_.template Push<Ch>() = '\0';
stack_.template Pop<Ch>(1);
return stack_.template Bottom<Ch>();
}
//! Get the size of string in bytes in the string buffer.
size_t GetSize() const { return stack_.GetSize(); }
//! Get the length of string in Ch in the string buffer.
size_t GetLength() const { return stack_.GetSize() / sizeof(Ch); }
static const size_t kDefaultCapacity = 256;
mutable internal::Stack<Allocator> stack_;
private:
// Prohibit copy constructor & assignment operator.
GenericStringBuffer(const GenericStringBuffer&);
GenericStringBuffer& operator=(const GenericStringBuffer&);
};
//! String buffer with UTF8 encoding
typedef GenericStringBuffer<UTF8<> > StringBuffer;
template<typename Encoding, typename Allocator>
inline void PutReserve(GenericStringBuffer<Encoding, Allocator>& stream, size_t count) {
stream.Reserve(count);
}
template<typename Encoding, typename Allocator>
inline void PutUnsafe(GenericStringBuffer<Encoding, Allocator>& stream, typename Encoding::Ch c) {
stream.PutUnsafe(c);
}
//! Implement specialized version of PutN() with memset() for better performance.
template<>
inline void PutN(GenericStringBuffer<UTF8<> >& stream, char c, size_t n) {
std::memset(stream.stack_.Push<char>(n), c, n * sizeof(c));
}
RAPIDJSON_NAMESPACE_END
#if defined(__clang__)
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_STRINGBUFFER_H_

624
rapidjson/writer.h Normal file
View File

@ -0,0 +1,624 @@
// Tencent is pleased to support the open source community by making RapidJSON available.
//
// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. All rights reserved.
//
// Licensed under the MIT License (the "License"); you may not use this file except
// in compliance with the License. You may obtain a copy of the License at
//
// http://opensource.org/licenses/MIT
//
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#ifndef RAPIDJSON_WRITER_H_
#define RAPIDJSON_WRITER_H_
#include "stream.h"
#include "internal/stack.h"
#include "internal/strfunc.h"
#include "internal/dtoa.h"
#include "internal/itoa.h"
#include "stringbuffer.h"
#include <new> // placement new
#if defined(RAPIDJSON_SIMD) && defined(_MSC_VER)
#include <intrin.h>
#pragma intrinsic(_BitScanForward)
#endif
#ifdef RAPIDJSON_SSE42
#include <nmmintrin.h>
#elif defined(RAPIDJSON_SSE2)
#include <emmintrin.h>
#endif
#ifdef _MSC_VER
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(4127) // conditional expression is constant
#endif
#ifdef __clang__
RAPIDJSON_DIAG_PUSH
RAPIDJSON_DIAG_OFF(padded)
RAPIDJSON_DIAG_OFF(unreachable-code)
RAPIDJSON_DIAG_OFF(c++98-compat)
#endif
RAPIDJSON_NAMESPACE_BEGIN
///////////////////////////////////////////////////////////////////////////////
// WriteFlag
/*! \def RAPIDJSON_WRITE_DEFAULT_FLAGS
\ingroup RAPIDJSON_CONFIG
\brief User-defined kWriteDefaultFlags definition.
User can define this as any \c WriteFlag combinations.
*/
#ifndef RAPIDJSON_WRITE_DEFAULT_FLAGS
#define RAPIDJSON_WRITE_DEFAULT_FLAGS kWriteNoFlags
#endif
//! Combination of writeFlags
enum WriteFlag {
kWriteNoFlags = 0, //!< No flags are set.
kWriteValidateEncodingFlag = 1, //!< Validate encoding of JSON strings.
kWriteNanAndInfFlag = 2, //!< Allow writing of Infinity, -Infinity and NaN.
kWriteDefaultFlags = RAPIDJSON_WRITE_DEFAULT_FLAGS //!< Default write flags. Can be customized by defining RAPIDJSON_WRITE_DEFAULT_FLAGS
};
//! JSON writer
/*! Writer implements the concept Handler.
It generates JSON text by events to an output os.
User may programmatically calls the functions of a writer to generate JSON text.
On the other side, a writer can also be passed to objects that generates events,
for example Reader::Parse() and Document::Accept().
\tparam OutputStream Type of output stream.
\tparam SourceEncoding Encoding of source string.
\tparam TargetEncoding Encoding of output stream.
\tparam StackAllocator Type of allocator for allocating memory of stack.
\note implements Handler concept
*/
template<typename OutputStream, typename SourceEncoding = UTF8<>, typename TargetEncoding = UTF8<>, typename StackAllocator = CrtAllocator, unsigned writeFlags = kWriteDefaultFlags>
class Writer {
public:
typedef typename SourceEncoding::Ch Ch;
static const int kDefaultMaxDecimalPlaces = 324;
//! Constructor
/*! \param os Output stream.
\param stackAllocator User supplied allocator. If it is null, it will create a private one.
\param levelDepth Initial capacity of stack.
*/
explicit
Writer(OutputStream& os, StackAllocator* stackAllocator = 0, size_t levelDepth = kDefaultLevelDepth) :
os_(&os), level_stack_(stackAllocator, levelDepth * sizeof(Level)), maxDecimalPlaces_(kDefaultMaxDecimalPlaces), hasRoot_(false) {}
explicit
Writer(StackAllocator* allocator = 0, size_t levelDepth = kDefaultLevelDepth) :
os_(0), level_stack_(allocator, levelDepth * sizeof(Level)), maxDecimalPlaces_(kDefaultMaxDecimalPlaces), hasRoot_(false) {}
#if RAPIDJSON_HAS_CXX11_RVALUE_REFS
Writer(Writer&& rhs) :
os_(rhs.os_), level_stack_(std::move(rhs.level_stack_)), maxDecimalPlaces_(rhs.maxDecimalPlaces_), hasRoot_(rhs.hasRoot_) {
rhs.os_ = 0;
}
#endif
//! Reset the writer with a new stream.
/*!
This function reset the writer with a new stream and default settings,
in order to make a Writer object reusable for output multiple JSONs.
\param os New output stream.
\code
Writer<OutputStream> writer(os1);
writer.StartObject();
// ...
writer.EndObject();
writer.Reset(os2);
writer.StartObject();
// ...
writer.EndObject();
\endcode
*/
void Reset(OutputStream& os) {
os_ = &os;
hasRoot_ = false;
level_stack_.Clear();
}
//! Checks whether the output is a complete JSON.
/*!
A complete JSON has a complete root object or array.
*/
bool IsComplete() const {
return hasRoot_ && level_stack_.Empty();
}
int GetMaxDecimalPlaces() const {
return maxDecimalPlaces_;
}
//! Sets the maximum number of decimal places for double output.
/*!
This setting truncates the output with specified number of decimal places.
For example,
\code
writer.SetMaxDecimalPlaces(3);
writer.StartArray();
writer.Double(0.12345); // "0.123"
writer.Double(0.0001); // "0.0"
writer.Double(1.234567890123456e30); // "1.234567890123456e30" (do not truncate significand for positive exponent)
writer.Double(1.23e-4); // "0.0" (do truncate significand for negative exponent)
writer.EndArray();
\endcode
The default setting does not truncate any decimal places. You can restore to this setting by calling
\code
writer.SetMaxDecimalPlaces(Writer::kDefaultMaxDecimalPlaces);
\endcode
*/
void SetMaxDecimalPlaces(int maxDecimalPlaces) {
maxDecimalPlaces_ = maxDecimalPlaces;
}
/*!@name Implementation of Handler
\see Handler
*/
//@{
bool Null() { Prefix(kNullType); return EndValue(WriteNull()); }
bool Bool(bool b) { Prefix(b ? kTrueType : kFalseType); return EndValue(WriteBool(b)); }
bool Int(int i) { Prefix(kNumberType); return EndValue(WriteInt(i)); }
bool Uint(unsigned u) { Prefix(kNumberType); return EndValue(WriteUint(u)); }
bool Int64(int64_t i64) { Prefix(kNumberType); return EndValue(WriteInt64(i64)); }
bool Uint64(uint64_t u64) { Prefix(kNumberType); return EndValue(WriteUint64(u64)); }
//! Writes the given \c double value to the stream
/*!
\param d The value to be written.
\return Whether it is succeed.
*/
bool Double(double d) { Prefix(kNumberType); return EndValue(WriteDouble(d)); }
bool RawNumber(const Ch* str, SizeType length, bool copy = false) {
RAPIDJSON_ASSERT(str != 0);
(void)copy;
Prefix(kNumberType);
return EndValue(WriteString(str, length));
}
bool String(const Ch* str, SizeType length, bool copy = false) {
RAPIDJSON_ASSERT(str != 0);
(void)copy;
Prefix(kStringType);
return EndValue(WriteString(str, length));
}
#if RAPIDJSON_HAS_STDSTRING
bool String(const std::basic_string<Ch>& str) {
return String(str.data(), SizeType(str.size()));
}
#endif
bool StartObject() {
Prefix(kObjectType);
new (level_stack_.template Push<Level>()) Level(false);
return WriteStartObject();
}
bool Key(const Ch* str, SizeType length, bool copy = false) { return String(str, length, copy); }
bool EndObject(SizeType memberCount = 0) {
(void)memberCount;
RAPIDJSON_ASSERT(level_stack_.GetSize() >= sizeof(Level));
RAPIDJSON_ASSERT(!level_stack_.template Top<Level>()->inArray);
level_stack_.template Pop<Level>(1);
return EndValue(WriteEndObject());
}
bool StartArray() {
Prefix(kArrayType);
new (level_stack_.template Push<Level>()) Level(true);
return WriteStartArray();
}
bool EndArray(SizeType elementCount = 0) {
(void)elementCount;
RAPIDJSON_ASSERT(level_stack_.GetSize() >= sizeof(Level));
RAPIDJSON_ASSERT(level_stack_.template Top<Level>()->inArray);
level_stack_.template Pop<Level>(1);
return EndValue(WriteEndArray());
}
//@}
/*! @name Convenience extensions */
//@{
//! Simpler but slower overload.
bool String(const Ch* str) { return String(str, internal::StrLen(str)); }
bool Key(const Ch* str) { return Key(str, internal::StrLen(str)); }
//@}
//! Write a raw JSON value.
/*!
For user to write a stringified JSON as a value.
\param json A well-formed JSON value. It should not contain null character within [0, length - 1] range.
\param length Length of the json.
\param type Type of the root of json.
*/
bool RawValue(const Ch* json, size_t length, Type type) {
RAPIDJSON_ASSERT(json != 0);
Prefix(type);
return EndValue(WriteRawValue(json, length));
}
protected:
//! Information for each nested level
struct Level {
Level(bool inArray_) : valueCount(0), inArray(inArray_) {}
size_t valueCount; //!< number of values in this level
bool inArray; //!< true if in array, otherwise in object
};
static const size_t kDefaultLevelDepth = 32;
bool WriteNull() {
PutReserve(*os_, 4);
PutUnsafe(*os_, 'n'); PutUnsafe(*os_, 'u'); PutUnsafe(*os_, 'l'); PutUnsafe(*os_, 'l'); return true;
}
bool WriteBool(bool b) {
if (b) {
PutReserve(*os_, 4);
PutUnsafe(*os_, 't'); PutUnsafe(*os_, 'r'); PutUnsafe(*os_, 'u'); PutUnsafe(*os_, 'e');
}
else {
PutReserve(*os_, 5);
PutUnsafe(*os_, 'f'); PutUnsafe(*os_, 'a'); PutUnsafe(*os_, 'l'); PutUnsafe(*os_, 's'); PutUnsafe(*os_, 'e');
}
return true;
}
bool WriteInt(int i) {
char buffer[11];
const char* end = internal::i32toa(i, buffer);
PutReserve(*os_, static_cast<size_t>(end - buffer));
for (const char* p = buffer; p != end; ++p)
PutUnsafe(*os_, static_cast<typename TargetEncoding::Ch>(*p));
return true;
}
bool WriteUint(unsigned u) {
char buffer[10];
const char* end = internal::u32toa(u, buffer);
PutReserve(*os_, static_cast<size_t>(end - buffer));
for (const char* p = buffer; p != end; ++p)
PutUnsafe(*os_, static_cast<typename TargetEncoding::Ch>(*p));
return true;
}
bool WriteInt64(int64_t i64) {
char buffer[21];
const char* end = internal::i64toa(i64, buffer);
PutReserve(*os_, static_cast<size_t>(end - buffer));
for (const char* p = buffer; p != end; ++p)
PutUnsafe(*os_, static_cast<typename TargetEncoding::Ch>(*p));
return true;
}
bool WriteUint64(uint64_t u64) {
char buffer[20];
char* end = internal::u64toa(u64, buffer);
PutReserve(*os_, static_cast<size_t>(end - buffer));
for (char* p = buffer; p != end; ++p)
PutUnsafe(*os_, static_cast<typename TargetEncoding::Ch>(*p));
return true;
}
bool WriteDouble(double d) {
if (internal::Double(d).IsNanOrInf()) {
if (!(writeFlags & kWriteNanAndInfFlag))
return false;
if (internal::Double(d).IsNan()) {
PutReserve(*os_, 3);
PutUnsafe(*os_, 'N'); PutUnsafe(*os_, 'a'); PutUnsafe(*os_, 'N');
return true;
}
if (internal::Double(d).Sign()) {
PutReserve(*os_, 9);
PutUnsafe(*os_, '-');
}
else
PutReserve(*os_, 8);
PutUnsafe(*os_, 'I'); PutUnsafe(*os_, 'n'); PutUnsafe(*os_, 'f');
PutUnsafe(*os_, 'i'); PutUnsafe(*os_, 'n'); PutUnsafe(*os_, 'i'); PutUnsafe(*os_, 't'); PutUnsafe(*os_, 'y');
return true;
}
char buffer[25];
char* end = internal::dtoa(d, buffer, maxDecimalPlaces_);
PutReserve(*os_, static_cast<size_t>(end - buffer));
for (char* p = buffer; p != end; ++p)
PutUnsafe(*os_, static_cast<typename TargetEncoding::Ch>(*p));
return true;
}
bool WriteString(const Ch* str, SizeType length) {
static const typename TargetEncoding::Ch hexDigits[16] = { '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'A', 'B', 'C', 'D', 'E', 'F' };
static const char escape[256] = {
#define Z16 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0
//0 1 2 3 4 5 6 7 8 9 A B C D E F
'u', 'u', 'u', 'u', 'u', 'u', 'u', 'u', 'b', 't', 'n', 'u', 'f', 'r', 'u', 'u', // 00
'u', 'u', 'u', 'u', 'u', 'u', 'u', 'u', 'u', 'u', 'u', 'u', 'u', 'u', 'u', 'u', // 10
0, 0, '"', 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 20
Z16, Z16, // 30~4F
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,'\\', 0, 0, 0, // 50
Z16, Z16, Z16, Z16, Z16, Z16, Z16, Z16, Z16, Z16 // 60~FF
#undef Z16
};
if (TargetEncoding::supportUnicode)
PutReserve(*os_, 2 + length * 6); // "\uxxxx..."
else
PutReserve(*os_, 2 + length * 12); // "\uxxxx\uyyyy..."
PutUnsafe(*os_, '\"');
GenericStringStream<SourceEncoding> is(str);
while (ScanWriteUnescapedString(is, length)) {
const Ch c = is.Peek();
if (!TargetEncoding::supportUnicode && static_cast<unsigned>(c) >= 0x80) {
// Unicode escaping
unsigned codepoint;
if (RAPIDJSON_UNLIKELY(!SourceEncoding::Decode(is, &codepoint)))
return false;
PutUnsafe(*os_, '\\');
PutUnsafe(*os_, 'u');
if (codepoint <= 0xD7FF || (codepoint >= 0xE000 && codepoint <= 0xFFFF)) {
PutUnsafe(*os_, hexDigits[(codepoint >> 12) & 15]);
PutUnsafe(*os_, hexDigits[(codepoint >> 8) & 15]);
PutUnsafe(*os_, hexDigits[(codepoint >> 4) & 15]);
PutUnsafe(*os_, hexDigits[(codepoint ) & 15]);
}
else {
RAPIDJSON_ASSERT(codepoint >= 0x010000 && codepoint <= 0x10FFFF);
// Surrogate pair
unsigned s = codepoint - 0x010000;
unsigned lead = (s >> 10) + 0xD800;
unsigned trail = (s & 0x3FF) + 0xDC00;
PutUnsafe(*os_, hexDigits[(lead >> 12) & 15]);
PutUnsafe(*os_, hexDigits[(lead >> 8) & 15]);
PutUnsafe(*os_, hexDigits[(lead >> 4) & 15]);
PutUnsafe(*os_, hexDigits[(lead ) & 15]);
PutUnsafe(*os_, '\\');
PutUnsafe(*os_, 'u');
PutUnsafe(*os_, hexDigits[(trail >> 12) & 15]);
PutUnsafe(*os_, hexDigits[(trail >> 8) & 15]);
PutUnsafe(*os_, hexDigits[(trail >> 4) & 15]);
PutUnsafe(*os_, hexDigits[(trail ) & 15]);
}
}
else if ((sizeof(Ch) == 1 || static_cast<unsigned>(c) < 256) && RAPIDJSON_UNLIKELY(escape[static_cast<unsigned char>(c)])) {
is.Take();
PutUnsafe(*os_, '\\');
PutUnsafe(*os_, static_cast<typename TargetEncoding::Ch>(escape[static_cast<unsigned char>(c)]));
if (escape[static_cast<unsigned char>(c)] == 'u') {
PutUnsafe(*os_, '0');
PutUnsafe(*os_, '0');
PutUnsafe(*os_, hexDigits[static_cast<unsigned char>(c) >> 4]);
PutUnsafe(*os_, hexDigits[static_cast<unsigned char>(c) & 0xF]);
}
}
else if (RAPIDJSON_UNLIKELY(!(writeFlags & kWriteValidateEncodingFlag ?
Transcoder<SourceEncoding, TargetEncoding>::Validate(is, *os_) :
Transcoder<SourceEncoding, TargetEncoding>::TranscodeUnsafe(is, *os_))))
return false;
}
PutUnsafe(*os_, '\"');
return true;
}
bool ScanWriteUnescapedString(GenericStringStream<SourceEncoding>& is, size_t length) {
return RAPIDJSON_LIKELY(is.Tell() < length);
}
bool WriteStartObject() { os_->Put('{'); return true; }
bool WriteEndObject() { os_->Put('}'); return true; }
bool WriteStartArray() { os_->Put('['); return true; }
bool WriteEndArray() { os_->Put(']'); return true; }
bool WriteRawValue(const Ch* json, size_t length) {
PutReserve(*os_, length);
for (size_t i = 0; i < length; i++) {
RAPIDJSON_ASSERT(json[i] != '\0');
PutUnsafe(*os_, json[i]);
}
return true;
}
void Prefix(Type type) {
(void)type;
if (RAPIDJSON_LIKELY(level_stack_.GetSize() != 0)) { // this value is not at root
Level* level = level_stack_.template Top<Level>();
if (level->valueCount > 0) {
if (level->inArray)
os_->Put(','); // add comma if it is not the first element in array
else // in object
os_->Put((level->valueCount % 2 == 0) ? ',' : ':');
}
if (!level->inArray && level->valueCount % 2 == 0)
RAPIDJSON_ASSERT(type == kStringType); // if it's in object, then even number should be a name
level->valueCount++;
}
else {
RAPIDJSON_ASSERT(!hasRoot_); // Should only has one and only one root.
hasRoot_ = true;
}
}
// Flush the value if it is the top level one.
bool EndValue(bool ret) {
if (RAPIDJSON_UNLIKELY(level_stack_.Empty())) // end of json text
os_->Flush();
return ret;
}
OutputStream* os_;
internal::Stack<StackAllocator> level_stack_;
int maxDecimalPlaces_;
bool hasRoot_;
private:
// Prohibit copy constructor & assignment operator.
Writer(const Writer&);
Writer& operator=(const Writer&);
};
// Full specialization for StringStream to prevent memory copying
template<>
inline bool Writer<StringBuffer>::WriteInt(int i) {
char *buffer = os_->Push(11);
const char* end = internal::i32toa(i, buffer);
os_->Pop(static_cast<size_t>(11 - (end - buffer)));
return true;
}
template<>
inline bool Writer<StringBuffer>::WriteUint(unsigned u) {
char *buffer = os_->Push(10);
const char* end = internal::u32toa(u, buffer);
os_->Pop(static_cast<size_t>(10 - (end - buffer)));
return true;
}
template<>
inline bool Writer<StringBuffer>::WriteInt64(int64_t i64) {
char *buffer = os_->Push(21);
const char* end = internal::i64toa(i64, buffer);
os_->Pop(static_cast<size_t>(21 - (end - buffer)));
return true;
}
template<>
inline bool Writer<StringBuffer>::WriteUint64(uint64_t u) {
char *buffer = os_->Push(20);
const char* end = internal::u64toa(u, buffer);
os_->Pop(static_cast<size_t>(20 - (end - buffer)));
return true;
}
template<>
inline bool Writer<StringBuffer>::WriteDouble(double d) {
if (internal::Double(d).IsNanOrInf()) {
// Note: This code path can only be reached if (RAPIDJSON_WRITE_DEFAULT_FLAGS & kWriteNanAndInfFlag).
if (!(kWriteDefaultFlags & kWriteNanAndInfFlag))
return false;
if (internal::Double(d).IsNan()) {
PutReserve(*os_, 3);
PutUnsafe(*os_, 'N'); PutUnsafe(*os_, 'a'); PutUnsafe(*os_, 'N');
return true;
}
if (internal::Double(d).Sign()) {
PutReserve(*os_, 9);
PutUnsafe(*os_, '-');
}
else
PutReserve(*os_, 8);
PutUnsafe(*os_, 'I'); PutUnsafe(*os_, 'n'); PutUnsafe(*os_, 'f');
PutUnsafe(*os_, 'i'); PutUnsafe(*os_, 'n'); PutUnsafe(*os_, 'i'); PutUnsafe(*os_, 't'); PutUnsafe(*os_, 'y');
return true;
}
char *buffer = os_->Push(25);
char* end = internal::dtoa(d, buffer, maxDecimalPlaces_);
os_->Pop(static_cast<size_t>(25 - (end - buffer)));
return true;
}
#if defined(RAPIDJSON_SSE2) || defined(RAPIDJSON_SSE42)
template<>
inline bool Writer<StringBuffer>::ScanWriteUnescapedString(StringStream& is, size_t length) {
if (length < 16)
return RAPIDJSON_LIKELY(is.Tell() < length);
if (!RAPIDJSON_LIKELY(is.Tell() < length))
return false;
const char* p = is.src_;
const char* end = is.head_ + length;
const char* nextAligned = reinterpret_cast<const char*>((reinterpret_cast<size_t>(p) + 15) & static_cast<size_t>(~15));
const char* endAligned = reinterpret_cast<const char*>(reinterpret_cast<size_t>(end) & static_cast<size_t>(~15));
if (nextAligned > end)
return true;
while (p != nextAligned)
if (*p < 0x20 || *p == '\"' || *p == '\\') {
is.src_ = p;
return RAPIDJSON_LIKELY(is.Tell() < length);
}
else
os_->PutUnsafe(*p++);
// The rest of string using SIMD
static const char dquote[16] = { '\"', '\"', '\"', '\"', '\"', '\"', '\"', '\"', '\"', '\"', '\"', '\"', '\"', '\"', '\"', '\"' };
static const char bslash[16] = { '\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\' };
static const char space[16] = { 0x19, 0x19, 0x19, 0x19, 0x19, 0x19, 0x19, 0x19, 0x19, 0x19, 0x19, 0x19, 0x19, 0x19, 0x19, 0x19 };
const __m128i dq = _mm_loadu_si128(reinterpret_cast<const __m128i *>(&dquote[0]));
const __m128i bs = _mm_loadu_si128(reinterpret_cast<const __m128i *>(&bslash[0]));
const __m128i sp = _mm_loadu_si128(reinterpret_cast<const __m128i *>(&space[0]));
for (; p != endAligned; p += 16) {
const __m128i s = _mm_load_si128(reinterpret_cast<const __m128i *>(p));
const __m128i t1 = _mm_cmpeq_epi8(s, dq);
const __m128i t2 = _mm_cmpeq_epi8(s, bs);
const __m128i t3 = _mm_cmpeq_epi8(_mm_max_epu8(s, sp), sp); // s < 0x20 <=> max(s, 0x19) == 0x19
const __m128i x = _mm_or_si128(_mm_or_si128(t1, t2), t3);
unsigned short r = static_cast<unsigned short>(_mm_movemask_epi8(x));
if (RAPIDJSON_UNLIKELY(r != 0)) { // some of characters is escaped
SizeType len;
#ifdef _MSC_VER // Find the index of first escaped
unsigned long offset;
_BitScanForward(&offset, r);
len = offset;
#else
len = static_cast<SizeType>(__builtin_ffs(r) - 1);
#endif
char* q = reinterpret_cast<char*>(os_->PushUnsafe(len));
for (size_t i = 0; i < len; i++)
q[i] = p[i];
p += len;
break;
}
_mm_storeu_si128(reinterpret_cast<__m128i *>(os_->PushUnsafe(16)), s);
}
is.src_ = p;
return RAPIDJSON_LIKELY(is.Tell() < length);
}
#endif // defined(RAPIDJSON_SSE2) || defined(RAPIDJSON_SSE42)
RAPIDJSON_NAMESPACE_END
#ifdef _MSC_VER
RAPIDJSON_DIAG_POP
#endif
#ifdef __clang__
RAPIDJSON_DIAG_POP
#endif
#endif // RAPIDJSON_RAPIDJSON_H_

87
socks.h Normal file
View File

@ -0,0 +1,87 @@
#pragma once
#ifdef _WIN32
#ifndef _WIN32_WINNT
#define _WIN32_WINNT 0x0601 /* Windows 7 */
#endif
#include <winsock2.h>
#include <ws2tcpip.h>
#include <windows.h>
inline void sock_init()
{
static bool bWSAInit = false;
if (!bWSAInit)
{
WSADATA wsaData;
WSAStartup(MAKEWORD(2, 2), &wsaData);
bWSAInit = true;
}
}
inline void sock_close(SOCKET s)
{
shutdown(s, SD_BOTH);
closesocket(s);
}
inline const char* sock_strerror(char* buf, size_t len)
{
buf[0] = '\0';
FormatMessageA(
FORMAT_MESSAGE_FROM_SYSTEM | FORMAT_MESSAGE_IGNORE_INSERTS | FORMAT_MESSAGE_MAX_WIDTH_MASK,
NULL, WSAGetLastError(),
MAKELANGID(LANG_NEUTRAL, SUBLANG_DEFAULT),
(LPSTR)buf, len, NULL);
return buf;
}
inline const char* sock_gai_strerror(int err, char* buf, size_t len)
{
buf[0] = '\0';
FormatMessageA(
FORMAT_MESSAGE_FROM_SYSTEM | FORMAT_MESSAGE_IGNORE_INSERTS | FORMAT_MESSAGE_MAX_WIDTH_MASK,
NULL, (DWORD)err,
MAKELANGID(LANG_NEUTRAL, SUBLANG_DEFAULT),
(LPSTR)buf, len, NULL);
return buf;
}
#else
/* Assume that any non-Windows platform uses POSIX-style sockets instead. */
#include <sys/socket.h>
#include <arpa/inet.h>
#include <netdb.h> /* Needed for getaddrinfo() and freeaddrinfo() */
#include <unistd.h> /* Needed for close() */
#include <errno.h>
#include <string.h>
inline void sock_init() {}
typedef int SOCKET;
#define INVALID_SOCKET (-1)
#define SOCKET_ERROR (-1)
inline void sock_close(SOCKET s)
{
shutdown(s, SHUT_RDWR);
close(s);
}
inline const char* sock_strerror(char* buf, size_t len)
{
buf[0] = '\0';
return strerror_r(errno, buf, len);
}
inline const char* sock_gai_strerror(int err, char* buf, size_t len)
{
buf[0] = '\0';
return gai_strerror(err);
}
#endif

49
thdq.hpp Normal file
View File

@ -0,0 +1,49 @@
#pragma once
#include <queue>
#include <thread>
#include <mutex>
#include <condition_variable>
template <typename T>
class thdq
{
public:
T pop()
{
std::unique_lock<std::mutex> mlock(mutex_);
while (queue_.empty()) { cond_.wait(mlock); }
auto item = std::move(queue_.front());
queue_.pop();
return item;
}
void pop(T& item)
{
std::unique_lock<std::mutex> mlock(mutex_);
while (queue_.empty()) { cond_.wait(mlock); }
item = queue_.front();
queue_.pop();
}
void push(const T& item)
{
std::unique_lock<std::mutex> mlock(mutex_);
queue_.push(item);
mlock.unlock();
cond_.notify_one();
}
void push(T&& item)
{
std::unique_lock<std::mutex> mlock(mutex_);
queue_.push(std::move(item));
mlock.unlock();
cond_.notify_one();
}
private:
std::queue<T> queue_;
std::mutex mutex_;
std::condition_variable cond_;
};

164
xmr-stak-amd.cbp Normal file
View File

@ -0,0 +1,164 @@
<?xml version="1.0" encoding="UTF-8" standalone="yes" ?>
<CodeBlocks_project_file>
<FileVersion major="1" minor="6" />
<Project>
<Option title="xmr-stak-amd" />
<Option pch_mode="2" />
<Option compiler="gcc" />
<Build>
<Target title="Debug">
<Option output="bin/Debug/miner" prefix_auto="1" extension_auto="1" />
<Option object_output="obj/Debug/" />
<Option type="1" />
<Option compiler="gcc" />
<Option parameters="config-debug.txt" />
<Compiler>
<Add option="-march=corei7-avx" />
<Add option="-std=c++11" />
<Add option="-m64" />
<Add option="-g" />
<Add directory="include" />
</Compiler>
<Linker>
<Add option="-m64" />
</Linker>
</Target>
<Target title="Release">
<Option output="bin/Release/miner" prefix_auto="1" extension_auto="1" />
<Option object_output="obj/Release/" />
<Option type="1" />
<Option compiler="gcc" />
<Option parameters="config-debug.txt" />
<Compiler>
<Add option="-march=westmere" />
<Add option="-O3" />
<Add option="-std=c++11" />
<Add option="-m64" />
<Add option="-DNDEBUG" />
<Add directory="include" />
</Compiler>
<Linker>
<Add option="-s" />
<Add option="-m64" />
</Linker>
</Target>
<Target title="Release_test">
<Option output="bin/Release_test/miner" prefix_auto="1" extension_auto="1" />
<Option object_output="obj/Release_test/" />
<Option type="1" />
<Option compiler="gcc" />
<Option parameters="config-debug.txt" />
<Compiler>
<Add option="-march=westmere" />
<Add option="-O3" />
<Add option="-std=c++11" />
<Add option="-m64" />
<Add directory="include" />
</Compiler>
<Linker>
<Add option="-s" />
<Add option="-m64" />
</Linker>
</Target>
</Build>
<Compiler>
<Add option="-Wall" />
</Compiler>
<Linker>
<Add library="pthread" />
<Add library="libmicrohttpd" />
<Add library="OpenCL" />
</Linker>
<Unit filename="amd_gpu/gpu.c">
<Option compilerVar="CC" />
</Unit>
<Unit filename="amd_gpu/gpu.h" />
<Unit filename="cli-miner.cpp" />
<Unit filename="console.cpp" />
<Unit filename="console.h" />
<Unit filename="crypto/c_blake256.c">
<Option compilerVar="CC" />
</Unit>
<Unit filename="crypto/c_blake256.h" />
<Unit filename="crypto/c_groestl.c">
<Option compilerVar="CC" />
</Unit>
<Unit filename="crypto/c_groestl.h" />
<Unit filename="crypto/c_jh.c">
<Option compilerVar="CC" />
</Unit>
<Unit filename="crypto/c_jh.h" />
<Unit filename="crypto/c_keccak.c">
<Option compilerVar="CC" />
</Unit>
<Unit filename="crypto/c_keccak.h" />
<Unit filename="crypto/c_skein.c">
<Option compilerVar="CC" />
</Unit>
<Unit filename="crypto/c_skein.h" />
<Unit filename="crypto/cryptonight.h" />
<Unit filename="crypto/cryptonight_aesni.h" />
<Unit filename="crypto/cryptonight_common.cpp" />
<Unit filename="crypto/groestl_tables.h" />
<Unit filename="crypto/hash.h" />
<Unit filename="crypto/int-util.h" />
<Unit filename="crypto/skein_port.h" />
<Unit filename="crypto/soft_aes.c">
<Option compilerVar="CC" />
</Unit>
<Unit filename="donate-level.h" />
<Unit filename="executor.cpp" />
<Unit filename="executor.h" />
<Unit filename="httpd.cpp" />
<Unit filename="httpd.h" />
<Unit filename="jconf.cpp" />
<Unit filename="jconf.h" />
<Unit filename="jext.h" />
<Unit filename="jpsock.cpp" />
<Unit filename="jpsock.h" />
<Unit filename="minethd.cpp" />
<Unit filename="minethd.h" />
<Unit filename="msgstruct.h" />
<Unit filename="rapidjson/allocators.h" />
<Unit filename="rapidjson/document.h" />
<Unit filename="rapidjson/encodedstream.h" />
<Unit filename="rapidjson/encodings.h" />
<Unit filename="rapidjson/error/en.h" />
<Unit filename="rapidjson/error/error.h" />
<Unit filename="rapidjson/filereadstream.h" />
<Unit filename="rapidjson/filewritestream.h" />
<Unit filename="rapidjson/fwd.h" />
<Unit filename="rapidjson/internal/biginteger.h" />
<Unit filename="rapidjson/internal/diyfp.h" />
<Unit filename="rapidjson/internal/dtoa.h" />
<Unit filename="rapidjson/internal/ieee754.h" />
<Unit filename="rapidjson/internal/itoa.h" />
<Unit filename="rapidjson/internal/meta.h" />
<Unit filename="rapidjson/internal/pow10.h" />
<Unit filename="rapidjson/internal/regex.h" />
<Unit filename="rapidjson/internal/stack.h" />
<Unit filename="rapidjson/internal/strfunc.h" />
<Unit filename="rapidjson/internal/strtod.h" />
<Unit filename="rapidjson/internal/swap.h" />
<Unit filename="rapidjson/istreamwrapper.h" />
<Unit filename="rapidjson/memorybuffer.h" />
<Unit filename="rapidjson/memorystream.h" />
<Unit filename="rapidjson/msinttypes/inttypes.h" />
<Unit filename="rapidjson/msinttypes/stdint.h" />
<Unit filename="rapidjson/ostreamwrapper.h" />
<Unit filename="rapidjson/pointer.h" />
<Unit filename="rapidjson/prettywriter.h" />
<Unit filename="rapidjson/rapidjson.h" />
<Unit filename="rapidjson/reader.h" />
<Unit filename="rapidjson/schema.h" />
<Unit filename="rapidjson/stream.h" />
<Unit filename="rapidjson/stringbuffer.h" />
<Unit filename="rapidjson/writer.h" />
<Unit filename="socks.h" />
<Unit filename="thdq.hpp" />
<Extensions>
<code_completion />
<debugger />
</Extensions>
</Project>
</CodeBlocks_project_file>