commit e2c517b529886d3367a43a8ef5eed0f87ec4af00 Author: Scott Duensing Date: Fri Dec 19 20:11:05 2025 -0600 Here we go again. diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..c0a2c22 --- /dev/null +++ b/.gitattributes @@ -0,0 +1,30 @@ +# +# Roo/E, the Kangaroo Punch Portable GUI Toolkit +# Copyright (C) 2026 Scott Duensing +# +# http://kangaroopunch.com +# +# +# This file is part of Roo/E. +# +# Roo/E is free software: you can redistribute it and/or modify it under the +# terms of the GNU Affero General Public License as published by the Free +# Software Foundation, either version 3 of the License, or (at your option) +# any later version. +# +# Roo/E 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 Affero General Public License +# along with Roo/E. If not, see . +# + +*.fnt filter=lfs diff=lfs merge=lfs -text +*.png filter=lfs diff=lfs merge=lfs -text +*.jpg filter=lfs diff=lfs merge=lfs -text +*.gif filter=lfs diff=lfs merge=lfs -text +*.zip filter=lfs diff=lfs merge=lfs -text +*.EXE filter=lfs diff=lfs merge=lfs -text +*.a filter=lfs diff=lfs merge=lfs -text diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..b4b72b9 --- /dev/null +++ b/.gitignore @@ -0,0 +1,28 @@ +# +# Roo/E, the Kangaroo Punch Portable GUI Toolkit +# Copyright (C) 2026 Scott Duensing +# +# http://kangaroopunch.com +# +# +# This file is part of Roo/E. +# +# Roo/E is free software: you can redistribute it and/or modify it under the +# terms of the GNU Affero General Public License as published by the Free +# Software Foundation, either version 3 of the License, or (at your option) +# any later version. +# +# Roo/E 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 Affero General Public License +# along with Roo/E. If not, see . +# + +.idea/ +bin/ +obj/ + +*~ diff --git a/3rdparty/CWSDPMI/CWSDSTUB.EXE b/3rdparty/CWSDPMI/CWSDSTUB.EXE new file mode 100644 index 0000000..ff1812e --- /dev/null +++ b/3rdparty/CWSDPMI/CWSDSTUB.EXE @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:15cb50f3cc7022cd1148b4569813b33f4b56c3f31ebdd296c7f97906d2543625 +size 22528 diff --git a/3rdparty/CWSDPMI/csdpmi7b.zip b/3rdparty/CWSDPMI/csdpmi7b.zip new file mode 100644 index 0000000..415ce08 --- /dev/null +++ b/3rdparty/CWSDPMI/csdpmi7b.zip @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:deacda0488e1cdd7c4a9f32fab45662b34c0ed6b2d7d4d13bc07041b62004a8c +size 71339 diff --git a/3rdparty/CWSDPMI/csdpmi7s.zip b/3rdparty/CWSDPMI/csdpmi7s.zip new file mode 100644 index 0000000..77e7a1a --- /dev/null +++ b/3rdparty/CWSDPMI/csdpmi7s.zip @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:882d29200a720ef5ad3591afc27f7d66a5fc575dddcd6590acea346813aa6d7f +size 89872 diff --git a/3rdparty/memwatch/FAQ b/3rdparty/memwatch/FAQ new file mode 100644 index 0000000..efcec58 --- /dev/null +++ b/3rdparty/memwatch/FAQ @@ -0,0 +1,133 @@ +Frequently Asked Questions for memwatch + +Q. I'm not getting any log file! What's wrong?? + +A. Did you define MEMWATCH when compiling all files? + Did you include memwatch.h in all the files? + If you did, then...: + + Memwatch creates the file when it initializes. If you're not + getting the log file, it's because a) memwatch is not + initializing or b) it's initializing, but can't create the + file. + + Memwatch has two functions, mwInit() and mwTerm(), that + initialize and terminate memwatch, respectively. They are + nestable. You USUALLY don't need to call mwInit() and + mwTerm(), since memwatch will auto-initialize on the first + call to a memory function, and then add mwTerm() to the + atexit() list. + + You can call mwInit() and mwTerm() manually, if it's not + initializing properly or if your system doesn't support + atexit(). Call mwInit() as soon as you can, and mwTerm() at + the logical no-error ending of your program. Call mwAbort() + if the program is stopping due to an error; this will + terminate memwatch even if more than one call to mwTerm() is + outstanding. + + If you are using C++, remember that global and static C++ + objects constructors execute before main() when considering + where to put mwInit(). Also, their destructors execute after + main(). You may want to create a global object very early + with mwInit() in the constructor and mwTerm() in the + destructor. Too bad C++ does not guarantee initialization + order for global objects. + + If this didn't help, try adding a call to mwDoFlush(1) after + mwInit(). If THAT didn't help, then memwatch is unable to + create the log file. Check write permissions. + + If you can't use a log file, you can still use memwatch by + redirecting the output to a function of your choice. See the + next question. + +Q. I'd like memwatch's output to pipe to my fave debugger! How? + +A. Call mwSetOutFunc() with the address of a "void func(int c)" + function. You should also consider doing something about + the ARI handler, see memwatch.h for more details about that. + +Q. Why isn't there any C++ support? + +A. Because C++ is for sissies! =) Just kidding. + C++ comes with overridable allocation/deallocation + built-in. You can define your own new/delete operators + for any class, and thus circumvent memwatch, or confuse + it to no end. Also, the keywords "new" and "delete" may + appear in declarations in C++, making the preprocessor + replacement approach shaky. You can do it, but it's not + very stable. + + If someone were to write a rock solid new/delete checker + for C++, there is no conflict with memwatch; use them both. + +Q. I'm getting "WILD free" errors, but the code is bug-free! + +A. If memwatch's free() recieves a pointer that wasn't allocated + by memwatch, a "WILD free" message appears. If the source of + the memory buffer is outside of memwatch (a non-standard + library function, for instance), you can use mwFree_() to + release it. mwFree_() calls free() on the pointer given if + memwatch can't recognize it, instead of blocking it. + + Another source of "WILD free" messages is that if memwatch + is terminated before all memory allocated is freed, memwatch + will have forgotten about it, and thus generate the errors. + This is commonly caused by having memwatch auto-initialize, + and then using atexit() to clean up. When auto-initializing, + memwatch registers mwTerm() with atexit(), but if mwTerm() + runs before all memory is freed, then you will get "unfreed" + and "WILD free" messages when your own atexit()-registered + cleanup code runs, and frees the memory. + +Q. I'm getting "unfreed" errors, but the code is bug-free! + +A. You can get erroneous "unfreed" messages if memwatch + terminates before all memory has been freed. Try using + mwInit() and mwTerm() instead of auto-initialization. + + If you _are_ using mwInit() and mwTerm(), it may be that + some code in your program executes before mwInit() or + after mwTerm(). Make sure that mwInit() is the first thing + executed, and mwTerm() the last. + +Q. When compiling memwatch I get these 'might get clobbered' + errors, and something about a longjmp() inside memwatch. + +A. When using gcc or egcs with the optimization to inline + functions, this warning occurs. This is because gcc and + egcs inlines memwatch's functions with setjmp/longjmp, + causing the calling functions to become unstable. + + The gcc/egcs maintainers have been informed of this + problem, but until they modify the inline optimization + so that it leaves setjmp functions alone, make sure to + compile memwatch without inline function optimizations. + + gcc 2.95.2 can be patched for this, and I have been told + it will be fixed in an upcoming version. + +Q. My program crashes with SIGSEGV or alignment errors, but + only when I compile with memwatch enabled! + +A. You are using a 64-bit (or higher) platform, and memwatch + was unable to detect and adjust for this. You'll have to + either compile with a suitable define for mwROUNDALLOC, + I suggest (number of bits / 8), or define mwROUNDALLOC + directly in memwatch.c. + + Also, please check your limits.h file for the relevant + #defines, and tell me what they are. + +Q. When I include string.h after memwatch.h, I get errors + related to strdup(), what gives? + +A. Most, but probably not all, platforms are nice about + including files multiple times, so I could probably + avoid these errors by including string.h from memwatch.h. + But since I want to be on the safe side, I don't. + + To fix this, simply include string.h before memwatch.h, + or modify memwatch.h to include string.h. + diff --git a/3rdparty/memwatch/Makefile b/3rdparty/memwatch/Makefile new file mode 100644 index 0000000..bc1be19 --- /dev/null +++ b/3rdparty/memwatch/Makefile @@ -0,0 +1,2 @@ +test: + $(CC) -DMEMWATCH -DMW_STDIO test.c memwatch.c diff --git a/3rdparty/memwatch/README b/3rdparty/memwatch/README new file mode 100644 index 0000000..aeb86ce --- /dev/null +++ b/3rdparty/memwatch/README @@ -0,0 +1,99 @@ +README for MEMWATCH 2.69 + + This file should be enough to get you started, and should be + enough for small projects. For more info, see the files USING + and the FAQ. If this is not enough, see memwatch.h, which is + well documented. + + Memwatch is licensed under the GPL from version 2.69 + onwards. Please read the file gpl.txt for more details. + + If you choose to use memwatch to validate your projects, I + would like to hear about it. Please drop me a line at + johan@linkdata.se about the project itself, the hardware, + operating system, compiler and any URL(s) you feel is + appropriate. + +***** To run the test program: + + Look at the source code for test.c first. It does some really + nasty things, and I want you to be aware of that. If memwatch + can't capture SIGSEGV (General Protection Fault for Windoze), + your program will dump core (crash for Windoze). + + Once you've done that, you can build the test program. + + Linux and other *nixes with gcc: + + gcc -o test -DMEMWATCH -DMEMWATCH_STDIO test.c memwatch.c + + Windows 95, Windows NT with MS Visual C++: + + cl -DMEMWATCH -DMEMWATCH_STDIO test.c memwatch.c + + Then simply run the test program. + + ./test + + +***** Quick-start instructions: + + 1. Make sure that memwatch.h is included in all of the + source code files. If you have an include file that + all of the source code uses, you might be able to include + memwatch.h from there. + + 2. Recompile the program with MEMWATCH defined. See your + compiler's documentation if you don't know how to do this. + The usual switch looks like "-DMEMWATCH". To have MEMWATCH + use stderr for some output (like, "Abort, Retry, Ignore?"), + please also define MW_STDIO (or MEMWATCH_STDIO, same thing). + + 3. Run the program and examine the output in the + log file "memwatch.log". If you didn't get a log file, + you probably didn't do step 1 and 2 correctly, or your + program crashed before memwatch flushed the file buffer. + To have memwatch _always_ flush the buffer, add a call + to "mwDoFlush(1)" at the top of your main function. + + 4. There is no fourth step... but remember that there + are limits to what memwatch can do, and that you need + to be aware of them: + +***** Limits to memwatch: + + Memwatch cannot catch all wild pointer writes. It can catch + those it could make itself due to your program trashing + memwatch's internal data structures. It can catch, sort of, + wild writes into No Mans Land buffers (see the header file for + more info). Anything else and you're going to get core dumped, + or data corruption if you're lucky. + + There are other limits of course, but that one is the most + serious one, and the one that you're likely to be suffering + from. + +***** Can use memwatch with XXXXX? + + Probably the answer is yes. It's been tested with several + different platforms and compilers. It may not work on yours + though... but there's only one way to find out. + +***** Need more assistance? + + I don't want e-mail on "how to program in C", or "I've got a + bug, help me". I _do_ want you to send email to me if you + find a bug in memwatch, or if it won't compile cleanly on your + system (assuming it's an ANSI-C compiler of course). + + If you need help with using memwatch, read the header file. + If, after reading the header file, you still can't resolve the + problem, please mail me with the details. + + I can be reached at "johan@linkdata.se". + + The latest version of memwatch should be found at + "http://www.linkdata.se/". + + Johan Lindh + diff --git a/3rdparty/memwatch/USING b/3rdparty/memwatch/USING new file mode 100644 index 0000000..3d78e95 --- /dev/null +++ b/3rdparty/memwatch/USING @@ -0,0 +1,213 @@ +Using memwatch +============== + +What is it? + + Memwatch is primarily a memory leak detector for C. Besides + detecting leaks, it can do a bunch of other stuff, but lets + stay to the basics. If you _really_ want to know all the + gory details, you should check out the header file, + memwatch.h, and the source code. It's actually got some + comments! (Whoa, what a concept!) + +How do I get the latest version? + + http://www.linkdata.se/sourcecode.html + ftp://ftp.linkdata.se/pub/memwatch/ + +How does it work? + + Using the C preprocessor, memwatch replaces all your + programs calls to ANSI C memory allocation functions with + calls to it's own functions, which keeps a record of all + allocations. + + Memwatch is very unobtrusive; unless the define MEMWATCH is + defined, memwatch removes all traces of itself from the + code (using the preprocessor). + + Memwatch normally writes it's data to the file + memwatch.log, but this can be overridden; see the section + on I/O, later. + +Can I use it for my C++ sources? + + You can, but it's not recommended. C++ allows individual + classes to have their own memory management, and the + preprocessor approach used by memwatch can cause havoc + with such class declarations if improperly used. + + If you have no such classes, or have them but still want + to test it, you can give it a try. + + First, re-enable the C++ support code in memwatch. + If you can't find it, you probably shouldn't be using + it. Then, in your source code, after including ALL + header files: + + #define new mwNew + #define delete mwDelete + + This will cause calls to new and delete in that source file + to be directed to memwatch. Also, be sure to read all the + text in memwatch.h regarding C++ support. + +Is this stuff thread-safe? + + I doubt it. As of version 2.66, there is rudimentary support + for threads, if you happen to be using Win32 or if you have + pthreads. Define WIN32 or MW_PTHREADS to signify this fact. + + This will cause a global mutex to be created, and memwatch + will lock it when accessing the global memory chain, but it's + still far from certified threadsafe. + +Initialization and cleanup + + In order to do it's work in a timely fashion, memwatch + needs to do some startup and cleanup work. mwInit() + initializes memwatch and mwTerm() terminates it. Memwatch + can auto-initialize, and will do so if you don't call + mwInit() yourself. If this is the case, memwatch will use + atexit() to register mwTerm() to the atexit-queue. + + The auto-init technique has a caveat; if you are using + atexit() yourself to do cleanup work, memwatch may + terminate before your program is done. To be on the safe + side, use mwInit() and mwTerm(). + + mwInit() and mwTerm() is nestable, so you can call mwInit() + several times, requiring mwTerm() to be called an equal + number of times to terminate memwatch. + + In case of the program aborting in a controlled way, you + may want to call mwAbort() instead of mwTerm(). mwAbort() + will terminate memwatch even if there are outstanding calls + to mwTerm(). + +I/O operations + + During normal operations, memwatch creates a file named + memwatch.log. Sometimes, memwatch.log can't be created; + then memwatch tries to create files name memwatNN.log, + where NN is between 01 and 99. If that fails, no log will + be produced. + + If you can't use a file log, or don't want to, no worry. + Just call mwSetOutFunc() with the address of a "void + func(int c)" function, and all output will be directed + there, character by character. + + Memwatch also has an Abort/Retry/Ignore handler that is + used when an ASSERT or VERIFY fails. The default handler + does no I/O, but automatically aborts the program. You can + use any handler you want; just send the address of a "int + func(const char*)" to mwSetAriFunc(). For more details on + that, see memwatch.h. + +TRACE/ASSERT/VERIFY macros + + Memwatch defines (if not already defined) the macros TRACE, + ASSERT and VERIFY. If you are already using macros with + these names, memwatch 2.61 and later will not override + them. Memwatch 2.61 and later will also always define the + macros mwTRACE, mwASSERT and mwVERIFY, so you can use these + to make sure you're talking to memwatch. Versions prior + to 2.61 will *OVERRIDE* existing TRACE, ASSERT and VERIFY. + + To make sure that existing TRACE, ASSERT and VERIFY macros + are preserved, you can define MW_NOTRACE, MW_NOASSERT and + MW_NOVERIFY. All versions of memwatch will abide by these. + +How slow can you go? + + Memwatch slows things down. Large allocations aren't + affected so that you can measure it, but small allocations + that would utilize a compilers small-allocator function + suddenly cannot, and so gets slowed down alot. As a worst + case, expect it to be 3-5 times slower. + + Free'ing gets hit worse, I'm afraid, as memwatch checks + a lot of stuff when freeing. Expect it to be 5-7 times + slower, no matter what the size of the allocation. + +Stress-testing the application + + You can simulate low-memory conditions using mwLimit(). + mwLimit() takes the maximum number of bytes to be + allocated; when the limit is hit, allocation requests will + fail, and a "limit" message will be logged. + + If you hit a real low-memory situation, memwatch logs that + too. Memwatch itself has some reserve memory tucked away so + it should continue running even in the worst conditions. + +Hunting down wild writes and other Nasty Things + + Wild writes are usually caused by using pointers that arent + initialized, or that were initialized, but then the memory + they points to is moved or freed. The best way to avoid + these kind of problems is to ALWAYS initialize pointers to + NULL, and after freeing a memory buffer, setting all + pointers that pointed to it to NULL. + + To aid in tracking down uninitialized pointers memwatch + zaps all memory with certain values. Recently allocated + memory (unless calloc'd, of course), contains 0xFE. + Recently freed memory contains 0xFD. So if your program + crashes when using memwatch and not without memwatch, it's + most likely because you are not initializing your allocated + buffers, or using the buffers after they've been freed. + + In the event that a wild pointer should damage memwatch's + internal data structures, memwatch employs checksums, + multiple copies of some values, and can also repair it's + own data structures. + + If you are a paranoid person, and as programmer you should + be, you can use memwatch's mwIsReadAddr() and + mwIsSafeAddr() functions to check the accessibility of + memory. These are implemented for both ANSI C systems and + Win32 systems. Just put an mwASSERT() around the check and + forget about it. + +Can I help? + + Well, sure. For instance, I like memwatch to compile + without any warnings or errors. If you are using an ANSI C + compliant compiler, and are getting warnings or errors, + please mail me the details and instructions on how to fix + them, if you can. + + Another thing you can do if you decide to use memwatch is + to mail me the name of the project(s) (and URL, if any), + hardware and operating system, compiler and what user + (organization). I will then post this info on the list of + memwatch users. + (http://www.linkdata.se/memwatchusers.html) + +Top five problems using memwatch + + 5. Passed a non-memwatch allocated pointer to memwatch's + free(). Symtom: Causes an erroneous "WILD free" log + entry to appear. Cure: Either include memwatch.h for + the file that allocates, or use mwFree_() to free it. + + 4. Relied on auto-initialization when using atexit(). + Symptom: Causes incorrect "unfreed" and "WILD free" + messages. Cure: Use mwInit() and mwTerm(). + + 3. Forgot to include memwatch.h in all files. Symptom: + Tends to generate "WILD free" and "unfreed" messages. + Cure: Make sure to include memwatch.h! + + 2. No write permissions in currect directory. Symptom: + Seems like memwatch 'just aint working'. Cure: Use + mwSetOutFunc() to redirect output. + + ...and the number one problem is... + + 1. Didn't define MEMWATCH when compiling. Symptom: + Memwatch dutifully disables itself. Cure: Try adding + -DMEMWATCH to the command line. + diff --git a/3rdparty/memwatch/gpl.txt b/3rdparty/memwatch/gpl.txt new file mode 100644 index 0000000..5b6e7c6 --- /dev/null +++ b/3rdparty/memwatch/gpl.txt @@ -0,0 +1,340 @@ + GNU GENERAL PUBLIC LICENSE + Version 2, June 1991 + + Copyright (C) 1989, 1991 Free Software Foundation, Inc. + 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + Everyone is permitted to copy and distribute verbatim copies + of this license document, but changing it is not allowed. + + Preamble + + The licenses for most software are designed to take away your +freedom to share and change it. By contrast, the GNU General Public +License is intended to guarantee your freedom to share and change free +software--to make sure the software is free for all its users. This +General Public License applies to most of the Free Software +Foundation's software and to any other program whose authors commit to +using it. (Some other Free Software Foundation software is covered by +the GNU Library General Public License instead.) You can apply it to +your programs, too. + + When we speak of free software, we are referring to freedom, not +price. Our General Public Licenses are designed to make sure that you +have the freedom to distribute copies of free software (and charge for +this service if you wish), that you receive source code or can get it +if you want it, that you can change the software or use pieces of it +in new free programs; and that you know you can do these things. + + To protect your rights, we need to make restrictions that forbid +anyone to deny you these rights or to ask you to surrender the rights. +These restrictions translate to certain responsibilities for you if you +distribute copies of the software, or if you modify it. + + For example, if you distribute copies of such a program, whether +gratis or for a fee, you must give the recipients all the rights that +you have. You must make sure that they, too, receive or can get the +source code. And you must show them these terms so they know their +rights. + + We protect your rights with two steps: (1) copyright the software, and +(2) offer you this license which gives you legal permission to copy, +distribute and/or modify the software. + + Also, for each author's protection and ours, we want to make certain +that everyone understands that there is no warranty for this free +software. If the software is modified by someone else and passed on, we +want its recipients to know that what they have is not the original, so +that any problems introduced by others will not reflect on the original +authors' reputations. + + Finally, any free program is threatened constantly by software +patents. We wish to avoid the danger that redistributors of a free +program will individually obtain patent licenses, in effect making the +program proprietary. To prevent this, we have made it clear that any +patent must be licensed for everyone's free use or not licensed at all. + + The precise terms and conditions for copying, distribution and +modification follow. + + GNU GENERAL PUBLIC LICENSE + TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION + + 0. This License applies to any program or other work which contains +a notice placed by the copyright holder saying it may be distributed +under the terms of this General Public License. The "Program", below, +refers to any such program or work, and a "work based on the Program" +means either the Program or any derivative work under copyright law: +that is to say, a work containing the Program or a portion of it, +either verbatim or with modifications and/or translated into another +language. (Hereinafter, translation is included without limitation in +the term "modification".) Each licensee is addressed as "you". + +Activities other than copying, distribution and modification are not +covered by this License; they are outside its scope. The act of +running the Program is not restricted, and the output from the Program +is covered only if its contents constitute a work based on the +Program (independent of having been made by running the Program). +Whether that is true depends on what the Program does. + + 1. You may copy and distribute verbatim copies of the Program's +source code as you receive it, in any medium, provided that you +conspicuously and appropriately publish on each copy an appropriate +copyright notice and disclaimer of warranty; keep intact all the +notices that refer to this License and to the absence of any warranty; +and give any other recipients of the Program a copy of this License +along with the Program. + +You may charge a fee for the physical act of transferring a copy, and +you may at your option offer warranty protection in exchange for a fee. + + 2. You may modify your copy or copies of the Program or any portion +of it, thus forming a work based on the Program, and copy and +distribute such modifications or work under the terms of Section 1 +above, provided that you also meet all of these conditions: + + a) You must cause the modified files to carry prominent notices + stating that you changed the files and the date of any change. + + b) You must cause any work that you distribute or publish, that in + whole or in part contains or is derived from the Program or any + part thereof, to be licensed as a whole at no charge to all third + parties under the terms of this License. + + c) If the modified program normally reads commands interactively + when run, you must cause it, when started running for such + interactive use in the most ordinary way, to print or display an + announcement including an appropriate copyright notice and a + notice that there is no warranty (or else, saying that you provide + a warranty) and that users may redistribute the program under + these conditions, and telling the user how to view a copy of this + License. (Exception: if the Program itself is interactive but + does not normally print such an announcement, your work based on + the Program is not required to print an announcement.) + +These requirements apply to the modified work as a whole. If +identifiable sections of that work are not derived from the Program, +and can be reasonably considered independent and separate works in +themselves, then this License, and its terms, do not apply to those +sections when you distribute them as separate works. But when you +distribute the same sections as part of a whole which is a work based +on the Program, the distribution of the whole must be on the terms of +this License, whose permissions for other licensees extend to the +entire whole, and thus to each and every part regardless of who wrote it. + +Thus, it is not the intent of this section to claim rights or contest +your rights to work written entirely by you; rather, the intent is to +exercise the right to control the distribution of derivative or +collective works based on the Program. + +In addition, mere aggregation of another work not based on the Program +with the Program (or with a work based on the Program) on a volume of +a storage or distribution medium does not bring the other work under +the scope of this License. + + 3. You may copy and distribute the Program (or a work based on it, +under Section 2) in object code or executable form under the terms of +Sections 1 and 2 above provided that you also do one of the following: + + a) Accompany it with the complete corresponding machine-readable + source code, which must be distributed under the terms of Sections + 1 and 2 above on a medium customarily used for software interchange; or, + + b) Accompany it with a written offer, valid for at least three + years, to give any third party, for a charge no more than your + cost of physically performing source distribution, a complete + machine-readable copy of the corresponding source code, to be + distributed under the terms of Sections 1 and 2 above on a medium + customarily used for software interchange; or, + + c) Accompany it with the information you received as to the offer + to distribute corresponding source code. (This alternative is + allowed only for noncommercial distribution and only if you + received the program in object code or executable form with such + an offer, in accord with Subsection b above.) + +The source code for a work means the preferred form of the work for +making modifications to it. For an executable work, complete source +code means all the source code for all modules it contains, plus any +associated interface definition files, plus the scripts used to +control compilation and installation of the executable. However, as a +special exception, the source code distributed need not include +anything that is normally distributed (in either source or binary +form) with the major components (compiler, kernel, and so on) of the +operating system on which the executable runs, unless that component +itself accompanies the executable. + +If distribution of executable or object code is made by offering +access to copy from a designated place, then offering equivalent +access to copy the source code from the same place counts as +distribution of the source code, even though third parties are not +compelled to copy the source along with the object code. + + 4. You may not copy, modify, sublicense, or distribute the Program +except as expressly provided under this License. Any attempt +otherwise to copy, modify, sublicense or distribute the Program is +void, and will automatically terminate your rights under this License. +However, parties who have received copies, or rights, from you under +this License will not have their licenses terminated so long as such +parties remain in full compliance. + + 5. You are not required to accept this License, since you have not +signed it. However, nothing else grants you permission to modify or +distribute the Program or its derivative works. These actions are +prohibited by law if you do not accept this License. Therefore, by +modifying or distributing the Program (or any work based on the +Program), you indicate your acceptance of this License to do so, and +all its terms and conditions for copying, distributing or modifying +the Program or works based on it. + + 6. Each time you redistribute the Program (or any work based on the +Program), the recipient automatically receives a license from the +original licensor to copy, distribute or modify the Program subject to +these terms and conditions. You may not impose any further +restrictions on the recipients' exercise of the rights granted herein. +You are not responsible for enforcing compliance by third parties to +this License. + + 7. If, as a consequence of a court judgment or allegation of patent +infringement or for any other reason (not limited to patent issues), +conditions are imposed on you (whether by court order, agreement or +otherwise) that contradict the conditions of this License, they do not +excuse you from the conditions of this License. If you cannot +distribute so as to satisfy simultaneously your obligations under this +License and any other pertinent obligations, then as a consequence you +may not distribute the Program at all. For example, if a patent +license would not permit royalty-free redistribution of the Program by +all those who receive copies directly or indirectly through you, then +the only way you could satisfy both it and this License would be to +refrain entirely from distribution of the Program. + +If any portion of this section is held invalid or unenforceable under +any particular circumstance, the balance of the section is intended to +apply and the section as a whole is intended to apply in other +circumstances. + +It is not the purpose of this section to induce you to infringe any +patents or other property right claims or to contest validity of any +such claims; this section has the sole purpose of protecting the +integrity of the free software distribution system, which is +implemented by public license practices. Many people have made +generous contributions to the wide range of software distributed +through that system in reliance on consistent application of that +system; it is up to the author/donor to decide if he or she is willing +to distribute software through any other system and a licensee cannot +impose that choice. + +This section is intended to make thoroughly clear what is believed to +be a consequence of the rest of this License. + + 8. If the distribution and/or use of the Program is restricted in +certain countries either by patents or by copyrighted interfaces, the +original copyright holder who places the Program under this License +may add an explicit geographical distribution limitation excluding +those countries, so that distribution is permitted only in or among +countries not thus excluded. In such case, this License incorporates +the limitation as if written in the body of this License. + + 9. The Free Software Foundation may publish revised and/or new versions +of the General Public License from time to time. Such new versions will +be similar in spirit to the present version, but may differ in detail to +address new problems or concerns. + +Each version is given a distinguishing version number. If the Program +specifies a version number of this License which applies to it and "any +later version", you have the option of following the terms and conditions +either of that version or of any later version published by the Free +Software Foundation. If the Program does not specify a version number of +this License, you may choose any version ever published by the Free Software +Foundation. + + 10. If you wish to incorporate parts of the Program into other free +programs whose distribution conditions are different, write to the author +to ask for permission. For software which is copyrighted by the Free +Software Foundation, write to the Free Software Foundation; we sometimes +make exceptions for this. Our decision will be guided by the two goals +of preserving the free status of all derivatives of our free software and +of promoting the sharing and reuse of software generally. + + NO WARRANTY + + 11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY +FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN +OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES +PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED +OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF +MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS +TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE +PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, +REPAIR OR CORRECTION. + + 12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING +WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR +REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, +INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING +OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED +TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY +YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER +PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE +POSSIBILITY OF SUCH DAMAGES. + + END OF TERMS AND CONDITIONS + + How to Apply These Terms to Your New Programs + + If you develop a new program, and you want it to be of the greatest +possible use to the public, the best way to achieve this is to make it +free software which everyone can redistribute and change under these terms. + + To do so, attach the following notices to the program. It is safest +to attach them to the start of each source file to most effectively +convey the exclusion of warranty; and each file should have at least +the "copyright" line and a pointer to where the full notice is found. + + + Copyright (C) + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + + +Also add information on how to contact you by electronic and paper mail. + +If the program is interactive, make it output a short notice like this +when it starts in an interactive mode: + + Gnomovision version 69, Copyright (C) year name of author + Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'. + This is free software, and you are welcome to redistribute it + under certain conditions; type `show c' for details. + +The hypothetical commands `show w' and `show c' should show the appropriate +parts of the General Public License. Of course, the commands you use may +be called something other than `show w' and `show c'; they could even be +mouse-clicks or menu items--whatever suits your program. + +You should also get your employer (if you work as a programmer) or your +school, if any, to sign a "copyright disclaimer" for the program, if +necessary. Here is a sample; alter the names: + + Yoyodyne, Inc., hereby disclaims all copyright interest in the program + `Gnomovision' (which makes passes at compilers) written by James Hacker. + + , 1 April 1989 + Ty Coon, President of Vice + +This General Public License does not permit incorporating your program into +proprietary programs. If your program is a subroutine library, you may +consider it more useful to permit linking proprietary applications with the +library. If this is what you want to do, use the GNU Library General +Public License instead of this License. diff --git a/3rdparty/memwatch/memwatch.c b/3rdparty/memwatch/memwatch.c new file mode 100644 index 0000000..0c36e3a --- /dev/null +++ b/3rdparty/memwatch/memwatch.c @@ -0,0 +1,2667 @@ +/* +** MEMWATCH.C +** Nonintrusive ANSI C memory leak / overwrite detection +** Copyright (C) 1992-2003 Johan Lindh +** All rights reserved. +** Version 2.71 + + This file is part of MEMWATCH. + + MEMWATCH is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2 of the License, or + (at your option) any later version. + + MEMWATCH 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 MEMWATCH; if not, write to the Free Software + Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + +** +** 920810 JLI [1.00] +** 920830 JLI [1.10 double-free detection] +** 920912 JLI [1.15 mwPuts, mwGrab/Drop, mwLimit] +** 921022 JLI [1.20 ASSERT and VERIFY] +** 921105 JLI [1.30 C++ support and TRACE] +** 921116 JLI [1.40 mwSetOutFunc] +** 930215 JLI [1.50 modified ASSERT/VERIFY] +** 930327 JLI [1.51 better auto-init & PC-lint support] +** 930506 JLI [1.55 MemWatch class, improved C++ support] +** 930507 JLI [1.60 mwTest & CHECK()] +** 930809 JLI [1.65 Abort/Retry/Ignore] +** 930820 JLI [1.70 data dump when unfreed] +** 931016 JLI [1.72 modified C++ new/delete handling] +** 931108 JLI [1.77 mwSetAssertAction() & some small changes] +** 940110 JLI [1.80 no-mans-land alloc/checking] +** 940328 JLI [2.00 version 2.0 rewrite] +** Improved NML (no-mans-land) support. +** Improved performance (especially for free()ing!). +** Support for 'read-only' buffers (checksums) +** ^^ NOTE: I never did this... maybe I should? +** FBI (free'd block info) tagged before freed blocks +** Exporting of the mwCounter variable +** mwBreakOut() localizes debugger support +** Allocation statistics (global, per-module, per-line) +** Self-repair ability with relinking +** 950913 JLI [2.10 improved garbage handling] +** 951201 JLI [2.11 improved auto-free in emergencies] +** 960125 JLI [X.01 implemented auto-checking using mwAutoCheck()] +** 960514 JLI [2.12 undefining of existing macros] +** 960515 JLI [2.13 possibility to use default new() & delete()] +** 960516 JLI [2.20 suppression of file flushing on unfreed msgs] +** 960516 JLI [2.21 better support for using MEMWATCH with DLL's] +** 960710 JLI [X.02 multiple logs and mwFlushNow()] +** 960801 JLI [2.22 merged X.01 version with current] +** 960805 JLI [2.30 mwIsXXXXAddr() to avoid unneeded GP's] +** 960805 JLI [2.31 merged X.02 version with current] +** 961002 JLI [2.32 support for realloc() + fixed STDERR bug] +** 961222 JLI [2.40 added mwMark() & mwUnmark()] +** 970101 JLI [2.41 added over/underflow checking after failed ASSERT/VERIFY] +** 970113 JLI [2.42 added support for PC-Lint 7.00g] +** 970207 JLI [2.43 added support for strdup()] +** 970209 JLI [2.44 changed default filename to lowercase] +** 970405 JLI [2.45 fixed bug related with atexit() and some C++ compilers] +** 970723 JLI [2.46 added MW_ARI_NULLREAD flag] +** 970813 JLI [2.47 stabilized marker handling] +** 980317 JLI [2.48 ripped out C++ support; wasn't working good anyway] +** 980318 JLI [2.50 improved self-repair facilities & SIGSEGV support] +** 980417 JLI [2.51 more checks for invalid addresses] +** 980512 JLI [2.52 moved MW_ARI_NULLREAD to occur before aborting] +** 990112 JLI [2.53 added check for empty heap to mwIsOwned] +** 990217 JLI [2.55 improved the emergency repairs diagnostics and NML] +** 990224 JLI [2.56 changed ordering of members in structures] +** 990303 JLI [2.57 first maybe-fixit-for-hpux test] +** 990516 JLI [2.58 added 'static' to the definition of mwAutoInit] +** 990517 JLI [2.59 fixed some high-sensitivity warnings] +** 990610 JLI [2.60 fixed some more high-sensitivity warnings] +** 990715 JLI [2.61 changed TRACE/ASSERT/VERIFY macro names] +** 991001 JLI [2.62 added CHECK_BUFFER() and mwTestBuffer()] +** 991007 JLI [2.63 first shot at a 64-bit compatible version] +** 991009 JLI [2.64 undef's strdup() if defined, mwStrdup made const] +** 000704 JLI [2.65 added some more detection for 64-bits] +** 010502 JLI [2.66 incorporated some user fixes] +** [mwRelink() could print out garbage pointer (thanks mac@phobos.ca)] +** [added array destructor for C++ (thanks rdasilva@connecttel.com)] +** [added mutex support (thanks rdasilva@connecttel.com)] +** 010531 JLI [2.67 fix: mwMutexXXX() was declared even if MW_HAVE_MUTEX was not defined] +** 010619 JLI [2.68 fix: mwRealloc() could leave the mutex locked] +** 020918 JLI [2.69 changed to GPL, added C++ array allocation by Howard Cohen] +** 030212 JLI [2.70 mwMalloc() bug for very large allocations (4GB on 32bits)] +** 030520 JLI [2.71 added ULONG_LONG_MAX as a 64-bit detector (thanks Sami Salonen)] +*/ + +#define __MEMWATCH_C 1 + +#ifdef MW_NOCPP +#define MEMWATCH_NOCPP +#endif +#ifdef MW_STDIO +#define MEMWATCH_STDIO +#endif + +/*********************************************************************** +** Include files +***********************************************************************/ + +#include +#include +#include +#include +#include +#include +#include +#include +#include "memwatch.h" + +#ifndef toupper +#include +#endif + +#if defined(WIN32) || defined(__WIN32__) +#define MW_HAVE_MUTEX 1 +#include +#endif + +#if defined(MW_PTHREADS) || defined(HAVE_PTHREAD_H) +#define MW_HAVE_MUTEX 1 +#include +#endif + +/*********************************************************************** +** Defines & other weird stuff +***********************************************************************/ + +/*lint -save -e767 */ +#define VERSION "2.71" /* the current version number */ +#define CHKVAL(mw) (0xFE0180L^(long)mw->count^(long)mw->size^(long)mw->line) +#define FLUSH() mwFlush() +#define TESTS(f,l) if(mwTestAlways) (void)mwTestNow(f,l,1) +#define PRECHK 0x01234567L +#define POSTCHK 0x76543210L +#define mwBUFFER_TO_MW(p) ( (mwData*) (void*) ( ((char*)p)-mwDataSize-mwOverflowZoneSize ) ) +/*lint -restore */ + +#define MW_NML 0x0001 + +#ifdef _MSC_VER +#define COMMIT "c" /* Microsoft C requires the 'c' to perform as desired */ +#else +#define COMMIT "" /* Normal ANSI */ +#endif /* _MSC_VER */ + +#ifdef __cplusplus +#define CPPTEXT "++" +#else +#define CPPTEXT "" +#endif /* __cplusplus */ + +#ifdef MEMWATCH_STDIO +#define mwSTDERR stderr +#else +#define mwSTDERR mwLog +#endif + +#ifdef MW_HAVE_MUTEX +#define MW_MUTEX_INIT() mwMutexInit() +#define MW_MUTEX_TERM() mwMutexTerm() +#define MW_MUTEX_LOCK() mwMutexLock() +#define MW_MUTEX_UNLOCK() mwMutexUnlock() +#else +#define MW_MUTEX_INIT() +#define MW_MUTEX_TERM() +#define MW_MUTEX_LOCK() +#define MW_MUTEX_UNLOCK() +#endif + +/*********************************************************************** +** If you really, really know what you're doing, +** you can predefine these things yourself. +***********************************************************************/ + +#ifndef mwBYTE_DEFINED +# if CHAR_BIT != 8 +# error need CHAR_BIT to be 8! +# else +typedef unsigned char mwBYTE; +# define mwBYTE_DEFINED 1 +# endif +#endif + +#if defined(ULONGLONG_MAX) || defined(ULLONG_MAX) || defined(_UI64_MAX) || defined(ULONG_LONG_MAX) +# define mw64BIT 1 +# define mwROUNDALLOC_DEFAULT 8 +#else +# if UINT_MAX <= 0xFFFFUL +# define mw16BIT 1 +# define mwROUNDALLOC_DEFAULT 2 +# else +# if ULONG_MAX > 0xFFFFFFFFUL +# define mw64BIT 1 +# define mwROUNDALLOC_DEFAULT 8 +# else +# define mw32BIT 1 +# define mwROUNDALLOC_DEFAULT 4 +# endif +# endif +#endif + +/* mwROUNDALLOC is the number of bytes to */ +/* round up to, to ensure that the end of */ +/* the buffer is suitable for storage of */ +/* any kind of object */ +#ifndef mwROUNDALLOC +# define mwROUNDALLOC mwROUNDALLOC_DEFAULT +#endif + +#ifndef mwDWORD_DEFINED +#if ULONG_MAX == 0xFFFFFFFFUL +typedef unsigned long mwDWORD; +#define mwDWORD_DEFINED "unsigned long" +#endif +#endif + +#ifndef mwDWORD_DEFINED +#if UINT_MAX == 0xFFFFFFFFUL +typedef unsigned int mwDWORD; +#define mwDWORD_DEFINED "unsigned int" +#endif +#endif + +#ifndef mwDWORD_DEFINED +#if USHRT_MAX == 0xFFFFFFFFUL +typedef unsigned short mwDWORD; +#define mwDWORD_DEFINED "unsigned short" +#endif +#endif + +#ifndef mwBYTE_DEFINED +#error "can't find out the correct type for a 8 bit scalar" +#endif + +#ifndef mwDWORD_DEFINED +#error "can't find out the correct type for a 32 bit scalar" +#endif + +/*********************************************************************** +** Typedefs & structures +***********************************************************************/ + +/* main data holding area, precedes actual allocation */ +typedef struct mwData_ mwData; +struct mwData_ { + mwData* prev; /* previous allocation in chain */ + mwData* next; /* next allocation in chain */ + const char* file; /* file name where allocated */ + long count; /* action count */ + long check; /* integrity check value */ +#if 0 + long crc; /* data crc value */ +#endif + size_t size; /* size of allocation */ + int line; /* line number where allocated */ + unsigned flag; /* flag word */ + }; + +/* statistics structure */ +typedef struct mwStat_ mwStat; +struct mwStat_ { + mwStat* next; /* next statistic buffer */ + const char* file; + long total; /* total bytes allocated */ + long num; /* total number of allocations */ + long max; /* max allocated at one time */ + long curr; /* current allocations */ + int line; + }; + +/* grabbing structure, 1K in size */ +typedef struct mwGrabData_ mwGrabData; +struct mwGrabData_ { + mwGrabData* next; + int type; + char blob[ 1024 - sizeof(mwGrabData*) - sizeof(int) ]; + }; + +typedef struct mwMarker_ mwMarker; +struct mwMarker_ { + void *host; + char *text; + mwMarker *next; + int level; + }; + +#if defined(WIN32) || defined(__WIN32__) +typedef HANDLE mwMutex; +#endif + +#if defined(MW_PTHREADS) || defined(HAVE_PTHREAD_H) +typedef pthread_mutex_t mwMutex; +#endif + +/*********************************************************************** +** Static variables +***********************************************************************/ + +static int mwInited = 0; +static int mwInfoWritten = 1; +static int mwUseAtexit = 0; +static FILE* mwLog = NULL; +static int mwFlushing = 0; +static int mwStatLevel = MW_STAT_DEFAULT; +static int mwNML = MW_NML_DEFAULT; +static int mwFBI = 0; +static long mwAllocLimit = 0L; +static int mwUseLimit = 0; + +static long mwNumCurAlloc = 0L; +static mwData* mwHead = NULL; +static mwData* mwTail = NULL; +static int mwDataSize = 0; +static unsigned char mwOverflowZoneTemplate[] = "mEmwAtch"; +static int mwOverflowZoneSize = mwROUNDALLOC; + +static void (*mwOutFunction)(int) = NULL; +static int (*mwAriFunction)(const char*) = NULL; +static int mwAriAction = MW_ARI_ABORT; + +static char mwPrintBuf[MW_TRACE_BUFFER+8]; + +static unsigned long mwCounter = 0L; +static long mwErrors = 0L; + +static int mwTestFlags = 0; +static int mwTestAlways = 0; + +static FILE* mwLogB1 = NULL; +static int mwFlushingB1 = 0; + +static mwStat* mwStatList = NULL; +static long mwStatTotAlloc = 0L; +static long mwStatMaxAlloc = 0L; +static long mwStatNumAlloc = 0L; +static long mwStatCurAlloc = 0L; +static long mwNmlNumAlloc = 0L; +static long mwNmlCurAlloc = 0L; + +static mwGrabData* mwGrabList = NULL; +static long mwGrabSize = 0L; + +static void * mwLastFree[MW_FREE_LIST]; +static const char *mwLFfile[MW_FREE_LIST]; +static int mwLFline[MW_FREE_LIST]; +static int mwLFcur = 0; + +static mwMarker* mwFirstMark = NULL; + +static FILE* mwLogB2 = NULL; +static int mwFlushingB2 = 0; + +#ifdef MW_HAVE_MUTEX +static mwMutex mwGlobalMutex; +#endif + +/*********************************************************************** +** Static function declarations +***********************************************************************/ + +static void mwAutoInit( void ); +static FILE* mwLogR( void ); +/* static */ void mwLogW( FILE* ); +static int mwFlushR( void ); +static void mwFlushW( int ); +static void mwFlush( void ); +static void mwIncErr( void ); +static void mwUnlink( mwData*, const char* file, int line ); +static int mwRelink( mwData*, const char* file, int line ); +static int mwIsHeapOK( mwData *mw ); +static int mwIsOwned( mwData* mw, const char* file, int line ); +static int mwTestBuf( mwData* mw, const char* file, int line ); +static void mwDefaultOutFunc( int ); +static void mwWrite( const char* format, ... ); +static void mwLogFile( const char* name ); +static size_t mwFreeUp( size_t, int ); +static const void *mwTestMem( const void *, unsigned, int ); +static int mwStrCmpI( const char *s1, const char *s2 ); +static int mwTestNow( const char *file, int line, int always_invoked ); +static void mwDropAll( void ); +static const char *mwGrabType( int type ); +static unsigned mwGrab_( unsigned kb, int type, int silent ); +static unsigned mwDrop_( unsigned kb, int type, int silent ); +static int mwARI( const char* text ); +static void mwStatReport( void ); +static mwStat* mwStatGet( const char*, int, int ); +static void mwStatAlloc( size_t, const char*, int ); +static void mwStatFree( size_t, const char*, int ); +static int mwCheckOF( const void * p ); +static void mwWriteOF( void * p ); +static char mwDummy( char c ); +#ifdef MW_HAVE_MUTEX +static void mwMutexInit( void ); +static void mwMutexTerm( void ); +static void mwMutexLock( void ); +static void mwMutexUnlock( void ); +#endif + +/*********************************************************************** +** System functions +***********************************************************************/ + +void mwInit( void ) { + time_t tid; + + if( mwInited++ > 0 ) return; + + MW_MUTEX_INIT(); + + /* start a log if none is running */ + if( mwLogR() == NULL ) mwLogFile( "memwatch.log" ); + if( mwLogR() == NULL ) { + int i; + char buf[32]; + /* oops, could not open it! */ + /* probably because it's already open */ + /* so we try some other names */ + for( i=1; i<100; i++ ) { + sprintf( buf, "memwat%02d.log", i ); + mwLogFile( buf ); + if( mwLogR() != NULL ) break; + } + } + + /* initialize the statistics */ + mwStatList = NULL; + mwStatTotAlloc = 0L; + mwStatCurAlloc = 0L; + mwStatMaxAlloc = 0L; + mwStatNumAlloc = 0L; + mwNmlCurAlloc = 0L; + mwNmlNumAlloc = 0L; + + /* calculate the buffer size to use for a mwData */ + mwDataSize = sizeof(mwData); + while( mwDataSize % mwROUNDALLOC ) mwDataSize ++; + + /* write informational header if needed */ + if( !mwInfoWritten ) { + mwInfoWritten = 1; + (void) time( &tid ); + mwWrite( + "\n=============" + " MEMWATCH " VERSION " Copyright (C) 1992-1999 Johan Lindh " + "=============\n"); + mwWrite( "\nStarted at %s\n", ctime( &tid ) ); + +/**************************************************************** Generic */ + mwWrite( "Modes: " ); +#ifdef mwNew + mwWrite( "C++ " ); +#endif /* mwNew */ +#ifdef __STDC__ + mwWrite( "__STDC__ " ); +#endif /* __STDC__ */ +#ifdef mw16BIT + mwWrite( "16-bit " ); +#endif +#ifdef mw32BIT + mwWrite( "32-bit " ); +#endif +#ifdef mw64BIT + mwWrite( "64-bit " ); +#endif + mwWrite( "mwDWORD==(" mwDWORD_DEFINED ")\n" ); + mwWrite( "mwROUNDALLOC==%d sizeof(mwData)==%d mwDataSize==%d\n", + mwROUNDALLOC, sizeof(mwData), mwDataSize ); +/**************************************************************** Generic */ + +/************************************************************ Microsoft C */ +#ifdef _MSC_VER + mwWrite( "Compiled using Microsoft C" CPPTEXT + " %d.%02d\n", _MSC_VER / 100, _MSC_VER % 100 ); +#endif /* _MSC_VER */ +/************************************************************ Microsoft C */ + +/************************************************************** Borland C */ +#ifdef __BORLANDC__ + mwWrite( "Compiled using Borland C" +#ifdef __cplusplus + "++ %d.%01d\n", __BCPLUSPLUS__/0x100, (__BCPLUSPLUS__%0x100)/0x10 ); +#else + " %d.%01d\n", __BORLANDC__/0x100, (__BORLANDC__%0x100)/0x10 ); +#endif /* __cplusplus */ +#endif /* __BORLANDC__ */ +/************************************************************** Borland C */ + +/************************************************************** Watcom C */ +#ifdef __WATCOMC__ + mwWrite( "Compiled using Watcom C %d.%02d ", + __WATCOMC__/100, __WATCOMC__%100 ); +#ifdef __FLAT__ + mwWrite( "(32-bit flat model)" ); +#endif /* __FLAT__ */ + mwWrite( "\n" ); +#endif /* __WATCOMC__ */ +/************************************************************** Watcom C */ + + mwWrite( "\n" ); + FLUSH(); + } + + if( mwUseAtexit ) (void) atexit( mwAbort ); + return; + } + +void mwAbort( void ) { + mwData *mw; + mwMarker *mrk; + char *data; + time_t tid; + int c, i, j; + int errors; + + tid = time( NULL ); + mwWrite( "\nStopped at %s\n", ctime( &tid) ); + + if( !mwInited ) + mwWrite( "internal: mwAbort(): MEMWATCH not initialized!\n" ); + + /* release the grab list */ + mwDropAll(); + + /* report mwMarked items */ + while( mwFirstMark ) { + mrk = mwFirstMark->next; + mwWrite( "mark: %p: %s\n", mwFirstMark->host, mwFirstMark->text ); + free( mwFirstMark->text ); + free( mwFirstMark ); + mwFirstMark = mrk; + mwErrors ++; + } + + /* release all still allocated memory */ + errors = 0; + while( mwHead != NULL && errors < 3 ) { + if( !mwIsOwned(mwHead, __FILE__, __LINE__ ) ) { + if( errors < 3 ) + { + errors ++; + mwWrite( "internal: NML/unfreed scan restarting\n" ); + FLUSH(); + mwHead = mwHead; + continue; + } + mwWrite( "internal: NML/unfreed scan aborted, heap too damaged\n" ); + FLUSH(); + break; + } + mwFlushW(0); + if( !(mwHead->flag & MW_NML) ) { + mwErrors++; + data = ((char*)mwHead)+mwDataSize; + mwWrite( "unfreed: <%ld> %s(%d), %ld bytes at %p ", + mwHead->count, mwHead->file, mwHead->line, (long)mwHead->size, data+mwOverflowZoneSize ); + if( mwCheckOF( data ) ) { + mwWrite( "[underflowed] "); + FLUSH(); + } + if( mwCheckOF( (data+mwOverflowZoneSize+mwHead->size) ) ) { + mwWrite( "[overflowed] "); + FLUSH(); + } + mwWrite( " \t{" ); + j = 16; if( mwHead->size < 16 ) j = (int) mwHead->size; + for( i=0;i<16;i++ ) { + if( i 126 ) c = '.'; + mwWrite( "%c", c ); + } + mwWrite( "}\n" ); + mw = mwHead; + mwUnlink( mw, __FILE__, __LINE__ ); + free( mw ); + } + else { + data = ((char*)mwHead) + mwDataSize + mwOverflowZoneSize; + if( mwTestMem( data, mwHead->size, MW_VAL_NML ) ) { + mwErrors++; + mwWrite( "wild pointer: <%ld> NoMansLand %p alloc'd at %s(%d)\n", + mwHead->count, data + mwOverflowZoneSize, mwHead->file, mwHead->line ); + FLUSH(); + } + mwNmlNumAlloc --; + mwNmlCurAlloc -= mwHead->size; + mw = mwHead; + mwUnlink( mw, __FILE__, __LINE__ ); + free( mw ); + } + } + + if( mwNmlNumAlloc ) mwWrite("internal: NoMansLand block counter %ld, not zero\n", mwNmlNumAlloc ); + if( mwNmlCurAlloc ) mwWrite("internal: NoMansLand byte counter %ld, not zero\n", mwNmlCurAlloc ); + + /* report statistics */ + mwStatReport(); + FLUSH(); + + mwInited = 0; + mwHead = mwTail = NULL; + if( mwErrors ) + fprintf(mwSTDERR,"MEMWATCH detected %ld anomalies\n",mwErrors); + mwLogFile( NULL ); + mwErrors = 0; + + MW_MUTEX_TERM(); + + } + +void mwTerm( void ) { + if( mwInited == 1 ) + { + mwAbort(); + return; + } + if( !mwInited ) + mwWrite("internal: mwTerm(): MEMWATCH has not been started!\n"); + else + mwInited --; + } + +void mwStatistics( int level ) +{ + mwAutoInit(); + if( level<0 ) level=0; + if( mwStatLevel != level ) + { + mwWrite( "statistics: now collecting on a %s basis\n", + level<1?"global":(level<2?"module":"line") ); + mwStatLevel = level; + } +} + +void mwAutoCheck( int onoff ) { + mwAutoInit(); + mwTestAlways = onoff; + if( onoff ) mwTestFlags = MW_TEST_ALL; + } + +void mwSetOutFunc( void (*func)(int) ) { + mwAutoInit(); + mwOutFunction = func; + } + +static void mwWriteOF( void *p ) +{ + int i; + unsigned char *ptr; + ptr = (unsigned char*) p; + for( i=0; inext is not always set? +*/ +void * mwMark( void *p, const char *desc, const char *file, unsigned line ) { + mwMarker *mrk; + unsigned n, isnew; + char *buf; + int tot, oflow = 0; + char wherebuf[128]; + + mwAutoInit(); + TESTS(NULL,0); + + if( desc == NULL ) desc = "unknown"; + if( file == NULL ) file = "unknown"; + + tot = sprintf( wherebuf, "%.48s called from %s(%d)", desc, file, line ); + if( tot >= (int)sizeof(wherebuf) ) { wherebuf[sizeof(wherebuf)-1] = 0; oflow = 1; } + + if( p == NULL ) { + mwWrite("mark: %s(%d), no mark for NULL:'%s' may be set\n", file, line, desc ); + return p; + } + + if( mwFirstMark != NULL && !mwIsReadAddr( mwFirstMark, sizeof( mwMarker ) ) ) + { + mwWrite("mark: %s(%d), mwFirstMark (%p) is trashed, can't mark for %s\n", + file, line, mwFirstMark, desc ); + return p; + } + + for( mrk=mwFirstMark; mrk; mrk=mrk->next ) + { + if( mrk->next != NULL && !mwIsReadAddr( mrk->next, sizeof( mwMarker ) ) ) + { + mwWrite("mark: %s(%d), mark(%p)->next(%p) is trashed, can't mark for %s\n", + file, line, mrk, mrk->next, desc ); + return p; + } + if( mrk->host == p ) break; + } + + if( mrk == NULL ) { + isnew = 1; + mrk = (mwMarker*) malloc( sizeof( mwMarker ) ); + if( mrk == NULL ) { + mwWrite("mark: %s(%d), no mark for %p:'%s', out of memory\n", file, line, p, desc ); + return p; + } + mrk->next = NULL; + n = 0; + } + else { + isnew = 0; + n = strlen( mrk->text ); + } + + n += strlen( wherebuf ); + buf = (char*) malloc( n+3 ); + if( buf == NULL ) { + if( isnew ) free( mrk ); + mwWrite("mark: %s(%d), no mark for %p:'%s', out of memory\n", file, line, p, desc ); + return p; + } + + if( isnew ) { + memcpy( buf, wherebuf, n+1 ); + mrk->next = mwFirstMark; + mrk->host = p; + mrk->text = buf; + mrk->level = 1; + mwFirstMark = mrk; + } + else { + strcpy( buf, mrk->text ); + strcat( buf, ", " ); + strcat( buf, wherebuf ); + free( mrk->text ); + mrk->text = buf; + mrk->level ++; + } + + if( oflow ) { + mwIncErr(); + mwTrace( " [WARNING: OUTPUT BUFFER OVERFLOW - SYSTEM UNSTABLE]\n" ); + } + return p; + } + +void* mwUnmark( void *p, const char *file, unsigned line ) { + mwMarker *mrk, *prv; + mrk = mwFirstMark; + prv = NULL; + while( mrk ) { + if( mrk->host == p ) { + if( mrk->level < 2 ) { + if( prv ) prv->next = mrk->next; + else mwFirstMark = mrk->next; + free( mrk->text ); + free( mrk ); + return p; + } + mrk->level --; + return p; + } + prv = mrk; + mrk = mrk->next; + } + mwWrite("mark: %s(%d), no mark found for %p\n", file, line, p ); + return p; + } + + +/*********************************************************************** +** Abort/Retry/Ignore handlers +***********************************************************************/ + +#pragma GCC diagnostic push +#pragma GCC diagnostic ignored "-Wunused-result" +static int mwARI( const char *estr ) { + char inbuf[81]; + int c; + fprintf(mwSTDERR, "\n%s\nMEMWATCH: Abort, Retry or Ignore? ", estr); + (void) fgets(inbuf,sizeof(inbuf),stdin); + for( c=0; inbuf[c] && inbuf[c] <= ' '; c++ ) ; + c = inbuf[c]; + if( c == 'R' || c == 'r' ) { + mwBreakOut( estr ); + return MW_ARI_RETRY; + } + if( c == 'I' || c == 'i' ) return MW_ARI_IGNORE; + return MW_ARI_ABORT; + } +#pragma GCC diagnostic pop + +/* standard ARI handler (exported) */ +int mwAriHandler( const char *estr ) { + mwAutoInit(); + return mwARI( estr ); + } + +/* used to set the ARI function */ +void mwSetAriFunc( int (*func)(const char *) ) { + mwAutoInit(); + mwAriFunction = func; + } + +/*********************************************************************** +** Allocation handlers +***********************************************************************/ + +void* mwMalloc( size_t size, const char* file, int line) { + size_t needed; + mwData *mw; + char *ptr; + void *p; + + mwAutoInit(); + + MW_MUTEX_LOCK(); + + TESTS(file,line); + + mwCounter ++; + needed = mwDataSize + mwOverflowZoneSize*2 + size; + if( needed < size ) + { + /* theoretical case: req size + mw overhead exceeded size_t limits */ + return NULL; + } + + /* if this allocation would violate the limit, fail it */ + if( mwUseLimit && ((long)size + mwStatCurAlloc > mwAllocLimit) ) { + mwWrite( "limit fail: <%ld> %s(%d), %ld wanted %ld available\n", + mwCounter, file, line, (long)size, mwAllocLimit - mwStatCurAlloc ); + mwIncErr(); + FLUSH(); + MW_MUTEX_UNLOCK(); + return NULL; + } + + mw = (mwData*) malloc( needed ); + if( mw == NULL ) { + if( mwFreeUp(needed,0) >= needed ) { + mw = (mwData*) malloc(needed); + if( mw == NULL ) { + mwWrite( "internal: mwFreeUp(%u) reported success, but malloc() fails\n", needed ); + mwIncErr(); + FLUSH(); + } + } + if( mw == NULL ) { + mwWrite( "fail: <%ld> %s(%d), %ld wanted %ld allocated\n", + mwCounter, file, line, (long)size, mwStatCurAlloc ); + mwIncErr(); + FLUSH(); + MW_MUTEX_UNLOCK(); + return NULL; + } + } + + mw->count = mwCounter; + mw->prev = NULL; + mw->next = mwHead; + mw->file = file; + mw->size = size; + mw->line = line; + mw->flag = 0; + mw->check = CHKVAL(mw); + + if( mwHead ) mwHead->prev = mw; + mwHead = mw; + if( mwTail == NULL ) mwTail = mw; + + ptr = ((char*)mw) + mwDataSize; + mwWriteOF( ptr ); /* '*(long*)ptr = PRECHK;' */ + ptr += mwOverflowZoneSize; + p = ptr; + memset( ptr, MW_VAL_NEW, size ); + ptr += size; + mwWriteOF( ptr ); /* '*(long*)ptr = POSTCHK;' */ + + mwNumCurAlloc ++; + mwStatCurAlloc += (long) size; + mwStatTotAlloc += (long) size; + if( mwStatCurAlloc > mwStatMaxAlloc ) + mwStatMaxAlloc = mwStatCurAlloc; + mwStatNumAlloc ++; + + if( mwStatLevel ) mwStatAlloc( size, file, line ); + + MW_MUTEX_UNLOCK(); + return p; + } + +void* mwRealloc( void *p, size_t size, const char* file, int line) { + int oldUseLimit, i; + mwData *mw; + char *ptr; + + mwAutoInit(); + + if( p == NULL ) return mwMalloc( size, file, line ); + if( size == 0 ) { mwFree( p, file, line ); return NULL; } + + MW_MUTEX_LOCK(); + + /* do the quick ownership test */ + mw = (mwData*) mwBUFFER_TO_MW( p ); + if( mwIsOwned( mw, file, line ) ) { + + /* if the buffer is an NML, treat this as a double-free */ + if( mw->flag & MW_NML ) + { + mwIncErr(); + if( *((unsigned char*)(mw)+mwDataSize+mwOverflowZoneSize) != MW_VAL_NML ) + { + mwWrite( "internal: <%ld> %s(%d), no-mans-land MW-%p is corrupted\n", + mwCounter, file, line, mw ); + } + goto check_dbl_free; + } + + /* if this allocation would violate the limit, fail it */ + if( mwUseLimit && ((long)size + mwStatCurAlloc - (long)mw->size > mwAllocLimit) ) { + TESTS(file,line); + mwCounter ++; + mwWrite( "limit fail: <%ld> %s(%d), %ld wanted %ld available\n", + mwCounter, file, line, (unsigned long)size - mw->size, mwAllocLimit - mwStatCurAlloc ); + mwIncErr(); + FLUSH(); + MW_MUTEX_UNLOCK(); + return NULL; + } + + /* fake realloc operation */ + oldUseLimit = mwUseLimit; + mwUseLimit = 0; + ptr = (char*) mwMalloc( size, file, line ); + if( ptr != NULL ) { + if( size < mw->size ) + memcpy( ptr, p, size ); + else + memcpy( ptr, p, mw->size ); + mwFree( p, file, line ); + } + mwUseLimit = oldUseLimit; + MW_MUTEX_UNLOCK(); + return (void*) ptr; + } + + /* Unknown pointer! */ + + /* using free'd pointer? */ +check_dbl_free: + for(i=0;i %s(%d), %p was" + " freed from %s(%d)\n", + mwCounter, file, line, p, + mwLFfile[i], mwLFline[i] ); + FLUSH(); + MW_MUTEX_UNLOCK(); + return NULL; + } + } + + /* some weird pointer */ + mwIncErr(); + mwWrite( "realloc: <%ld> %s(%d), unknown pointer %p\n", + mwCounter, file, line, p ); + FLUSH(); + MW_MUTEX_UNLOCK(); + return NULL; + } + +char *mwStrdup( const char* str, const char* file, int line ) { + size_t len; + char *newstring; + + MW_MUTEX_LOCK(); + + if( str == NULL ) { + mwIncErr(); + mwWrite( "strdup: <%ld> %s(%d), strdup(NULL) called\n", + mwCounter, file, line ); + FLUSH(); + MW_MUTEX_UNLOCK(); + return NULL; + } + + len = strlen( str ) + 1; + newstring = (char*) mwMalloc( len, file, line ); + if( newstring != NULL ) memcpy( newstring, str, len ); + MW_MUTEX_UNLOCK(); + return newstring; + } + +void mwFree( void* p, const char* file, int line ) { + int i; + mwData* mw; + char buffer[ sizeof(mwData) + (mwROUNDALLOC*3) + 64 ]; + + /* this code is in support of C++ delete */ + if( file == NULL ) { + mwFree_( p ); + MW_MUTEX_UNLOCK(); + return; + } + + mwAutoInit(); + + MW_MUTEX_LOCK(); + TESTS(file,line); + mwCounter ++; + + /* on NULL free, write a warning and return */ + if( p == NULL ) { + mwWrite( "NULL free: <%ld> %s(%d), NULL pointer free'd\n", + mwCounter, file, line ); + FLUSH(); + MW_MUTEX_UNLOCK(); + return; + } + + /* do the quick ownership test */ + mw = (mwData*) mwBUFFER_TO_MW( p ); + + if( mwIsOwned( mw, file, line ) ) { + (void) mwTestBuf( mw, file, line ); + + /* if the buffer is an NML, treat this as a double-free */ + if( mw->flag & MW_NML ) + { + if( *(((unsigned char*)mw)+mwDataSize+mwOverflowZoneSize) != MW_VAL_NML ) + { + mwWrite( "internal: <%ld> %s(%d), no-mans-land MW-%p is corrupted\n", + mwCounter, file, line, mw ); + } + goto check_dbl_free; + } + + /* update the statistics */ + mwNumCurAlloc --; + mwStatCurAlloc -= (long) mw->size; + if( mwStatLevel ) mwStatFree( mw->size, mw->file, mw->line ); + + /* we should either free the allocation or keep it as NML */ + if( mwNML ) { + mw->flag |= MW_NML; + mwNmlNumAlloc ++; + mwNmlCurAlloc += (long) mw->size; + memset( ((char*)mw)+mwDataSize+mwOverflowZoneSize, MW_VAL_NML, mw->size ); + } + else { + /* unlink the allocation, and enter the post-free data */ + mwUnlink( mw, file, line ); + memset( mw, MW_VAL_DEL, + mw->size + mwDataSize+mwOverflowZoneSize+mwOverflowZoneSize ); + if( mwFBI ) { + memset( mw, '.', mwDataSize + mwOverflowZoneSize ); + sprintf( buffer, "FBI<%ld>%s(%d)", mwCounter, file, line ); + strncpy( (char*)(void*)mw, buffer, mwDataSize + mwOverflowZoneSize ); + } + free( mw ); + } + + /* add the pointer to the last-free track */ + mwLFfile[ mwLFcur ] = file; + mwLFline[ mwLFcur ] = line; + mwLastFree[ mwLFcur++ ] = p; + if( mwLFcur == MW_FREE_LIST ) mwLFcur = 0; + + MW_MUTEX_UNLOCK(); + return; + } + + /* check for double-freeing */ +check_dbl_free: + for(i=0;i %s(%d), %p was" + " freed from %s(%d)\n", + mwCounter, file, line, p, + mwLFfile[i], mwLFline[i] ); + FLUSH(); + MW_MUTEX_UNLOCK(); + return; + } + } + + /* some weird pointer... block the free */ + mwIncErr(); + mwWrite( "WILD free: <%ld> %s(%d), unknown pointer %p\n", + mwCounter, file, line, p ); + FLUSH(); + MW_MUTEX_UNLOCK(); + return; + } + +void* mwCalloc( size_t a, size_t b, const char *file, int line ) { + void *p; + size_t size = a * b; + p = mwMalloc( size, file, line ); + if( p == NULL ) return NULL; + memset( p, 0, size ); + return p; + } + +void mwFree_( void *p ) { + MW_MUTEX_LOCK(); + TESTS(NULL,0); + MW_MUTEX_UNLOCK(); + free(p); + } + +void* mwMalloc_( size_t size ) { + MW_MUTEX_LOCK(); + TESTS(NULL,0); + MW_MUTEX_UNLOCK(); + return malloc( size ); + } + +void* mwRealloc_( void *p, size_t size ) { + MW_MUTEX_LOCK(); + TESTS(NULL,0); + MW_MUTEX_UNLOCK(); + return realloc( p, size ); + } + +void* mwCalloc_( size_t a, size_t b ) { + MW_MUTEX_LOCK(); + TESTS(NULL,0); + MW_MUTEX_UNLOCK(); + return calloc( a, b ); + } + +void mwFlushNow( void ) { + if( mwLogR() ) fflush( mwLogR() ); + return; + } + +void mwDoFlush( int onoff ) { + mwFlushW( onoff<1?0:onoff ); + if( onoff ) if( mwLogR() ) fflush( mwLogR() ); + return; + } + +void mwLimit( long lim ) { + TESTS(NULL,0); + mwWrite("limit: old limit = "); + if( !mwAllocLimit ) mwWrite( "none" ); + else mwWrite( "%ld bytes", mwAllocLimit ); + mwWrite( ", new limit = "); + if( !lim ) { + mwWrite( "none\n" ); + mwUseLimit = 0; + } + else { + mwWrite( "%ld bytes\n", lim ); + mwUseLimit = 1; + } + mwAllocLimit = lim; + FLUSH(); + } + +void mwSetAriAction( int action ) { + MW_MUTEX_LOCK(); + TESTS(NULL,0); + mwAriAction = action; + MW_MUTEX_UNLOCK(); + return; + } + +int mwAssert( int exp, const char *exps, const char *fn, int ln ) { + int i; + char buffer[MW_TRACE_BUFFER+8]; + if( exp ) { + return 0; + } + mwAutoInit(); + MW_MUTEX_LOCK(); + TESTS(fn,ln); + mwIncErr(); + mwCounter++; + mwWrite( "assert trap: <%ld> %s(%d), %s\n", mwCounter, fn, ln, exps ); + if( mwAriFunction != NULL ) { + sprintf( buffer, "MEMWATCH: assert trap: %s(%d), %s", fn, ln, exps ); + i = (*mwAriFunction)(buffer); + switch( i ) { + case MW_ARI_IGNORE: + mwWrite( "assert trap: <%ld> IGNORED - execution continues\n", mwCounter ); + MW_MUTEX_UNLOCK(); + return 0; + case MW_ARI_RETRY: + mwWrite( "assert trap: <%ld> RETRY - executing again\n", mwCounter ); + MW_MUTEX_UNLOCK(); + return 1; + } + } + else { + if( mwAriAction & MW_ARI_IGNORE ) { + mwWrite( "assert trap: <%ld> AUTO IGNORED - execution continues\n", mwCounter ); + MW_MUTEX_UNLOCK(); + return 0; + } + fprintf(mwSTDERR,"\nMEMWATCH: assert trap: %s(%d), %s\n", fn, ln, exps ); + } + + FLUSH(); + (void) mwTestNow( fn, ln, 1 ); + FLUSH(); + + if( mwAriAction & MW_ARI_NULLREAD ) { + /* This is made in an attempt to kick in */ + /* any debuggers or OS stack traces */ + FLUSH(); + /*lint -save -e413 */ + i = *((int*)NULL); + mwDummy( (char)i ); + /*lint -restore */ + } + + MW_MUTEX_UNLOCK(); + exit(255); + /* NOT REACHED - the return statement is in to keep */ + /* stupid compilers from squeaking about differing return modes. */ + /* Smart compilers instead say 'code unreachable...' */ + /*lint -save -e527 */ + return 0; + /*lint -restore */ + } + +int mwVerify( int exp, const char *exps, const char *fn, int ln ) { + int i; + char buffer[MW_TRACE_BUFFER+8]; + if( exp ) { + return 0; + } + mwAutoInit(); + MW_MUTEX_LOCK(); + TESTS(fn,ln); + mwIncErr(); + mwCounter++; + mwWrite( "verify trap: <%ld> %s(%d), %s\n", mwCounter, fn, ln, exps ); + if( mwAriFunction != NULL ) { + sprintf( buffer, "MEMWATCH: verify trap: %s(%d), %s", fn, ln, exps ); + i = (*mwAriFunction)(buffer); + if( i == 0 ) { + mwWrite( "verify trap: <%ld> IGNORED - execution continues\n", mwCounter ); + MW_MUTEX_UNLOCK(); + return 0; + } + if( i == 1 ) { + mwWrite( "verify trap: <%ld> RETRY - executing again\n", mwCounter ); + MW_MUTEX_UNLOCK(); + return 1; + } + } + else { + if( mwAriAction & MW_ARI_NULLREAD ) { + /* This is made in an attempt to kick in */ + /* any debuggers or OS stack traces */ + FLUSH(); + /*lint -save -e413 */ + i = *((int*)NULL); + mwDummy( (char)i ); + /*lint -restore */ + } + if( mwAriAction & MW_ARI_IGNORE ) { + mwWrite( "verify trap: <%ld> AUTO IGNORED - execution continues\n", mwCounter ); + MW_MUTEX_UNLOCK(); + return 0; + } + fprintf(mwSTDERR,"\nMEMWATCH: verify trap: %s(%d), %s\n", fn, ln, exps ); + } + FLUSH(); + (void) mwTestNow( fn, ln, 1 ); + FLUSH(); + MW_MUTEX_UNLOCK(); + exit(255); + /* NOT REACHED - the return statement is in to keep */ + /* stupid compilers from squeaking about differing return modes. */ + /* Smart compilers instead say 'code unreachable...' */ + /*lint -save -e527 */ + return 0; + /*lint -restore */ + } + +void mwTrace( const char *format, ... ) { + int tot, oflow = 0; + va_list mark; + + mwAutoInit(); + MW_MUTEX_LOCK(); + TESTS(NULL,0); + if( mwOutFunction == NULL ) mwOutFunction = mwDefaultOutFunc; + + va_start( mark, format ); + tot = vsprintf( mwPrintBuf, format, mark ); + va_end( mark ); + if( tot >= MW_TRACE_BUFFER ) { mwPrintBuf[MW_TRACE_BUFFER] = 0; oflow = 1; } + for(tot=0;mwPrintBuf[tot];tot++) + (*mwOutFunction)( mwPrintBuf[tot] ); + if( oflow ) { + mwIncErr(); + mwTrace( " [WARNING: OUTPUT BUFFER OVERFLOW - SYSTEM UNSTABLE]\n" ); + } + + FLUSH(); + MW_MUTEX_UNLOCK(); + } + + +/*********************************************************************** +** Grab & Drop +***********************************************************************/ + +unsigned mwGrab( unsigned kb ) { + TESTS(NULL,0); + return mwGrab_( kb, MW_VAL_GRB, 0 ); + } + +unsigned mwDrop( unsigned kb ) { + TESTS(NULL,0); + return mwDrop_( kb, MW_VAL_GRB, 0 ); + } + +static void mwDropAll() { + TESTS(NULL,0); + (void) mwDrop_( 0, MW_VAL_GRB, 0 ); + (void) mwDrop_( 0, MW_VAL_NML, 0 ); + if( mwGrabList != NULL ) + mwWrite( "internal: the grab list is not empty after mwDropAll()\n"); + } + +static const char *mwGrabType( int type ) { + switch( type ) { + case MW_VAL_GRB: + return "grabbed"; + case MW_VAL_NML: + return "no-mans-land"; + default: + /* do nothing */ + ; + } + return ""; + } + +static unsigned mwGrab_( unsigned kb, int type, int silent ) { + unsigned i = kb; + mwGrabData *gd; + if( !kb ) i = kb = 65000U; + + for(;kb;kb--) { + if( mwUseLimit && + (mwStatCurAlloc + mwGrabSize + (long)sizeof(mwGrabData) > mwAllocLimit) ) { + if( !silent ) { + mwWrite("grabbed: all allowed memory to %s (%u kb)\n", + mwGrabType(type), i-kb); + FLUSH(); + } + return i-kb; + } + gd = (mwGrabData*) malloc( sizeof(mwGrabData) ); + if( gd == NULL ) { + if( !silent ) { + mwWrite("grabbed: all available memory to %s (%u kb)\n", + mwGrabType(type), i-kb); + FLUSH(); + } + return i-kb; + } + mwGrabSize += (long) sizeof(mwGrabData); + gd->next = mwGrabList; + memset( gd->blob, type, sizeof(gd->blob) ); + gd->type = type; + mwGrabList = gd; + } + if( !silent ) { + mwWrite("grabbed: %u kilobytes of %s memory\n", i, mwGrabType(type) ); + FLUSH(); + } + return i; + } + +static unsigned mwDrop_( unsigned kb, int type, int silent ) { + unsigned i = kb; + mwGrabData *gd,*tmp,*pr; + const void *p; + + if( mwGrabList == NULL && kb == 0 ) return 0; + if( !kb ) i = kb = 60000U; + + pr = NULL; + gd = mwGrabList; + for(;kb;) { + if( gd == NULL ) { + if( i-kb > 0 && !silent ) { + mwWrite("dropped: all %s memory (%u kb)\n", mwGrabType(type), i-kb); + FLUSH(); + } + return i-kb; + } + if( gd->type == type ) { + if( pr ) pr->next = gd->next; + kb --; + tmp = gd; + if( mwGrabList == gd ) mwGrabList = gd->next; + gd = gd->next; + p = mwTestMem( tmp->blob, sizeof( tmp->blob ), type ); + if( p != NULL ) { + mwWrite( "wild pointer: <%ld> %s memory hit at %p\n", + mwCounter, mwGrabType(type), p ); + FLUSH(); + } + mwGrabSize -= (long) sizeof(mwGrabData); + free( tmp ); + } + else { + pr = gd; + gd = gd->next; + } + } + if( !silent ) { + mwWrite("dropped: %u kilobytes of %s memory\n", i, mwGrabType(type) ); + FLUSH(); + } + return i; + } + +/*********************************************************************** +** No-Mans-Land +***********************************************************************/ + +void mwNoMansLand( int level ) { + mwAutoInit(); + TESTS(NULL,0); + switch( level ) { + case MW_NML_NONE: + (void) mwDrop_( 0, MW_VAL_NML, 0 ); + break; + case MW_NML_FREE: + break; + case MW_NML_ALL: + (void) mwGrab_( 0, MW_VAL_NML, 0 ); + break; + default: + return; + } + mwNML = level; + } + +/*********************************************************************** +** Static functions +***********************************************************************/ + +static void mwAutoInit( void ) +{ + if( mwInited ) return; + mwUseAtexit = 1; + mwInit(); + return; +} + +static FILE *mwLogR() { + if( (mwLog == mwLogB1) && (mwLog == mwLogB2) ) return mwLog; + if( mwLog == mwLogB1 ) mwLogB2 = mwLog; + if( mwLog == mwLogB2 ) mwLogB1 = mwLog; + if( mwLogB1 == mwLogB2 ) mwLog = mwLogB1; + if( (mwLog == mwLogB1) && (mwLog == mwLogB2) ) { + mwWrite("internal: log file handle damaged and recovered\n"); + FLUSH(); + return mwLog; + } + fprintf(mwSTDERR,"\nMEMWATCH: log file handle destroyed, using mwSTDERR\n" ); + mwLog = mwLogB1 = mwLogB2 = mwSTDERR; + return mwSTDERR; + } + +/* static */ void mwLogW( FILE *p ) { + mwLog = mwLogB1 = mwLogB2 = p; + } + +static int mwFlushR() { + if( (mwFlushing == mwFlushingB1) && (mwFlushing == mwFlushingB2) ) return mwFlushing; + if( mwFlushing == mwFlushingB1 ) mwFlushingB2 = mwFlushing; + if( mwFlushing == mwFlushingB2 ) mwFlushingB1 = mwFlushing; + if( mwFlushingB1 == mwFlushingB2 ) mwFlushing = mwFlushingB1; + if( (mwFlushing == mwFlushingB1) && (mwFlushing == mwFlushingB2) ) { + mwWrite("internal: flushing flag damaged and recovered\n"); + FLUSH(); + return mwFlushing; + } + mwWrite("internal: flushing flag destroyed, so set to true\n"); + mwFlushing = mwFlushingB1 = mwFlushingB2 = 1; + return 1; + } + +static void mwFlushW( int n ) { + mwFlushing = mwFlushingB1 = mwFlushingB2 = n; + } + +static void mwIncErr() { + mwErrors++; + mwFlushW( mwFlushR()+1 ); + FLUSH(); + } + +static void mwFlush() { + if( mwLogR() == NULL ) return; +#ifdef MW_FLUSH + fflush( mwLogR() ); +#else + if( mwFlushR() ) fflush( mwLogR() ); +#endif + return; + } + +static void mwUnlink( mwData* mw, const char* file, int line ) { + if( mw->prev == NULL ) { + if( mwHead != mw ) + mwWrite( "internal: <%ld> %s(%d), MW-%p: link1 NULL, but not head\n", + mwCounter, file, line, mw ); + mwHead = mw->next; + } + else { + if( mw->prev->next != mw ) + mwWrite( "internal: <%ld> %s(%d), MW-%p: link1 failure\n", + mwCounter, file, line, mw ); + else mw->prev->next = mw->next; + } + if( mw->next == NULL ) { + if( mwTail != mw ) + mwWrite( "internal: <%ld> %s(%d), MW-%p: link2 NULL, but not tail\n", + mwCounter, file, line, mw ); + mwTail = mw->prev; + } + else { + if( mw->next->prev != mw ) + mwWrite( "internal: <%ld> %s(%d), MW-%p: link2 failure\n", + mwCounter, file, line, mw ); + else mw->next->prev = mw->prev; + } + } + +/* +** Relinking tries to repair a damaged mw block. +** Returns nonzero if it thinks it successfully +** repaired the heap chain. +*/ +static int mwRelink( mwData* mw, const char* file, int line ) { + int fails; + mwData *mw1, *mw2; + long count, size; + mwStat *ms; + + if( file == NULL ) file = "unknown"; + + if( mw == NULL ) { + mwWrite("relink: cannot repair MW at NULL\n"); + FLUSH(); + goto emergency; + } + + if( !mwIsSafeAddr(mw, mwDataSize) ) { + mwWrite("relink: MW-%p is a garbage pointer\n", mw); + FLUSH(); + goto emergency; + } + + mwWrite("relink: <%ld> %s(%d) attempting to repair MW-%p...\n", mwCounter, file, line, mw ); + FLUSH(); + fails = 0; + + /* Repair from head */ + if( mwHead != mw ) { + if( !mwIsSafeAddr( mwHead, mwDataSize ) ) { + mwWrite("relink: failed for MW-%p; head pointer destroyed\n", mw ); + FLUSH(); + goto emergency; + } + for( mw1=mwHead; mw1; mw1=mw1->next ) { + if( mw1->next == mw ) { + mw->prev = mw1; + break; + } + if( mw1->next && + ( !mwIsSafeAddr(mw1->next, mwDataSize ) || mw1->next->prev != mw1) ) { + mwWrite("relink: failed for MW-%p; forward chain fragmented at MW-%p: 'next' is %p\n", mw, mw1, mw1->next ); + FLUSH(); + goto emergency; + } + } + if( mw1 == NULL ) { + mwWrite("relink: MW-%p not found in forward chain search\n", mw ); + FLUSH(); + fails ++; + } + } + else + { + mwWrite( "relink: MW-%p is the head (first) allocation\n", mw ); + if( mw->prev != NULL ) + { + mwWrite( "relink: MW-%p prev pointer is non-NULL, you have a wild pointer\n", mw ); + mw->prev = NULL; + } + } + + /* Repair from tail */ + if( mwTail != mw ) { + if( !mwIsSafeAddr( mwTail, mwDataSize ) ) { + mwWrite("relink: failed for MW-%p; tail pointer destroyed\n", mw ); + FLUSH(); + goto emergency; + } + for( mw1=mwTail; mw1; mw1=mw1->prev ) { + if( mw1->prev == mw ) { + mw->next = mw1; + break; + } + if( mw1->prev && (!mwIsSafeAddr(mw1->prev, mwDataSize ) || mw1->prev->next != mw1) ) { + mwWrite("relink: failed for MW-%p; reverse chain fragmented at MW-%p, 'prev' is %p\n", mw, mw1, mw1->prev ); + FLUSH(); + goto emergency; + } + } + if( mw1 == NULL ) { + mwWrite("relink: MW-%p not found in reverse chain search\n", mw ); + FLUSH(); + fails ++; + } + } + else + { + mwWrite( "relink: MW-%p is the tail (last) allocation\n", mw ); + if( mw->next != NULL ) + { + mwWrite( "relink: MW-%p next pointer is non-NULL, you have a wild pointer\n", mw ); + mw->next = NULL; + } + } + + if( fails > 1 ) { + mwWrite("relink: heap appears intact, MW-%p probably garbage pointer\n", mw ); + FLUSH(); + goto verifyok; + } + + /* restore MW info where possible */ + if( mwIsReadAddr( mw->file, 1 ) ) { + ms = mwStatGet( mw->file, -1, 0 ); + if( ms == NULL ) mw->file = ""; + } + mw->check = CHKVAL(mw); + goto verifyok; + + /* Emergency repair */ + emergency: + + if( mwHead == NULL && mwTail == NULL ) + { + if( mwStatCurAlloc == 0 ) + mwWrite("relink: <%ld> %s(%d) heap is empty, nothing to repair\n", mwCounter, file, line ); + else + mwWrite("relink: <%ld> %s(%d) heap damaged beyond repair\n", mwCounter, file, line ); + FLUSH(); + return 0; + } + + mwWrite("relink: <%ld> %s(%d) attempting emergency repairs...\n", mwCounter, file, line ); + FLUSH(); + + if( mwHead == NULL || mwTail == NULL ) + { + if( mwHead == NULL ) mwWrite("relink: mwHead is NULL, but mwTail is %p\n", mwTail ); + else mwWrite("relink: mwTail is NULL, but mwHead is %p\n", mwHead ); + } + + mw1=NULL; + if( mwHead != NULL ) + { + if( !mwIsReadAddr( mwHead, mwDataSize ) || mwHead->check != CHKVAL(mwHead) ) + { + mwWrite("relink: mwHead (MW-%p) is damaged, skipping forward scan\n", mwHead ); + mwHead = NULL; + goto scan_reverse; + } + if( mwHead->prev != NULL ) + { + mwWrite("relink: the mwHead pointer's 'prev' member is %p, not NULL\n", mwHead->prev ); + } + for( mw1=mwHead; mw1; mw1=mw1->next ) + { + if( mw1->next ) + { + if( !mwIsReadAddr(mw1->next,mwDataSize) || + (!mw1->next->check) != CHKVAL(mw1) || + mw1->next->prev != mw1 ) + { + mwWrite("relink: forward chain's last intact MW is MW-%p, %ld %sbytes at %s(%d)\n", + mw1, mw1->size, (mw->flag & MW_NML)?"NoMansLand ":"", mw1->file, mw1->line ); + if( mwIsReadAddr(mw1->next,mwDataSize ) ) + { + mwWrite("relink: forward chain's first damaged MW is MW-%p, %ld %sbytes at %s(%d)\n", + mw1->next, mw1->size, (mw->flag & MW_NML)?"NoMansLand ":"", + mwIsReadAddr(mw1->file,16)?mw1->file:"", mw1->line ); + } + else + { + mwWrite("relink: the 'next' pointer of this MW points to %p, which is out-of-legal-access\n", + mw1->next ); + } + break; + } + } + } + } + + +scan_reverse: + mw2=NULL; + if( mwTail != NULL ) + { + if( !mwIsReadAddr(mwTail,mwDataSize) || mwTail->check != CHKVAL(mwTail) ) + { + mwWrite("relink: mwTail (%p) is damaged, skipping reverse scan\n", mwTail ); + mwTail = NULL; + goto analyze; + } + if( mwTail->next != NULL ) + { + mwWrite("relink: the mwTail pointer's 'next' member is %p, not NULL\n", mwTail->next ); + } + for( mw2=mwTail; mw2; mw2=mw2->prev ) + { + if( mw2->prev ) + { + if( !mwIsReadAddr(mw2->prev,mwDataSize) || + (!mw2->prev->check) != CHKVAL(mw2) || + mw2->prev->next != mw2 ) + { + mwWrite("relink: reverse chain's last intact MW is MW-%p, %ld %sbytes at %s(%d)\n", + mw2, mw2->size, (mw->flag & MW_NML)?"NoMansLand ":"", mw2->file, mw2->line ); + if( mwIsReadAddr(mw2->prev,mwDataSize ) ) + { + mwWrite("relink: reverse chain's first damaged MW is MW-%p, %ld %sbytes at %s(%d)\n", + mw2->prev, mw2->size, (mw->flag & MW_NML)?"NoMansLand ":"", + mwIsReadAddr(mw2->file,16)?mw2->file:"", mw2->line ); + } + else + { + mwWrite("relink: the 'prev' pointer of this MW points to %p, which is out-of-legal-access\n", + mw2->prev ); + } + break; + } + } + } + } + +analyze: + if( mwHead == NULL && mwTail == NULL ) + { + mwWrite("relink: both head and tail pointers damaged, aborting program\n"); + mwFlushW(1); + FLUSH(); + abort(); + } + if( mwHead == NULL ) + { + mwHead = mw2; + mwWrite("relink: heap truncated, MW-%p designated as new mwHead\n", mw2 ); + mw2->prev = NULL; + mw1 = mw2 = NULL; + } + if( mwTail == NULL ) + { + mwTail = mw1; + mwWrite("relink: heap truncated, MW-%p designated as new mwTail\n", mw1 ); + mw1->next = NULL; + mw1 = mw2 = NULL; + } + if( mw1 == NULL && mw2 == NULL && + mwHead->prev == NULL && mwTail->next == NULL ) { + mwWrite("relink: verifying heap integrity...\n" ); + FLUSH(); + goto verifyok; + } + if( mw1 && mw2 && mw1 != mw2 ) { + mw1->next = mw2; + mw2->prev = mw1; + mwWrite("relink: emergency repairs successful, assessing damage...\n"); + FLUSH(); + } + else { + mwWrite("relink: heap totally destroyed, aborting program\n"); + mwFlushW(1); + FLUSH(); + abort(); + } + + /* Verify by checking that the number of active allocations */ + /* match the number of entries in the chain */ +verifyok: + if( !mwIsHeapOK( NULL ) ) { + mwWrite("relink: heap verification FAILS - aborting program\n"); + mwFlushW(1); + FLUSH(); + abort(); + } + for( size=count=0, mw1=mwHead; mw1; mw1=mw1->next ) { + count ++; + size += (long) mw1->size; + } + if( count == mwNumCurAlloc ) { + mwWrite("relink: successful, "); + if( size == mwStatCurAlloc ) { + mwWrite("no allocations lost\n"); + } + else { + if( mw != NULL ) { + mwWrite("size information lost for MW-%p\n", mw); + mw->size = 0; + } + } + } + else { + mwWrite("relink: partial, %ld MW-blocks of %ld bytes lost\n", + mwNmlNumAlloc+mwNumCurAlloc-count, mwNmlCurAlloc+mwStatCurAlloc-size ); + return 0; + } + + return 1; + } + +/* +** If mwData* is NULL: +** Returns 0 if heap chain is broken. +** Returns 1 if heap chain is intact. +** If mwData* is not NULL: +** Returns 0 if mwData* is missing or if chain is broken. +** Returns 1 if chain is intact and mwData* is found. +*/ +static int mwIsHeapOK( mwData *includes_mw ) { + int found = 0; + mwData *mw; + + for( mw = mwHead; mw; mw=mw->next ) { + if( includes_mw == mw ) found++; + if( !mwIsSafeAddr( mw, mwDataSize ) ) return 0; + if( mw->prev ) { + if( !mwIsSafeAddr( mw->prev, mwDataSize ) ) return 0; + if( mw==mwHead || mw->prev->next != mw ) return 0; + } + if( mw->next ) { + if( !mwIsSafeAddr( mw->next, mwDataSize ) ) return 0; + if( mw==mwTail || mw->next->prev != mw ) return 0; + } + else if( mw!=mwTail ) return 0; + } + + if( includes_mw != NULL && !found ) return 0; + + return 1; + } + +static int mwIsOwned( mwData* mw, const char *file, int line ) { + int retv; + mwStat *ms; + + /* see if the address is legal according to OS */ + if( !mwIsSafeAddr( mw, mwDataSize ) ) return 0; + + /* make sure we have _anything_ allocated */ + if( mwHead == NULL && mwTail == NULL && mwStatCurAlloc == 0 ) + return 0; + + /* calculate checksum */ + if( mw->check != CHKVAL(mw) ) { + /* may be damaged checksum, see if block is in heap */ + if( mwIsHeapOK( mw ) ) { + /* damaged checksum, repair it */ + mwWrite( "internal: <%ld> %s(%d), checksum for MW-%p is incorrect\n", + mwCounter, file, line, mw ); + mwIncErr(); + if( mwIsReadAddr( mw->file, 1 ) ) { + ms = mwStatGet( mw->file, -1, 0 ); + if( ms == NULL ) mw->file = ""; + } + else mw->file = ""; + mw->size = 0; + mw->check = CHKVAL(mw); + return 1; + } + /* no, it's just some garbage data */ + return 0; + } + + /* check that the non-NULL pointers are safe */ + if( mw->prev && !mwIsSafeAddr( mw->prev, mwDataSize ) ) mwRelink( mw, file, line ); + if( mw->next && !mwIsSafeAddr( mw->next, mwDataSize ) ) mwRelink( mw, file, line ); + + /* safe address, checksum OK, proceed with heap checks */ + + /* see if the block is in the heap */ + retv = 0; + if( mw->prev ) { if( mw->prev->next == mw ) retv ++; } + else { if( mwHead == mw ) retv++; } + if( mw->next ) { if( mw->next->prev == mw ) retv ++; } + else { if( mwTail == mw ) retv++; } + if( mw->check == CHKVAL(mw) ) retv ++; + if( retv > 2 ) return 1; + + /* block not in heap, check heap for corruption */ + + if( !mwIsHeapOK( mw ) ) { + if( mwRelink( mw, file, line ) ) + return 1; + } + + /* unable to repair */ + mwWrite( "internal: <%ld> %s(%d), mwIsOwned fails for MW-%p\n", + mwCounter, file, line, mw ); + mwIncErr(); + + return 0; + } + +/* +** mwTestBuf: +** Checks a buffers links and pre/postfixes. +** Writes errors found to the log. +** Returns zero if no errors found. +*/ +static int mwTestBuf( mwData* mw, const char* file, int line ) { + int retv = 0; + char *p; + + if( file == NULL ) file = "unknown"; + + if( !mwIsSafeAddr( mw, mwDataSize + mwOverflowZoneSize ) ) { + mwWrite( "internal: <%ld> %s(%d): pointer MW-%p is invalid\n", + mwCounter, file, line, mw ); + mwIncErr(); + return 2; + } + + if( mw->check != CHKVAL(mw) ) { + mwWrite( "internal: <%ld> %s(%d), info trashed; relinking\n", + mwCounter, file, line ); + mwIncErr(); + if( !mwRelink( mw, file, line ) ) return 2; + } + + if( mw->prev && mw->prev->next != mw ) { + mwWrite( "internal: <%ld> %s(%d), buffer <%ld> %s(%d) link1 broken\n", + mwCounter,file,line, (long)mw->size, mw->count, mw->file, mw->line ); + mwIncErr(); + if( !mwRelink( mw, file, line ) ) retv = 2; + } + if( mw->next && mw->next->prev != mw ) { + mwWrite( "internal: <%ld> %s(%d), buffer <%ld> %s(%d) link2 broken\n", + mwCounter,file,line, (long)mw->size, mw->count, mw->file, mw->line ); + mwIncErr(); + if( !mwRelink( mw, file, line ) ) retv = 2; + } + + p = ((char*)mw) + mwDataSize; + if( mwCheckOF( p ) ) { + mwWrite( "underflow: <%ld> %s(%d), %ld bytes alloc'd at <%ld> %s(%d)\n", + mwCounter,file,line, (long)mw->size, mw->count, mw->file, mw->line ); + mwIncErr(); + retv = 1; + } + p += mwOverflowZoneSize + mw->size; + if( mwIsReadAddr( p, mwOverflowZoneSize ) && mwCheckOF( p ) ) { + mwWrite( "overflow: <%ld> %s(%d), %ld bytes alloc'd at <%ld> %s(%d)\n", + mwCounter,file,line, (long)mw->size, mw->count, mw->file, mw->line ); + mwIncErr(); + retv = 1; + } + + return retv; + } + +static void mwDefaultOutFunc( int c ) { + if( mwLogR() ) fputc( c, mwLogR() ); + } + +static void mwWrite( const char *format, ... ) { + int tot, oflow = 0; + va_list mark; + mwAutoInit(); + if( mwOutFunction == NULL ) mwOutFunction = mwDefaultOutFunc; + va_start( mark, format ); + tot = vsprintf( mwPrintBuf, format, mark ); + va_end( mark ); + if( tot >= MW_TRACE_BUFFER ) { mwPrintBuf[MW_TRACE_BUFFER] = 0; oflow = 1; } + for(tot=0;mwPrintBuf[tot];tot++) + (*mwOutFunction)( mwPrintBuf[tot] ); + if( oflow ) { + mwWrite( "\ninternal: mwWrite(): WARNING! OUTPUT EXCEEDED %u CHARS: SYSTEM UNSTABLE\n", MW_TRACE_BUFFER-1 ); + FLUSH(); + } + return; + } + +static void mwLogFile( const char *name ) { + time_t tid; + (void) time( &tid ); + if( mwLogR() != NULL ) { + fclose( mwLogR() ); + mwLogW( NULL ); + } + if( name == NULL ) return; + mwLogW( fopen( name, "a" COMMIT ) ); + if( mwLogR() == NULL ) + mwWrite( "logfile: failed to open/create file '%s'\n", name ); + } + +/* +** Try to free NML memory until a contiguous allocation of +** 'needed' bytes can be satisfied. If this is not enough +** and the 'urgent' parameter is nonzero, grabbed memory is +** also freed. +*/ +static size_t mwFreeUp( size_t needed, int urgent ) { + void *p; + mwData *mw, *mw2; + char *data; + + /* free grabbed NML memory */ + for(;;) { + if( mwDrop_( 1, MW_VAL_NML, 1 ) == 0 ) break; + p = malloc( needed ); + if( p == NULL ) continue; + free( p ); + return needed; + } + + /* free normal NML memory */ + mw = mwHead; + while( mw != NULL ) { + if( !(mw->flag & MW_NML) ) mw = mw->next; + else { + data = ((char*)mw)+mwDataSize+mwOverflowZoneSize; + if( mwTestMem( data, mw->size, MW_VAL_NML ) ) { + mwIncErr(); + mwWrite( "wild pointer: <%ld> NoMansLand %p alloc'd at %s(%d)\n", + mw->count, data + mwOverflowZoneSize, mw->file, mw->line ); + } + mw2 = mw->next; + mwUnlink( mw, "mwFreeUp", 0 ); + free( mw ); + mw = mw2; + p = malloc( needed ); + if( p == NULL ) continue; + free( p ); + return needed; + } + } + + /* if not urgent (for internal purposes), fail */ + if( !urgent ) return 0; + + /* free grabbed memory */ + for(;;) { + if( mwDrop_( 1, MW_VAL_GRB, 1 ) == 0 ) break; + p = malloc( needed ); + if( p == NULL ) continue; + free( p ); + return needed; + } + + return 0; + } + +static const void * mwTestMem( const void *p, unsigned len, int c ) { + const unsigned char *ptr; + ptr = (const unsigned char *) p; + while( len-- ) { + if( *ptr != (unsigned char)c ) return (const void*)ptr; + ptr ++; + } + return NULL; + } + +static int mwStrCmpI( const char *s1, const char *s2 ) { + if( s1 == NULL || s2 == NULL ) return 0; + while( *s1 ) { + if( toupper(*s2) == toupper(*s1) ) { s1++; s2++; continue; } + return 1; + } + return 0; + } + +#define AIPH() if( always_invoked ) { mwWrite("autocheck: <%ld> %s(%d) ", mwCounter, file, line ); always_invoked = 0; } + +static int mwTestNow( const char *file, int line, int always_invoked ) { + int retv = 0; + mwData *mw; + char *data; + + if( file && !always_invoked ) + mwWrite("check: <%ld> %s(%d), checking %s%s%s\n", + mwCounter, file, line, + (mwTestFlags & MW_TEST_CHAIN) ? "chain ": "", + (mwTestFlags & MW_TEST_ALLOC) ? "alloc ": "", + (mwTestFlags & MW_TEST_NML) ? "nomansland ": "" + ); + + if( mwTestFlags & MW_TEST_CHAIN ) { + for( mw = mwHead; mw; mw=mw->next ) { + if( !mwIsSafeAddr(mw, mwDataSize) ) { + AIPH(); + mwWrite("check: heap corruption detected\n"); + mwIncErr(); + return retv + 1; + } + if( mw->prev ) { + if( !mwIsSafeAddr(mw->prev, mwDataSize) ) { + AIPH(); + mwWrite("check: heap corruption detected\n"); + mwIncErr(); + return retv + 1; + } + if( mw==mwHead || mw->prev->next != mw ) { + AIPH(); + mwWrite("check: heap chain broken, prev link incorrect\n"); + mwIncErr(); + retv ++; + } + } + if( mw->next ) { + if( !mwIsSafeAddr(mw->next, mwDataSize) ) { + AIPH(); + mwWrite("check: heap corruption detected\n"); + mwIncErr(); + return retv + 1; + } + if( mw==mwTail || mw->next->prev != mw ) { + AIPH(); + mwWrite("check: heap chain broken, next link incorrect\n"); + mwIncErr(); + retv ++; + } + } + else if( mw!=mwTail ) { + AIPH(); + mwWrite("check: heap chain broken, tail incorrect\n"); + mwIncErr(); + retv ++; + } + } + } + if( mwTestFlags & MW_TEST_ALLOC ) { + for( mw = mwHead; mw; mw=mw->next ) { + if( mwTestBuf( mw, file, line ) ) retv ++; + } + } + if( mwTestFlags & MW_TEST_NML ) { + for( mw = mwHead; mw; mw=mw->next ) { + if( (mw->flag & MW_NML) ) { + data = ((char*)mw)+mwDataSize+mwOverflowZoneSize; + if( mwTestMem( data, mw->size, MW_VAL_NML ) ) { + mwIncErr(); + mwWrite( "wild pointer: <%ld> NoMansLand %p alloc'd at %s(%d)\n", + mw->count, data + mwOverflowZoneSize, mw->file, mw->line ); + } + } + } + } + + + if( file && !always_invoked && !retv ) + mwWrite("check: <%ld> %s(%d), complete; no errors\n", + mwCounter, file, line ); + return retv; + } + +/********************************************************************** +** Statistics +**********************************************************************/ + +static void mwStatReport() +{ + mwStat* ms, *ms2; + const char *modname; + int modnamelen; + + /* global statistics report */ + mwWrite( "\nMemory usage statistics (global):\n" ); + mwWrite( " N)umber of allocations made: %ld\n", mwStatNumAlloc ); + mwWrite( " L)argest memory usage : %ld\n", mwStatMaxAlloc ); + mwWrite( " T)otal of all alloc() calls: %ld\n", mwStatTotAlloc ); + mwWrite( " U)nfreed bytes totals : %ld\n", mwStatCurAlloc ); + FLUSH(); + + if( mwStatLevel < 1 ) return; + + /* on a per-module basis */ + mwWrite( "\nMemory usage statistics (detailed):\n"); + mwWrite( " Module/Line Number Largest Total Unfreed \n"); + for( ms=mwStatList; ms; ms=ms->next ) + { + if( ms->line == -1 ) + { + if( ms->file == NULL || !mwIsReadAddr(ms->file,22) ) modname = ""; + else modname = ms->file; + modnamelen = strlen(modname); + if( modnamelen > 42 ) + { + modname = modname + modnamelen - 42; + } + + mwWrite(" %-42s %-8ld %-8ld %-8ld %-8ld\n", + modname, ms->num, ms->max, ms->total, ms->curr ); + if( ms->file && mwStatLevel > 1 ) + { + for( ms2=mwStatList; ms2; ms2=ms2->next ) + { + if( ms2->line!=-1 && ms2->file!=NULL && !mwStrCmpI( ms2->file, ms->file ) ) + { + mwWrite( " %-8d %-8ld %-8ld %-8ld %-8ld\n", + ms2->line, ms2->num, ms2->max, ms2->total, ms2->curr ); + } + } + } + } + } +} + +static mwStat* mwStatGet( const char *file, int line, int makenew ) { + mwStat* ms; + + if( mwStatLevel < 2 ) line = -1; + + for( ms=mwStatList; ms!=NULL; ms=ms->next ) { + if( line != ms->line ) continue; + if( file==NULL ) { + if( ms->file == NULL ) break; + continue; + } + if( ms->file == NULL ) continue; + if( !strcmp( ms->file, file ) ) break; + } + + if( ms != NULL ) return ms; + + if( !makenew ) return NULL; + + ms = (mwStat*) malloc( sizeof(mwStat) ); + if( ms == NULL ) { + if( mwFreeUp( sizeof(mwStat), 0 ) < sizeof(mwStat) || + (ms=(mwStat*)malloc(sizeof(mwStat))) == NULL ) { + mwWrite("internal: memory low, statistics incomplete for '%s'\n", file ); + return NULL; + } + } + ms->file = file; + ms->line = line; + ms->total = 0L; + ms->max = 0L; + ms->num = 0L; + ms->curr = 0L; + ms->next = mwStatList; + mwStatList = ms; + return ms; + } + +static void mwStatAlloc( size_t size, const char* file, int line ) { + mwStat* ms; + + /* update the module statistics */ + ms = mwStatGet( file, -1, 1 ); + if( ms != NULL ) { + ms->total += (long) size; + ms->curr += (long) size; + ms->num ++; + if( ms->curr > ms->max ) ms->max = ms->curr; + } + + /* update the line statistics */ + if( mwStatLevel > 1 && line != -1 && file ) { + ms = mwStatGet( file, line, 1 ); + if( ms != NULL ) { + ms->total += (long) size; + ms->curr += (long) size; + ms->num ++; + if( ms->curr > ms->max ) ms->max = ms->curr; + } + } + + } + +static void mwStatFree( size_t size, const char* file, int line ) { + mwStat* ms; + + /* update the module statistics */ + ms = mwStatGet( file, -1, 1 ); + if( ms != NULL ) ms->curr -= (long) size; + + /* update the line statistics */ + if( mwStatLevel > 1 && line != -1 && file ) { + ms = mwStatGet( file, line, 1 ); + if( ms != NULL ) ms->curr -= (long) size; + } + } + +/*********************************************************************** +** Safe memory checkers +** +** Using ifdefs, implement the operating-system specific mechanism +** of identifying a piece of memory as legal to access with read +** and write priviliges. Default: return nonzero for non-NULL pointers. +***********************************************************************/ + +static char mwDummy( char c ) +{ + return c; +} + +#ifndef MW_SAFEADDR +#ifdef WIN32 +#define MW_SAFEADDR +#define WIN32_LEAN_AND_MEAN +#include +int mwIsReadAddr( const void *p, unsigned len ) +{ + if( p == NULL ) return 0; + if( IsBadReadPtr(p,len) ) return 0; + return 1; +} +int mwIsSafeAddr( void *p, unsigned len ) +{ + /* NOTE: For some reason, under Win95 the IsBad... */ + /* can return false for invalid pointers. */ + if( p == NULL ) return 0; + if( IsBadReadPtr(p,len) || IsBadWritePtr(p,len) ) return 0; + return 1; +} +#endif /* WIN32 */ +#endif /* MW_SAFEADDR */ + +#ifndef MW_SAFEADDR +#ifdef SIGSEGV +#define MW_SAFEADDR + +typedef void (*mwSignalHandlerPtr)( int ); +mwSignalHandlerPtr mwOldSIGSEGV = (mwSignalHandlerPtr) 0; +jmp_buf mwSIGSEGVjump; +static void mwSIGSEGV( int n ); + +static void mwSIGSEGV( int n ) +{ + n = n; + longjmp( mwSIGSEGVjump, 1 ); +} + +int mwIsReadAddr( const void *p, unsigned len ) +{ + const char *ptr; + + if( p == NULL ) return 0; + if( !len ) return 1; + + /* set up to catch the SIGSEGV signal */ + mwOldSIGSEGV = signal( SIGSEGV, mwSIGSEGV ); + + if( setjmp( mwSIGSEGVjump ) ) + { + signal( SIGSEGV, mwOldSIGSEGV ); + return 0; + } + + /* read all the bytes in the range */ + ptr = (const char *)p; + ptr += len; + + /* the reason for this rather strange construct is that */ + /* we want to keep the number of used parameters and locals */ + /* to a minimum. if we use len for a counter gcc will complain */ + /* it may get clobbered by longjmp() at high warning levels. */ + /* it's a harmless warning, but this way we don't have to see it. */ + do + { + ptr --; + if( *ptr == 0x7C ) (void) mwDummy( (char)0 ); + } while( (const void*) ptr != p ); + + /* remove the handler */ + signal( SIGSEGV, mwOldSIGSEGV ); + + return 1; +} +int mwIsSafeAddr( void *p, unsigned len ) +{ + char *ptr; + + if( p == NULL ) return 0; + if( !len ) return 1; + + /* set up to catch the SIGSEGV signal */ + mwOldSIGSEGV = signal( SIGSEGV, mwSIGSEGV ); + + if( setjmp( mwSIGSEGVjump ) ) + { + signal( SIGSEGV, mwOldSIGSEGV ); + return 0; + } + + /* read and write-back all the bytes in the range */ + ptr = (char *)p; + ptr += len; + + /* the reason for this rather strange construct is that */ + /* we want to keep the number of used parameters and locals */ + /* to a minimum. if we use len for a counter gcc will complain */ + /* it may get clobbered by longjmp() at high warning levels. */ + /* it's a harmless warning, but this way we don't have to see it. */ + do + { + ptr --; + *ptr = mwDummy( *ptr ); + } while( (void*) ptr != p ); + + /* remove the handler */ + signal( SIGSEGV, mwOldSIGSEGV ); + + return 1; +} +#endif /* SIGSEGV */ +#endif /* MW_SAFEADDR */ + +#ifndef MW_SAFEADDR +int mwIsReadAddr( const void *p, unsigned len ) +{ + if( p == NULL ) return 0; + if( len == 0 ) return 1; + return 1; +} +int mwIsSafeAddr( void *p, unsigned len ) +{ + if( p == NULL ) return 0; + if( len == 0 ) return 1; + return 1; +} +#endif + +/********************************************************************** +** Mutex handling +**********************************************************************/ + +#if defined(WIN32) || defined(__WIN32__) + +static void mwMutexInit( void ) +{ + mwGlobalMutex = CreateMutex( NULL, FALSE, NULL); + return; +} + +static void mwMutexTerm( void ) +{ + CloseHandle( mwGlobalMutex ); + return; +} + +static void mwMutexLock( void ) +{ + if( WaitForSingleObject(mwGlobalMutex, 1000 ) == WAIT_TIMEOUT ) + { + mwWrite( "mwMutexLock: timed out, possible deadlock\n" ); + } + return; +} + +static void mwMutexUnlock( void ) +{ + ReleaseMutex( mwGlobalMutex ); + return; +} + +#endif + +#if defined(MW_PTHREADS) || defined(HAVE_PTHREAD_H) + +static void mwMutexInit( void ) +{ + pthread_mutex_init( &mwGlobalMutex, NULL ); + return; +} + +static void mwMutexTerm( void ) +{ + pthread_mutex_destroy( &mwGlobalMutex ); + return; +} + +static void mwMutexLock( void ) +{ + pthread_mutex_lock(&mwGlobalMutex); + return; +} + +static void mwMutexUnlock( void ) +{ + pthread_mutex_unlock(&mwGlobalMutex); + return; +} + +#endif + +/********************************************************************** +** C++ new & delete +**********************************************************************/ + +#if 0 /* 980317: disabled C++ */ + +#ifdef __cplusplus +#ifndef MEMWATCH_NOCPP + +int mwNCur = 0; +const char *mwNFile = NULL; +int mwNLine = 0; + +class MemWatch { +public: + MemWatch(); + ~MemWatch(); + }; + +MemWatch::MemWatch() { + if( mwInited ) return; + mwUseAtexit = 0; + mwInit(); + } + +MemWatch::~MemWatch() { + if( mwUseAtexit ) return; + mwTerm(); + } + +/* +** This global new will catch all 'new' calls where MEMWATCH is +** not active. +*/ +void* operator new( unsigned size ) { + mwNCur = 0; + return mwMalloc( size, "", 0 ); + } + +/* +** This is the new operator that's called when a module uses mwNew. +*/ +void* operator new( unsigned size, const char *file, int line ) { + mwNCur = 0; + return mwMalloc( size, file, line ); + } + +/* +** This is the new operator that's called when a module uses mwNew[]. +** -- hjc 07/16/02 +*/ +void* operator new[] ( unsigned size, const char *file, int line ) { + mwNCur = 0; + return mwMalloc( size, file, line ); + } + +/* +** Since this delete operator will recieve ALL delete's +** even those from within libraries, we must accept +** delete's before we've been initialized. Nor can we +** reliably check for wild free's if the mwNCur variable +** is not set. +*/ +void operator delete( void *p ) { + if( p == NULL ) return; + if( !mwInited ) { + free( p ); + return; + } + if( mwNCur ) { + mwFree( p, mwNFile, mwNLine ); + mwNCur = 0; + return; + } + mwFree_( p ); + } + +void operator delete[]( void *p ) { + if( p == NULL ) return; + if( !mwInited ) { + free( p ); + return; + } + if( mwNCur ) { + mwFree( p, mwNFile, mwNLine ); + mwNCur = 0; + return; + } + mwFree_( p ); + } + +#endif /* MEMWATCH_NOCPP */ +#endif /* __cplusplus */ + +#endif /* 980317: disabled C++ */ + +/* MEMWATCH.C */ diff --git a/3rdparty/memwatch/memwatch.h b/3rdparty/memwatch/memwatch.h new file mode 100644 index 0000000..4a42cb9 --- /dev/null +++ b/3rdparty/memwatch/memwatch.h @@ -0,0 +1,710 @@ +/* +** MEMWATCH.H +** Nonintrusive ANSI C memory leak / overwrite detection +** Copyright (C) 1992-2002 Johan Lindh +** All rights reserved. +** Version 2.71 +** +************************************************************************ +** +** PURPOSE: +** +** MEMWATCH has been written to allow guys and gals that like to +** program in C a public-domain memory error control product. +** I hope you'll find it's as advanced as most commercial packages. +** The idea is that you use it during the development phase and +** then remove the MEMWATCH define to produce your final product. +** MEMWATCH is distributed in source code form in order to allow +** you to compile it for your platform with your own compiler. +** It's aim is to be 100% ANSI C, but some compilers are more stingy +** than others. If it doesn't compile without warnings, please mail +** me the configuration of operating system and compiler you are using +** along with a description of how to modify the source, and the version +** number of MEMWATCH that you are using. +** +************************************************************************ + + This file is part of MEMWATCH. + + MEMWATCH is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2 of the License, or + (at your option) any later version. + + MEMWATCH 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 MEMWATCH; if not, write to the Free Software + Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + +************************************************************************ +** +** REVISION HISTORY: +** +** 920810 JLI [1.00] +** 920830 JLI [1.10 double-free detection] +** 920912 JLI [1.15 mwPuts, mwGrab/Drop, mwLimit] +** 921022 JLI [1.20 ASSERT and VERIFY] +** 921105 JLI [1.30 C++ support and TRACE] +** 921116 JLI [1.40 mwSetOutFunc] +** 930215 JLI [1.50 modified ASSERT/VERIFY] +** 930327 JLI [1.51 better auto-init & PC-lint support] +** 930506 JLI [1.55 MemWatch class, improved C++ support] +** 930507 JLI [1.60 mwTest & CHECK()] +** 930809 JLI [1.65 Abort/Retry/Ignore] +** 930820 JLI [1.70 data dump when unfreed] +** 931016 JLI [1.72 modified C++ new/delete handling] +** 931108 JLI [1.77 mwSetAssertAction() & some small changes] +** 940110 JLI [1.80 no-mans-land alloc/checking] +** 940328 JLI [2.00 version 2.0 rewrite] +** Improved NML (no-mans-land) support. +** Improved performance (especially for free()ing!). +** Support for 'read-only' buffers (checksums) +** ^^ NOTE: I never did this... maybe I should? +** FBI (free'd block info) tagged before freed blocks +** Exporting of the mwCounter variable +** mwBreakOut() localizes debugger support +** Allocation statistics (global, per-module, per-line) +** Self-repair ability with relinking +** 950913 JLI [2.10 improved garbage handling] +** 951201 JLI [2.11 improved auto-free in emergencies] +** 960125 JLI [X.01 implemented auto-checking using mwAutoCheck()] +** 960514 JLI [2.12 undefining of existing macros] +** 960515 JLI [2.13 possibility to use default new() & delete()] +** 960516 JLI [2.20 suppression of file flushing on unfreed msgs] +** 960516 JLI [2.21 better support for using MEMWATCH with DLL's] +** 960710 JLI [X.02 multiple logs and mwFlushNow()] +** 960801 JLI [2.22 merged X.01 version with current] +** 960805 JLI [2.30 mwIsXXXXAddr() to avoid unneeded GP's] +** 960805 JLI [2.31 merged X.02 version with current] +** 961002 JLI [2.32 support for realloc() + fixed STDERR bug] +** 961222 JLI [2.40 added mwMark() & mwUnmark()] +** 970101 JLI [2.41 added over/underflow checking after failed ASSERT/VERIFY] +** 970113 JLI [2.42 added support for PC-Lint 7.00g] +** 970207 JLI [2.43 added support for strdup()] +** 970209 JLI [2.44 changed default filename to lowercase] +** 970405 JLI [2.45 fixed bug related with atexit() and some C++ compilers] +** 970723 JLI [2.46 added MW_ARI_NULLREAD flag] +** 970813 JLI [2.47 stabilized marker handling] +** 980317 JLI [2.48 ripped out C++ support; wasn't working good anyway] +** 980318 JLI [2.50 improved self-repair facilities & SIGSEGV support] +** 980417 JLI [2.51 more checks for invalid addresses] +** 980512 JLI [2.52 moved MW_ARI_NULLREAD to occur before aborting] +** 990112 JLI [2.53 added check for empty heap to mwIsOwned] +** 990217 JLI [2.55 improved the emergency repairs diagnostics and NML] +** 990224 JLI [2.56 changed ordering of members in structures] +** 990303 JLI [2.57 first maybe-fixit-for-hpux test] +** 990516 JLI [2.58 added 'static' to the definition of mwAutoInit] +** 990517 JLI [2.59 fixed some high-sensitivity warnings] +** 990610 JLI [2.60 fixed some more high-sensitivity warnings] +** 990715 JLI [2.61 changed TRACE/ASSERT/VERIFY macro names] +** 991001 JLI [2.62 added CHECK_BUFFER() and mwTestBuffer()] +** 991007 JLI [2.63 first shot at a 64-bit compatible version] +** 991009 JLI [2.64 undef's strdup() if defined, mwStrdup made const] +** 000704 JLI [2.65 added some more detection for 64-bits] +** 010502 JLI [2.66 incorporated some user fixes] +** [mwRelink() could print out garbage pointer (thanks mac@phobos.ca)] +** [added array destructor for C++ (thanks rdasilva@connecttel.com)] +** [added mutex support (thanks rdasilva@connecttel.com)] +** 010531 JLI [2.67 fix: mwMutexXXX() was declared even if MW_HAVE_MUTEX was not defined] +** 010619 JLI [2.68 fix: mwRealloc() could leave the mutex locked] +** 020918 JLI [2.69 changed to GPL, added C++ array allocation by Howard Cohen] +** 030212 JLI [2.70 mwMalloc() bug for very large allocations (4GB on 32bits)] +** 030520 JLI [2.71 added ULONG_LONG_MAX as a 64-bit detector (thanks Sami Salonen)] +** +** To use, simply include 'MEMWATCH.H' as a header file, +** and add MEMWATCH.C to your list of files, and define the macro +** 'MEMWATCH'. If this is not defined, MEMWATCH will disable itself. +** +** To call the standard C malloc / realloc / calloc / free; use mwMalloc_(), +** mwCalloc_() and mwFree_(). Note that mwFree_() will correctly +** free both malloc()'d memory as well as mwMalloc()'d. +** +** 980317: C++ support has been disabled. +** The code remains, but is not compiled. +** +** For use with C++, which allows use of inlining in header files +** and class specific new/delete, you must also define 'new' as +** 'mwNew' and 'delete' as 'mwDelete'. Do this *after* you include +** C++ header files from libraries, otherwise you can mess up their +** class definitions. If you don't define these, the C++ allocations +** will not have source file and line number information. Also note, +** most C++ class libraries implement their own C++ memory management, +** and don't allow anyone to override them. MFC belongs to this crew. +** In these cases, the only thing to do is to use MEMWATCH_NOCPP. +** +** You can capture output from MEMWATCH using mwSetOutFunc(). +** Just give it the adress of a "void myOutFunc(int c)" function, +** and all characters to be output will be redirected there. +** +** A failing ASSERT() or VERIFY() will normally always abort your +** program. This can be changed using mwSetAriFunc(). Give it a +** pointer to a "int myAriFunc(const char *)" function. Your function +** must ask the user whether to Abort, Retry or Ignore the trap. +** Return 2 to Abort, 1 to Retry or 0 to Ignore. Beware retry; it +** causes the expression to be evaluated again! MEMWATCH has a +** default ARI handler. It's disabled by default, but you can enable +** it by calling 'mwDefaultAri()'. Note that this will STILL abort +** your program unless you define MEMWATCH_STDIO to allow MEMWATCH +** to use the standard C I/O streams. Also, setting the ARI function +** will cause MEMWATCH *NOT* to write the ARI error to stderr. The +** error string is passed to the ARI function instead, as the +** 'const char *' parameter. +** +** You can disable MEMWATCH's ASSERT/VERIFY and/or TRACE implementations. +** This can be useful if you're using a debug terminal or smart debugger. +** Disable them by defining MW_NOASSERT, MW_NOVERIFY or MW_NOTRACE. +** +** MEMWATCH fills all allocated memory with the byte 0xFE, so if +** you're looking at erroneous data which are all 0xFE:s, the +** data probably was not initialized by you. The exception is +** calloc(), which will fill with zero's. All freed buffers are +** zapped with 0xFD. If this is what you look at, you're using +** data that has been freed. If this is the case, be aware that +** MEMWATCH places a 'free'd block info' structure immediately +** before the freed data. This block contains info about where +** the block was freed. The information is in readable text, +** in the format "FBIfilename(line)", for example: +** "FBI<267>test.c(12)". Using FBI's slows down free(), so it's +** disabled by default. Use mwFreeBufferInfo(1) to enable it. +** +** To aid in tracking down wild pointer writes, MEMWATCH can perform +** no-mans-land allocations. No-mans-land will contain the byte 0xFC. +** MEMWATCH will, when this is enabled, convert recently free'd memory +** into NML allocations. +** +** MEMWATCH protects it's own data buffers with checksums. If you +** get an internal error, it means you're overwriting wildly, +** or using an uninitialized pointer. +** +************************************************************************ +** +** Note when compiling with Microsoft C: +** - MSC ignores fflush() by default. This is overridden, so that +** the disk log will always be current. +** +** This utility has been tested with: +** PC-lint 7.0k, passed as 100% ANSI C compatible +** Microsoft Visual C++ on Win16 and Win32 +** Microsoft C on DOS +** SAS C on an Amiga 500 +** Gnu C on a PC running Red Hat Linux +** ...and using an (to me) unknown compiler on an Atari machine. +** +************************************************************************ +** +** Format of error messages in MEMWATCH.LOG: +** message: filename(linenumber), information +** +** Errors caught by MemWatch, when they are detected, and any +** actions taken besides writing to the log file MEMWATCH.LOG: +** +** Double-freeing: +** A pointer that was recently freed and has not since been +** reused was freed again. The place where the previous free() +** was executed is displayed. +** Detect: delete or free() using the offending pointer. +** Action: The delete or free() is cancelled, execution continues. +** Underflow: +** You have written just ahead of the allocated memory. +** The size and place of the allocation is displayed. +** Detect: delete or free() of the damaged buffer. +** Action: The buffer is freed, but there may be secondary damage. +** Overflow: +** Like underflow, but you've written after the end of the buffer. +** Detect: see Underflow. +** Action: see Underflow. +** WILD free: +** An unrecognized pointer was passed to delete or free(). +** The pointer may have been returned from a library function; +** in that case, use mwFree_() to force free() of it. +** Also, this may be a double-free, but the previous free was +** too long ago, causing MEMWATCH to 'forget' it. +** Detect: delete or free() of the offending pointer. +** Action: The delete or free() is cancelled, execution continues. +** NULL free: +** It's unclear to me whether or not freeing of NULL pointers +** is legal in ANSI C, therefore a warning is written to the log file, +** but the error counter remains the same. This is legal using C++, +** so the warning does not appear with delete. +** Detect: When you free(NULL). +** Action: The free() is cancelled. +** Failed: +** A request to allocate memory failed. If the allocation is +** small, this may be due to memory depletion, but is more likely +** to be memory fragmentation problems. The amount of memory +** allocated so far is displayed also. +** Detect: When you new, malloc(), realloc() or calloc() memory. +** Action: NULL is returned. +** Realloc: +** A request to re-allocate a memory buffer failed for reasons +** other than out-of-memory. The specific reason is shown. +** Detect: When you realloc() +** Action: realloc() is cancelled, NULL is returned +** Limit fail: +** A request to allocate memory failed since it would violate +** the limit set using mwLimit(). mwLimit() is used to stress-test +** your code under simulated low memory conditions. +** Detect: At new, malloc(), realloc() or calloc(). +** Action: NULL is returned. +** Assert trap: +** An ASSERT() failed. The ASSERT() macro works like C's assert() +** macro/function, except that it's interactive. See your C manual. +** Detect: On the ASSERT(). +** Action: Program ends with an advisory message to stderr, OR +** Program writes the ASSERT to the log and continues, OR +** Program asks Abort/Retry/Ignore? and takes that action. +** Verify trap: +** A VERIFY() failed. The VERIFY() macro works like ASSERT(), +** but if MEMWATCH is not defined, it still evaluates the +** expression, but it does not act upon the result. +** Detect: On the VERIFY(). +** Action: Program ends with an advisory message to stderr, OR +** Program writes the VERIFY to the log and continues, OR +** Program asks Abort/Retry/Ignore? and takes that action. +** Wild pointer: +** A no-mans-land buffer has been written into. MEMWATCH can +** allocate and distribute chunks of memory solely for the +** purpose of trying to catch random writes into memory. +** Detect: Always on CHECK(), but can be detected in several places. +** Action: The error is logged, and if an ARI handler is installed, +** it is executed, otherwise, execution continues. +** Unfreed: +** A memory buffer you allocated has not been freed. +** You are informed where it was allocated, and whether any +** over or underflow has occured. MemWatch also displays up to +** 16 bytes of the data, as much as it can, in hex and text. +** Detect: When MemWatch terminates. +** Action: The buffer is freed. +** Check: +** An error was detected during a CHECK() operation. +** The associated pointer is displayed along with +** the file and line where the CHECK() was executed. +** Followed immediately by a normal error message. +** Detect: When you CHECK() +** Action: Depends on the error +** Relink: +** After a MEMWATCH internal control block has been trashed, +** MEMWATCH tries to repair the damage. If successful, program +** execution will continue instead of aborting. Some information +** about the block may be gone permanently, though. +** Detect: N/A +** Action: Relink successful: program continues. +** Relink fails: program aborts. +** Internal: +** An internal error is flagged by MEMWATCH when it's control +** structures have been damaged. You are likely using an uninitialized +** pointer somewhere in your program, or are zapping memory all over. +** The message may give you additional diagnostic information. +** If possible, MEMWATCH will recover and continue execution. +** Detect: Various actions. +** Action: Whatever is needed +** Mark: +** The program terminated without umarking all marked pointers. Marking +** can be used to track resources other than memory. mwMark(pointer,text,...) +** when the resource is allocated, and mwUnmark(pointer) when it's freed. +** The 'text' is displayed for still marked pointers when the program +** ends. +** Detect: When MemWatch terminates. +** Action: The error is logged. +** +** +************************************************************************ +** +** The author may be reached by e-mail at the address below. If you +** mail me about source code changes in MEMWATCH, remember to include +** MW's version number. +** +** Johan Lindh +** johan@linkdata.se +** +** The latest version of MEMWATCH may be downloaded from +** http://www.linkdata.se/ +*/ + +#ifndef __MEMWATCH_H +#define __MEMWATCH_H + +/* Make sure that malloc(), realloc(), calloc() and free() are declared. */ +/*lint -save -e537 */ +#include +/*lint -restore */ + +/* strdup() */ +#include + +#ifdef __cplusplus +extern "C" { +#endif + + +/* +** Constants used +** All MEMWATCH constants start with the prefix MW_, followed by +** a short mnemonic which indicates where the constant is used, +** followed by a descriptive text about it. +*/ + +#define MW_ARI_NULLREAD 0x10 /* Null read (to start debugger) */ +#define MW_ARI_ABORT 0x04 /* ARI handler says: abort program! */ +#define MW_ARI_RETRY 0x02 /* ARI handler says: retry action! */ +#define MW_ARI_IGNORE 0x01 /* ARI handler says: ignore error! */ + +#define MW_VAL_NEW 0xFE /* value in newly allocated memory */ +#define MW_VAL_DEL 0xFD /* value in newly deleted memory */ +#define MW_VAL_NML 0xFC /* value in no-mans-land */ +#define MW_VAL_GRB 0xFB /* value in grabbed memory */ + +#define MW_TEST_ALL 0xFFFF /* perform all tests */ +#define MW_TEST_CHAIN 0x0001 /* walk the heap chain */ +#define MW_TEST_ALLOC 0x0002 /* test allocations & NML guards */ +#define MW_TEST_NML 0x0004 /* test all-NML areas for modifications */ + +#define MW_NML_NONE 0 /* no NML */ +#define MW_NML_FREE 1 /* turn FREE'd memory into NML */ +#define MW_NML_ALL 2 /* all unused memory is NML */ +#define MW_NML_DEFAULT 0 /* the default NML setting */ + +#define MW_STAT_GLOBAL 0 /* only global statistics collected */ +#define MW_STAT_MODULE 1 /* collect statistics on a module basis */ +#define MW_STAT_LINE 2 /* collect statistics on a line basis */ +#define MW_STAT_DEFAULT 0 /* the default statistics setting */ + +/* +** MemWatch internal constants +** You may change these and recompile MemWatch to change the limits +** of some parameters. Respect the recommended minimums! +*/ +#define MW_TRACE_BUFFER 2048 /* (min 160) size of TRACE()'s output buffer */ +#define MW_FREE_LIST 64 /* (min 4) number of free()'s to track */ + +/* +** Exported variables +** In case you have to remove the 'const' keyword because your compiler +** doesn't support it, be aware that changing the values may cause +** unpredictable behaviour. +** - mwCounter contains the current action count. You can use this to +** place breakpoints using a debugger, if you want. +*/ +#ifndef __MEMWATCH_C +extern const unsigned long mwCounter; +#endif + +/* +** System functions +** Normally, it is not nessecary to call any of these. MEMWATCH will +** automatically initialize itself on the first MEMWATCH function call, +** and set up a call to mwAbort() using atexit(). Some C++ implementations +** run the atexit() chain before the program has terminated, so you +** may have to use mwInit() or the MemWatch C++ class to get good +** behaviour. +** - mwInit() can be called to disable the atexit() usage. If mwInit() +** is called directly, you must call mwTerm() to end MemWatch, or +** mwAbort(). +** - mwTerm() is usually not nessecary to call; but if called, it will +** call mwAbort() if it finds that it is cancelling the 'topmost' +** mwInit() call. +** - mwAbort() cleans up after MEMWATCH, reports unfreed buffers, etc. +*/ +void mwInit( void ); +void mwTerm( void ); +void mwAbort( void ); + +/* +** Setup functions +** These functions control the operation of MEMWATCH's protective features. +** - mwFlushNow() causes MEMWATCH to flush it's buffers. +** - mwDoFlush() controls whether MEMWATCH flushes the disk buffers after +** writes. The default is smart flushing: MEMWATCH will not flush buffers +** explicitly until memory errors are detected. Then, all writes are +** flushed until program end or mwDoFlush(0) is called. +** - mwLimit() sets the allocation limit, an arbitrary limit on how much +** memory your program may allocate in bytes. Used to stress-test app. +** Also, in virtual-memory or multitasking environs, puts a limit on +** how much MW_NML_ALL can eat up. +** - mwGrab() grabs up X kilobytes of memory. Allocates actual memory, +** can be used to stress test app & OS both. +** - mwDrop() drops X kilobytes of grabbed memory. +** - mwNoMansLand() sets the behaviour of the NML logic. See the +** MW_NML_xxx for more information. The default is MW_NML_DEFAULT. +** - mwStatistics() sets the behaviour of the statistics collector. See +** the MW_STAT_xxx defines for more information. Default MW_STAT_DEFAULT. +** - mwFreeBufferInfo() enables or disables the tagging of free'd buffers +** with freeing information. This information is written in text form, +** using sprintf(), so it's pretty slow. Disabled by default. +** - mwAutoCheck() performs a CHECK() operation whenever a MemWatch function +** is used. Slows down performance, of course. +** - mwCalcCheck() calculates checksums for all data buffers. Slow! +** - mwDumpCheck() logs buffers where stored & calc'd checksums differ. Slow!! +** - mwMark() sets a generic marker. Returns the pointer given. +** - mwUnmark() removes a generic marker. If, at the end of execution, some +** markers are still in existence, these will be reported as leakage. +** returns the pointer given. +*/ +void mwFlushNow( void ); +void mwDoFlush( int onoff ); +void mwLimit( long bytes ); +unsigned mwGrab( unsigned kilobytes ); +unsigned mwDrop( unsigned kilobytes ); +void mwNoMansLand( int mw_nml_level ); +void mwStatistics( int level ); +void mwFreeBufferInfo( int onoff ); +void mwAutoCheck( int onoff ); +void mwCalcCheck( void ); +void mwDumpCheck( void ); +void * mwMark( void *p, const char *description, const char *file, unsigned line ); +void * mwUnmark( void *p, const char *file, unsigned line ); + +/* +** Testing/verification/tracing +** All of these macros except VERIFY() evaluates to a null statement +** if MEMWATCH is not defined during compilation. +** - mwIsReadAddr() checks a memory area for read privilige. +** - mwIsSafeAddr() checks a memory area for both read & write privilige. +** This function and mwIsReadAddr() is highly system-specific and +** may not be implemented. If this is the case, they will default +** to returning nonzero for any non-NULL pointer. +** - CHECK() does a complete memory integrity test. Slow! +** - CHECK_THIS() checks only selected components. +** - CHECK_BUFFER() checks the indicated buffer for errors. +** - mwASSERT() or ASSERT() If the expression evaluates to nonzero, execution continues. +** Otherwise, the ARI handler is called, if present. If not present, +** the default ARI action is taken (set with mwSetAriAction()). +** ASSERT() can be disabled by defining MW_NOASSERT. +** - mwVERIFY() or VERIFY() works just like ASSERT(), but when compiling without +** MEMWATCH the macro evaluates to the expression. +** VERIFY() can be disabled by defining MW_NOVERIFY. +** - mwTRACE() or TRACE() writes some text and data to the log. Use like printf(). +** TRACE() can be disabled by defining MW_NOTRACE. +*/ +int mwIsReadAddr( const void *p, unsigned len ); +int mwIsSafeAddr( void *p, unsigned len ); +int mwTest( const char *file, int line, int mw_test_flags ); +int mwTestBuffer( const char *file, int line, void *p ); +int mwAssert( int, const char*, const char*, int ); +int mwVerify( int, const char*, const char*, int ); + +/* +** User I/O functions +** - mwTrace() works like printf(), but dumps output either to the +** function specified with mwSetOutFunc(), or the log file. +** - mwPuts() works like puts(), dumps output like mwTrace(). +** - mwSetOutFunc() allows you to give the adress of a function +** where all user output will go. (exeption: see mwSetAriFunc) +** Specifying NULL will direct output to the log file. +** - mwSetAriFunc() gives MEMWATCH the adress of a function to call +** when an 'Abort, Retry, Ignore' question is called for. The +** actual error message is NOT printed when you've set this adress, +** but instead it is passed as an argument. If you call with NULL +** for an argument, the ARI handler is disabled again. When the +** handler is disabled, MEMWATCH will automatically take the +** action specified by mwSetAriAction(). +** - mwSetAriAction() sets the default ARI return value MEMWATCH should +** use if no ARI handler is specified. Defaults to MW_ARI_ABORT. +** - mwAriHandler() is an ANSI ARI handler you can use if you like. It +** dumps output to stderr, and expects input from stdin. +** - mwBreakOut() is called in certain cases when MEMWATCH feels it would +** be nice to break into a debugger. If you feel like MEMWATCH, place +** an execution breakpoint on this function. +*/ +void mwTrace( const char* format_string, ... ); +void mwPuts( const char* text ); +void mwSetOutFunc( void (*func)(int) ); +void mwSetAriFunc( int (*func)(const char*) ); +void mwSetAriAction( int mw_ari_value ); +int mwAriHandler( const char* cause ); +void mwBreakOut( const char* cause ); + +/* +** Allocation/deallocation functions +** These functions are the ones actually to perform allocations +** when running MEMWATCH, for both C and C++ calls. +** - mwMalloc() debugging allocator +** - mwMalloc_() always resolves to a clean call of malloc() +** - mwRealloc() debugging re-allocator +** - mwRealloc_() always resolves to a clean call of realloc() +** - mwCalloc() debugging allocator, fills with zeros +** - mwCalloc_() always resolves to a clean call of calloc() +** - mwFree() debugging free. Can only free memory which has +** been allocated by MEMWATCH. +** - mwFree_() resolves to a) normal free() or b) debugging free. +** Can free memory allocated by MEMWATCH and malloc() both. +** Does not generate any runtime errors. +*/ +void* mwMalloc( size_t, const char*, int ); +void* mwMalloc_( size_t ); +void* mwRealloc( void *, size_t, const char*, int ); +void* mwRealloc_( void *, size_t ); +void* mwCalloc( size_t, size_t, const char*, int ); +void* mwCalloc_( size_t, size_t ); +void mwFree( void*, const char*, int ); +void mwFree_( void* ); +char* mwStrdup( const char *, const char*, int ); + +/* +** Enable/disable precompiler block +** This block of defines and if(n)defs make sure that references +** to MEMWATCH is completely removed from the code if the MEMWATCH +** manifest constant is not defined. +*/ +#ifndef __MEMWATCH_C +#ifdef MEMWATCH + +#define mwASSERT(exp) while(mwAssert((int)(exp),#exp,__FILE__,__LINE__)) +#ifndef MW_NOASSERT +#ifndef ASSERT +#define ASSERT mwASSERT +#endif /* !ASSERT */ +#endif /* !MW_NOASSERT */ +#define mwVERIFY(exp) while(mwVerify((int)(exp),#exp,__FILE__,__LINE__)) +#ifndef MW_NOVERIFY +#ifndef VERIFY +#define VERIFY mwVERIFY +#endif /* !VERIFY */ +#endif /* !MW_NOVERIFY */ +#define mwTRACE mwTrace +#ifndef MW_NOTRACE +#ifndef TRACE +#define TRACE mwTRACE +#endif /* !TRACE */ +#endif /* !MW_NOTRACE */ + +/* some compilers use a define and not a function */ +/* for strdup(). */ +#ifdef strdup +#undef strdup +#endif + +#define malloc(n) mwMalloc(n,__FILE__,__LINE__) +#define strdup(p) mwStrdup(p,__FILE__,__LINE__) +#define realloc(p,n) mwRealloc(p,n,__FILE__,__LINE__) +#define calloc(n,m) mwCalloc(n,m,__FILE__,__LINE__) +#define free(p) mwFree(p,__FILE__,__LINE__) +#define CHECK() mwTest(__FILE__,__LINE__,MW_TEST_ALL) +#define CHECK_THIS(n) mwTest(__FILE__,__LINE__,n) +#define CHECK_BUFFER(b) mwTestBuffer(__FILE__,__LINE__,b) +#define MARK(p) mwMark(p,#p,__FILE__,__LINE__) +#define UNMARK(p) mwUnmark(p,__FILE__,__LINE__) + +#else /* MEMWATCH */ + +#define mwASSERT(exp) +#ifndef MW_NOASSERT +#ifndef ASSERT +#define ASSERT mwASSERT +#endif /* !ASSERT */ +#endif /* !MW_NOASSERT */ + +#define mwVERIFY(exp) exp +#ifndef MW_NOVERIFY +#ifndef VERIFY +#define VERIFY mwVERIFY +#endif /* !VERIFY */ +#endif /* !MW_NOVERIFY */ + +/*lint -esym(773,mwTRACE) */ +#define mwTRACE /*lint -save -e506 */ 1?(void)0:mwDummyTraceFunction /*lint -restore */ +#ifndef MW_NOTRACE +#ifndef TRACE +/*lint -esym(773,TRACE) */ +#define TRACE mwTRACE +#endif /* !TRACE */ +#endif /* !MW_NOTRACE */ + +extern void mwDummyTraceFunction(const char *,...); +/*lint -save -e652 */ +#define mwDoFlush(n) +#define mwPuts(s) +#define mwInit() +#define mwGrab(n) +#define mwDrop(n) +#define mwLimit(n) +#define mwTest(f,l) +#define mwSetOutFunc(f) +#define mwSetAriFunc(f) +#define mwDefaultAri() +#define mwNomansland() +#define mwStatistics(f) +#define mwMark(p,t,f,n) (p) +#define mwUnmark(p,f,n) (p) +#define mwMalloc(n,f,l) malloc(n) +#define mwStrdup(p,f,l) strdup(p) +#define mwRealloc(p,n,f,l) realloc(p,n) +#define mwCalloc(n,m,f,l) calloc(n,m) +#define mwFree(p) free(p) +#define mwMalloc_(n) malloc(n) +#define mwRealloc_(p,n) realloc(p,n) +#define mwCalloc_(n,m) calloc(n,m) +#define mwFree_(p) free(p) +#define mwAssert(e,es,f,l) +#define mwVerify(e,es,f,l) (e) +#define mwTrace mwDummyTrace +#define mwTestBuffer(f,l,b) (0) +#define CHECK() +#define CHECK_THIS(n) +#define CHECK_BUFFER(b) +#define MARK(p) (p) +#define UNMARK(p) (p) +/*lint -restore */ + +#endif /* MEMWATCH */ +#endif /* !__MEMWATCH_C */ + +#ifdef __cplusplus + } +#endif + +#if 0 /* 980317: disabled C++ */ + +/* +** C++ support section +** Implements the C++ support. Please note that in order to avoid +** messing up library classes, C++ support is disabled by default. +** You must NOT enable it until AFTER the inclusion of all header +** files belonging to code that are not compiled with MEMWATCH, and +** possibly for some that are! The reason for this is that a C++ +** class may implement it's own new() function, and the preprocessor +** would substitute this crucial declaration for MEMWATCH new(). +** You can forcibly deny C++ support by defining MEMWATCH_NOCPP. +** To enble C++ support, you must be compiling C++, MEMWATCH must +** be defined, MEMWATCH_NOCPP must not be defined, and finally, +** you must define 'new' to be 'mwNew', and 'delete' to be 'mwDelete'. +** Unlike C, C++ code can begin executing *way* before main(), for +** example if a global variable is created. For this reason, you can +** declare a global variable of the class 'MemWatch'. If this is +** is the first variable created, it will then check ALL C++ allocations +** and deallocations. Unfortunately, this evaluation order is not +** guaranteed by C++, though the compilers I've tried evaluates them +** in the order encountered. +*/ +#ifdef __cplusplus +#ifndef __MEMWATCH_C +#ifdef MEMWATCH +#ifndef MEMWATCH_NOCPP +extern int mwNCur; +extern const char *mwNFile; +extern int mwNLine; +class MemWatch { +public: + MemWatch(); + ~MemWatch(); + }; +void * operator new(size_t); +void * operator new(size_t,const char *,int); +void * operator new[] (size_t,const char *,int); // hjc 07/16/02 +void operator delete(void *); +#define mwNew new(__FILE__,__LINE__) +#define mwDelete (mwNCur=1,mwNFile=__FILE__,mwNLine=__LINE__),delete +#endif /* MEMWATCH_NOCPP */ +#endif /* MEMWATCH */ +#endif /* !__MEMWATCH_C */ +#endif /* __cplusplus */ + +#endif /* 980317: disabled C++ */ + +#endif /* __MEMWATCH_H */ + +/* EOF MEMWATCH.H */ diff --git a/3rdparty/memwatch/memwatch.lsm b/3rdparty/memwatch/memwatch.lsm new file mode 100644 index 0000000..a7742a3 --- /dev/null +++ b/3rdparty/memwatch/memwatch.lsm @@ -0,0 +1,15 @@ +Begin3 +Title: memwatch +Version: 2.71 +Entered-date: 2002-09-18 +Description: fault tolerant ANSI-C source code memory leak and corruption detection +Keywords: memwatch debugging library memory leak source code ansi c +Author: johan@linkdata.se +Maintained-by: johan@linkdata.se +Primary-site: ftp.linkdata.se /pub/memwatch + 42K memwatch-2.71.tar.gz +Alternate-site: +Original-site: ftp.linkdata.se /pub/memwatch +Platforms: all +Copying-policy: GPL +End diff --git a/3rdparty/memwatch/test.c b/3rdparty/memwatch/test.c new file mode 100644 index 0000000..b4d847c --- /dev/null +++ b/3rdparty/memwatch/test.c @@ -0,0 +1,116 @@ + +/* +** NOTE: Running this program in a Win32 or Unix environment +** will probably result in a segmentation fault or protection +** error. These errors may be caused by MEMWATCH when it is +** looking at memory to see if it owns it, or may be caused by +** the test program writing to memory it does not own. +** +** MEMWATCH has two functions called 'mwIsReadAddr()' and +** 'mwIsSafeAddr()', which are system-specific. +** If they are implemented for your system, and works +** correctly, MEMWATCH will identify garbage pointers and +** avoid causing segmentation faults, GP's etc. +** +** If they are NOT implemented, count on getting the core +** dumped when running this test program! As of this writing, +** the safe-address checking has been implemented for Win32 +** and ANSI-C compliant systems. The ANSI-C checking traps +** SIGSEGV and uses setjmp/longjmp to resume processing. +** +** Note for Win95 users: The Win32 IsBadReadPtr() and its +** similar functions can return incorrect values. This has +** not happened under WinNT, though, just Win95. +** +** 991009 Johan Lindh +** +*/ + +#include +#include +#include "memwatch.h" + +#ifndef SIGSEGV +#error "SIGNAL.H does not define SIGSEGV; running this program WILL cause a core dump/crash!" +#endif + +#ifndef MEMWATCH +#error "You really, really don't want to run this without memwatch. Trust me." +#endif + +#if !defined(MW_STDIO) && !defined(MEMWATCH_STDIO) +#error "Define MW_STDIO and try again, please." +#endif + +int main() +{ + char *p; + + /* Collect stats on a line number basis */ + mwStatistics( 2 ); + + /* Slows things down, but OK for this test prg */ + /* mwAutoCheck( 1 ); */ + + TRACE("Hello world!\n"); + + p = malloc(210); + free(p); + p = malloc(20); + p = malloc(200); /* causes unfreed error */ + p[-1] = 0; /* causes underflow error */ + free(p); + + p = malloc(100); + p[ -(int)(sizeof(long)*8) ] = -1; /* try to damage MW's heap chain */ + free( p ); /* should cause relink */ + + mwSetAriFunc( mwAriHandler ); + ASSERT(1==2); + + mwLimit(1000000); + mwNoMansLand( MW_NML_ALL ); + + /* These may cause a general protection fault (segmentation fault) */ + /* They're here to help test the no-mans-land protection */ + if( mwIsSafeAddr(p+50000,1) ) { + TRACE("Killing byte at %p\n", p+50000); + *(p+50000) = 0; + } + if( mwIsSafeAddr(p+30000,1) ) { + TRACE("Killing byte at %p\n", p+30000); + *(p+30000) = 0; + } + if( mwIsSafeAddr(p+1000,1) ) { + TRACE("Killing byte at %p\n", p+1000); + *(p+1000) = 0; + } + if( mwIsSafeAddr(p-100,1) ) { + TRACE("Killing byte at %p\n", p-100); + *(p-100) = 0; + } + + /* This may cause a GP fault as well, since MW data buffers */ + /* have been damaged in the above killing spree */ + CHECK(); + + p = malloc(12000); + p[-5] = 1; + p[-10] = 2; + p[-15] = 3; + p[-20] = 4; + + /* This may cause a GP fault since MW's buffer list may have */ + /* been damaged by above killing, and it will try to repair it. */ + free(p); + + p = realloc(p,10); /* causes realloc: free'd from error */ + + /* May cause GP since MW will inspect the memory to see if it owns it. */ + free( (void*)main ); + + return 0; +} + +/* Comment out the following line to compile. */ +#error "Hey! Don't just compile this program, read the comments first!" diff --git a/3rdparty/pthreads/fpth314b.zip b/3rdparty/pthreads/fpth314b.zip new file mode 100644 index 0000000..8a07f91 --- /dev/null +++ b/3rdparty/pthreads/fpth314b.zip @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8eb938be276f3d65cbf6fae6c7924296e0bc5531d6b3e3851ee930ae042af3cb +size 84267 diff --git a/3rdparty/pthreads/fpth314s.zip b/3rdparty/pthreads/fpth314s.zip new file mode 100644 index 0000000..c854bd7 --- /dev/null +++ b/3rdparty/pthreads/fpth314s.zip @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ac2c0e1c0ad3ec847ef40ff1f34d20dac1fb3aa232ec602daaf16026f5adc6c9 +size 297092 diff --git a/3rdparty/pthreads/include/pthread.h b/3rdparty/pthreads/include/pthread.h new file mode 100644 index 0000000..9b28dfc --- /dev/null +++ b/3rdparty/pthreads/include/pthread.h @@ -0,0 +1,677 @@ +/* Copyright (C) 1992-2000 the Florida State University + Distributed by the Florida State University under the terms of the + GNU Library General Public License. + +This file is part of Pthreads. + +Pthreads is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public +License as published by the Free Software Foundation (version 2). + +Pthreads is distributed "AS IS" 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 Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with Pthreads; see the file COPYING. If not, write +to the Free Software Foundation, 675 Mass Ave, Cambridge, +MA 02139, USA. + +Report problems and direct all questions to: + + pthreads-bugs@ada.cs.fsu.edu + + @(#)pthread.h 3.14 11/8/00 + +*/ + +#ifndef _pthread_pthread_h +#define _pthread_pthread_h + +/* + * Pthreads interface definition + */ + +#include +#include +#ifdef SOLARIS_NP +#if defined(_POSIX_PTHREAD_SEMANTICS_NP) && !defined(_POSIX_PTHREAD_SEMANTICS) +#define _POSIX_PTHREAD_SEMANTICS +#endif +#if defined(__PRAGMA_REDEFINE_EXTNAME_NP) && !defined(__PRAGMA_REDEFINE_EXTNAME) +#define __PRAGMA_REDEFINE_EXTNAME +#endif +#include +#endif +#include +/*#include moved up */ +#include +#include +#include /* needed for struct timeval */ +#ifdef _M_UNIX +/* + * Bug in latest SCO 3.2: gettimeofday return neg. value after alarm + */ +#undef timerisset +#define timerisset(tp) ((tp)->tv_sec > 0 || (tp)->tv_usec > 0) +#endif /* _M_UNIX */ +#include +#include + +/* + * When setjmp.h is included using SCO, sigjmp_buf will only be defined, if + * additional defines have been done, here _POSIX_SOURCE is used: + */ +#ifdef _M_UNIX +#ifndef _POSIX_SOURCE +#define _POSIX_SOURCE +#include +#undef _POSIX_SOURCE +#else /* _POSIX_SOURCE */ +#include +#endif /* _POSIX_SOURCE */ +#else /* !_M_UNIX */ +#include +#endif /* !_M_UNIX */ + +#ifdef IO_NP +#if defined(__FreeBSD__) || defined(_M_UNIX) || defined(__linux__) || defined(__dos__) +#include +#else /* !__FreeBSD__ && !_M_UNIX && !__linux__ && !__dos__ */ +#include +#include +#endif /* !__FreeBSD__ && !_M_UNIX && !__linux__ && !__dos__ */ +#endif /* IO_NP */ +#if defined(MALLOC_NP) || defined(STAND_ALONE) +#include +#ifdef malloc +#undef malloc +#endif +#ifdef calloc +#undef calloc +#endif +#ifdef free +#undef free +#endif +#ifdef cfree +#undef cfree +#endif +#endif /* MALLOC_NP */ + +#if defined(__FreeBSD__) +#include +#elif defined(__dos__) +#include +#elif defined(__linux__) +#include +#elif defined(_M_UNIX) +#include +#endif + +/* contentionscope attribute values */ +#define PTHREAD_SCOPE_SYSTEM 0 +#define PTHREAD_SCOPE_PROCESS 1 + +/* creation modes */ +#define PTHREAD_CREATE_JOINABLE 0 +#define PTHREAD_CREATE_DETACHED 1 + +/* inheritsched attribute values */ +#define PTHREAD_INHERIT_SCHED 0 +#define PTHREAD_EXPLICIT_SCHED 1 + +/* Allow variable stack sizes */ +#ifndef _POSIX_THREAD_ATTR_STACKSIZE +#define _POSIX_THREAD_ATTR_STACKSIZE +#endif + +#if defined (__GNUC__) || defined(__STDC__) || defined(__cplusplus) || defined(c_plusplus) + +#ifndef _C_PROTOTYPE +#define _C_PROTOTYPE(arglist) arglist +#endif /* !_C_PROTOTYPE */ + +typedef void *any_t; + +#else /* !(defined (__GNUC__) || defined(__STDC__) || defined(__cplusplus) || defined(c_plusplus)) */ + +#ifndef _C_PROTOTYPE +#define _C_PROTOTYPE(arglist) () +#endif /* !_C_PROTOTYPE */ + +typedef char *any_t; + +#ifndef const +#define const +#endif + +#ifndef volatile +#define volatile +#endif + +#endif /* defined(__GNUC__) || defined(__STDC__) || defined(__cplusplus) || defined(c_plusplus) */ + +#ifdef __linux__ +#include +#endif +#ifndef TRUE +#define TRUE 1 +#define FALSE 0 +#endif + +#if !defined(__sys_stdtypes_h) && !defined(_SIZE_T) && !defined(__FreeBSD__) && !defined(_M_UNIX) && !defined(__linux__) && !defined(__dos__) +typedef unsigned int size_t; +#endif + +/* + * Mutex objects. + */ + +typedef int pthread_protocol_t; + +typedef struct pthread_queue { + struct pthread *head; + struct pthread *tail; +} *pthread_queue_t; + +typedef struct { + struct pthread_queue queue; + char lock; + struct pthread *owner; + int flags; +#ifdef _POSIX_THREADS_PRIO_PROTECT + int prioceiling; + pthread_protocol_t protocol; + int prev_max_ceiling_prio; +#endif +} pthread_mutex_t; + +typedef struct { + int flags; +#ifdef _POSIX_THREADS_PRIO_PROTECT + int prioceiling; + pthread_protocol_t protocol; +#endif +} pthread_mutexattr_t; + +#ifdef _POSIX_THREADS_PRIO_PROTECT +#define PTHREAD_MUTEX_INITIALIZER { \ + { 0, 0 }, /* queue */ \ + 0, /* lock */ \ + 0, /* owner */ \ + 1, /* flags */ \ + 0, /* prioceiling */ \ + PTHREAD_PRIO_NONE /* protocol */ \ +} +#else +#define PTHREAD_MUTEX_INITIALIZER { \ + { 0, 0 }, /* queue */ \ + 0, /* lock */ \ + 0, /* owner */ \ + 1 /* flags */ \ + } +#endif + +/******************************/ +/* Mutex Attributes Functions */ +/* Mutex Functions */ +/******************************/ + +#ifdef __cplusplus +extern "C" { +#endif + +extern int pthread_mutex_lock _C_PROTOTYPE((pthread_mutex_t *__mutex)); +extern int pthread_mutex_trylock _C_PROTOTYPE((pthread_mutex_t *__mutex)); +extern int pthread_mutex_unlock _C_PROTOTYPE((pthread_mutex_t *__mutex)); +extern int pthread_mutex_init _C_PROTOTYPE((pthread_mutex_t *__mutex, + pthread_mutexattr_t *__attr)); +extern int pthread_mutex_destroy _C_PROTOTYPE((pthread_mutex_t *__mutex)); +extern int pthread_mutex_setprioceiling + _C_PROTOTYPE((pthread_mutex_t *__mutex, + int __prioceiling)); +extern int pthread_mutex_getprioceiling + _C_PROTOTYPE((pthread_mutex_t __mutex, + int *__prioceiling)); +extern int pthread_mutexattr_init _C_PROTOTYPE((pthread_mutexattr_t *__attr)); +extern int pthread_mutexattr_destroy + _C_PROTOTYPE((pthread_mutexattr_t *__attr)); +extern int pthread_mutexattr_setprotocol + _C_PROTOTYPE((pthread_mutexattr_t *__attr, + pthread_protocol_t __protocol)); +extern int pthread_mutexattr_getprotocol + _C_PROTOTYPE((pthread_mutexattr_t *__attr, + int *__protocol)); + +extern int pthread_mutexattr_setprioceiling + _C_PROTOTYPE((pthread_mutexattr_t *__attr, + int __prioceiling)); +extern int pthread_mutexattr_getprioceiling + _C_PROTOTYPE((pthread_mutexattr_t *__attr, + int *__prioceiling)); +extern int pthread_mutexattr_getpshared + _C_PROTOTYPE((pthread_mutexattr_t *__attr, + int *__pshared)); +extern int pthread_mutexattr_setpshared + _C_PROTOTYPE((pthread_mutexattr_t *__attr, + int __pshared)); + +#ifdef __cplusplus +} /* extern "C" */ +#endif + +/* + * Once Objects. + */ + +typedef struct { + short init; + short exec; + pthread_mutex_t mutex; +} pthread_once_t; + +#define PTHREAD_ONCE_INIT {FALSE,FALSE,PTHREAD_MUTEX_INITIALIZER}; + +/* + * Condition variables. + */ +typedef struct { + struct pthread_queue queue; + int flags; + int waiters; + pthread_mutex_t *mutex; +} pthread_cond_t; + +typedef struct { + int flags; +} pthread_condattr_t; + +#define PTHREAD_COND_INITIALIZER { \ + { 0, 0 }, /* queue */ \ + 1, /* flags */ \ + 0, /* waiters */ \ + 0 /* mutex */ \ +} + +/*******************************/ +/* Condition Functions */ +/*Condition Attribute Functions*/ +/*******************************/ + +#ifdef __cplusplus +extern "C" { +#endif + +extern int pthread_cond_destroy _C_PROTOTYPE((pthread_cond_t *__cond)); +extern int pthread_cond_init _C_PROTOTYPE((pthread_cond_t *__cond, + pthread_condattr_t *__attr)); +extern int pthread_condattr_init _C_PROTOTYPE((pthread_condattr_t *__attr)); +extern int pthread_condattr_destroy _C_PROTOTYPE((pthread_condattr_t *__attr)); +extern int pthread_cond_wait _C_PROTOTYPE((pthread_cond_t *__cond, + pthread_mutex_t *__mutex)); +extern int pthread_cond_timedwait _C_PROTOTYPE((pthread_cond_t *__cond, + pthread_mutex_t *__mutex, + struct timespec *__timeout)); +extern int pthread_cond_signal _C_PROTOTYPE((pthread_cond_t *__cond)); +extern int pthread_cond_broadcast _C_PROTOTYPE((pthread_cond_t *__cond)); +extern int pthread_condattr_getpshared + _C_PROTOTYPE((pthread_condattr_t *__attr, + int *__pshared)); +extern int pthread_condattr_setpshared + _C_PROTOTYPE((pthread_condattr_t *__attr, + int __pshared)); + +#ifdef __cplusplus +} /* extern "C" */ +#endif + +/* + * Threads. + */ + +typedef any_t (*pthread_func_t) _C_PROTOTYPE((any_t __arg)); +typedef void (*pthread_sighandler_t) _C_PROTOTYPE((int)); + +struct sched_param { + int sched_priority; +}; + +typedef struct { + int flags; +#ifdef _POSIX_THREAD_ATTR_STACKSIZE + int stacksize; +#endif + int contentionscope; + int inheritsched; + int detachstate; + int sched; + struct sched_param param; + struct timespec starttime, deadline, period; +} pthread_attr_t; + +typedef struct { + int state; + pthread_queue_t queue; + pthread_cond_t *cond; +} pthread_suspend_t; + +/* + * Queue indices + */ +#define PRIMARY_QUEUE 0 +#define ALL_QUEUE 1 +/* this slot may be used by the TIMER_QUEUE with index 2 */ +#define NUM_QUEUES 3 + +typedef int pthread_key_t; + +typedef struct pthread_cleanup *pthread_cleanup_t; + +typedef struct pthread { +#if defined (__FreeBSD__) || defined (_M_UNIX) || defined(__linux__) || defined(__dos__) + SYS_SIGJMP_BUF context; /* save area for context switch */ + SYS_SIGJMP_BUF body; /* save area for pthread_body */ +#else + sigjmp_buf context; /* save area for context switch */ + sigjmp_buf body; /* save area for pthread_body */ +#endif + volatile int terrno; /* thread-specific errno */ + volatile int ret; /* return value (EINTR --> -1) */ + char *stack_base; /* bottom of run-time stack */ + int state; /* thread state, -> pthread_asm.h */ + struct pthread *next[NUM_QUEUES]; /* links for queues */ + int num_timers; /* number of timers for thread */ + struct timeval interval; /* time left for SCHED_RR */ + struct p_siginfo sig_info[NNSIG]; /* info for user handlers */ + int sig; /* latest received signal */ + int code; /* latest received signal code */ + int osp, opc, obp; /* save area for old context sw */ + struct context_t *nscp; /* UNIX signal context (new sig) */ + struct context_t *scp; /* UNIX signal context (current) */ + struct pthread_queue joinq; /* queue to await termination */ + pthread_cond_t *cond; /* cond var. if in cond_wait */ + pthread_queue_t queue;/* primary queue thread is contained in */ + sigset_t mask; /* set of signals thread has masked out */ + sigset_t pending; /* set of signals pending on thread */ + sigset_t sigwaitset; /* set of signals thread is waiting for */ + pthread_func_t func; /* actual function to call upon activation */ + any_t arg; /* argument list to above function */ + any_t result; /* return value of above function */ + any_t key[_POSIX_DATAKEYS_MAX]; /* thread specific data */ + pthread_cleanup_t cleanup_top; /* stack of cleanup handlers */ + pthread_attr_t attr; /* attributes */ + int base_prio; /* Base priority of thread */ + int max_ceiling_prio; /* Max of ceiling prio among locked mutexes */ + int new_prio; /* New Priority */ + pthread_suspend_t suspend; /* save area for thread suspend */ +#if defined(IO_NP) && !defined(__FreeBSD__) && !defined(__dos__) +#if !defined(__linux__) && !defined(_M_UNIX) + volatile struct aio_result_t resultp;/*result of asynchronous I/O ops*/ +#endif /* !__linux__ && !_M_UNIX */ +#ifdef USE_POLL_NP + int wait_on_select; /* more information in fds */ + int nfds; /* pertinent file desc. set width */ + int how_many; /* how many amoung 0 .. width -1 */ + struct pollfd* fds; /* poll events */ +#else /* !USE_POLL_NP */ + int wait_on_select; /* more information in fds */ + int width; /* pertinent file desc. set width */ + int how_many; /* how many amoung 0 .. width -1 */ + fd_set readfds; /* read file descriptor set */ + fd_set writefds; /* write file descriptor set */ + fd_set exceptfds; /* except. file descriptor set */ +#endif /* !USE_POLL_NP */ +#endif /* IO_NP && !__FreeBSD__ && !__dos__ */ +#ifdef STAND_ALONE_NP + pthread_func_t timer_func; /* function to be called on timeout*/ + any_t timer_arg; /* arg of above function */ + struct timespec tp; /* wake-up time */ + int dummy; /* filler for stack alignment */ +#else /* !STAND_ALONE_NP */ + struct timeval tp; /* wake-up time */ +#endif /* !STAND_ALONE_NP */ +} *pthread_t; + +#ifdef __FreeBSD__ +#define IO_SIZE_T size_t +#elif defined (__dos__) +#define IO_SIZE_T size_t +#define SIGSUSPEND_CONST const +#define SIGACTION_CONST const +#define SIGPROCMASK_CONST const +#define SIGWAIT_CONST const +#define LONGJMP_CONST +#define SIGLONGJMP_CONST +#elif defined (__linux__) +#define IO_SIZE_T size_t +#elif defined (_M_UNIX) +#define IO_SIZE_T size_t +#else +#if defined(__GNUC__) && defined(_PARAMS) +#define IO_SIZE_T __SIZE_TYPE__ +/* + * for gcc 2.5.8 or before use: + * #define IO_SIZE_T unsigned int + */ +#define SIGSUSPEND_CONST +#define SIGACTION_CONST const +#define SIGPROCMASK_CONST const +#define SIGWAIT_CONST const +#define LONGJMP_CONST +#define SIGLONGJMP_CONST +#else +/* + * for older gcc, remove "const" for "SIGPROCMASK_CONST" + */ +#define IO_SIZE_T size_t +#define SIGSUSPEND_CONST const +#define SIGACTION_CONST const +#define SIGPROCMASK_CONST const +#define SIGWAIT_CONST const +#define LONGJMP_CONST +#define SIGLONGJMP_CONST +#endif +#endif + + +/******************************/ +/* Thread Functions */ +/* Thread Attribute Functions */ +/******************************/ + +#ifdef __cplusplus +extern "C" { +#endif + +extern pthread_t pthread_self _C_PROTOTYPE ((void)); +extern void pthread_init _C_PROTOTYPE ((void)); +extern int pthread_create _C_PROTOTYPE((pthread_t *___thread, + pthread_attr_t *__attr, + pthread_func_t __func, + any_t __arg)); +extern int pthread_equal _C_PROTOTYPE((pthread_t __t1, + pthread_t __t2)); +extern int pthread_detach _C_PROTOTYPE((pthread_t ___thread)); +extern int pthread_join _C_PROTOTYPE((pthread_t ___thread, + any_t *__status)); +extern int sched_yield _C_PROTOTYPE((void)); +extern void pthread_exit _C_PROTOTYPE((any_t __status)); +extern int pthread_attr_init _C_PROTOTYPE((pthread_attr_t *__attr)); +extern int pthread_attr_destroy _C_PROTOTYPE((pthread_attr_t *__attr)); +extern int pthread_getschedparam _C_PROTOTYPE((pthread_t ___thread, + int *__policy, + struct sched_param *__param)); +extern int pthread_setschedparam _C_PROTOTYPE((pthread_t ___thread, + int __policy, + struct sched_param *__param)); +extern int pthread_attr_setstacksize _C_PROTOTYPE((pthread_attr_t *__attr, + size_t __stacksize)); +extern int pthread_attr_getstacksize _C_PROTOTYPE((pthread_attr_t *__attr, + size_t *__stacksize)); +extern int pthread_attr_setscope _C_PROTOTYPE((pthread_attr_t *__attr, + int __contentionscope)); +extern int pthread_attr_setinheritsched _C_PROTOTYPE((pthread_attr_t *__attr, + int __inheritsched)); +extern int pthread_attr_setschedpolicy _C_PROTOTYPE((pthread_attr_t *__attr, + int __policy)); +extern int pthread_attr_setschedparam _C_PROTOTYPE((pthread_attr_t *__attr, + struct sched_param *__param)); +extern int pthread_attr_getscope _C_PROTOTYPE((pthread_attr_t *__attr, + int *__contentionscope)); +extern int pthread_attr_getinheritsched _C_PROTOTYPE((pthread_attr_t *__attr, + int *__inheritsched)); +extern int pthread_attr_getschedpolicy _C_PROTOTYPE((pthread_attr_t *__attr, + int *__policy)); +extern int pthread_attr_getschedparam _C_PROTOTYPE((pthread_attr_t *__attr, + struct sched_param *__param)); + +extern int pthread_attr_getstarttime_np _C_PROTOTYPE((pthread_attr_t *__attr, + struct timespec *__tp)); +extern int pthread_attr_setstarttime_np _C_PROTOTYPE((pthread_attr_t *__attr, + struct timespec *__tp)); +extern int pthread_attr_getdeadline_np _C_PROTOTYPE((pthread_attr_t *__attr, + struct timespec *__tp)); +extern int pthread_attr_setdeadline_np _C_PROTOTYPE((pthread_attr_t *__attr, + struct timespec *__tp, + pthread_func_t func)); +extern int pthread_attr_getperiod_np _C_PROTOTYPE((pthread_attr_t *__attr, + struct timespec *__tp)); +extern int pthread_attr_setperiod_np _C_PROTOTYPE((pthread_attr_t *__attr, + struct timespec *__tp, + pthread_func_t func)); + +extern int pthread_key_create _C_PROTOTYPE((pthread_key_t *__key, + void (*__func) + (any_t __value))); +extern int pthread_key_delete _C_PROTOTYPE((pthread_key_t __key)); +extern int pthread_setspecific _C_PROTOTYPE((pthread_key_t __key, + any_t __value)); +extern any_t pthread_getspecific _C_PROTOTYPE((pthread_key_t __key)); +extern void pthread_cleanup_push _C_PROTOTYPE((void (*__func) + (any_t __value), + any_t __arg)); +extern void pthread_cleanup_pop _C_PROTOTYPE((int __execute)); +extern int sched_get_priority_max _C_PROTOTYPE((int __policy)); +extern int sched_get_priority_min _C_PROTOTYPE((int __policy)); +extern int pthread_attr_setdetachstate _C_PROTOTYPE((pthread_attr_t *__attr, + int __detachstate)); +extern int pthread_attr_getdetachstate _C_PROTOTYPE((pthread_attr_t *__attr, + int *__detachstate)); +extern int pthread_once _C_PROTOTYPE((pthread_once_t *__once_c, + void (*__func) (void))); + +/* + * implementation-defined extensions + */ +extern void pthread_setsigcontext_np _C_PROTOTYPE((struct context_t *__scp, + jmp_buf __env, + int __val)); +extern int pthread_lock_stack_np _C_PROTOTYPE((pthread_t __p)); + +extern int pthread_suspend_np _C_PROTOTYPE((pthread_t __t)); +extern int pthread_resume_np _C_PROTOTYPE((pthread_t __t)); + +/******************************/ +/* Signal Functions */ +/******************************/ + +extern int sigwait _C_PROTOTYPE((SIGWAIT_CONST sigset_t *__set, + int *__sig)); +extern int sigprocmask _C_PROTOTYPE((int __how, + SIGPROCMASK_CONST sigset_t *__set, + sigset_t *__oset)); +extern int pthread_sigmask _C_PROTOTYPE((int __how, + SIGPROCMASK_CONST sigset_t *__set, + sigset_t *__oset)); +extern int sigpending _C_PROTOTYPE((sigset_t *__set)); +extern int sigsuspend _C_PROTOTYPE((SIGSUSPEND_CONST sigset_t *__set)); +extern int pause _C_PROTOTYPE((void)); +#ifdef __dos__ +/* we use a macro to rename these so we can still call the system + functions, since we don't have system-call hooks under DJGPP. */ +#define raise(sig) pthread_dummy_raise(sig) +#define sigprocmask(sig, set1, set2) pthread_dummy_sigprocmask(sig, set1, set2) +#define sigsuspend(set) pthread_dummy_sigsuspend(set) +#endif +#if defined(M_UNIX) +#undef raise +#endif +extern int raise _C_PROTOTYPE((int __sig)); +extern int pthread_kill _C_PROTOTYPE((pthread_t ___thread, + int __sig)); +extern int pthread_cancel _C_PROTOTYPE((pthread_t ___thread)); +extern int pthread_setcancelstate _C_PROTOTYPE((int __state, int *__oldstate)); +extern int pthread_setcanceltype _C_PROTOTYPE((int __type, int *__oldtype)); +extern void pthread_testcancel _C_PROTOTYPE((void)); +extern int sigaction _C_PROTOTYPE((int __sig, + SIGACTION_CONST struct sigaction *__act, + struct sigaction *__oact)); +#if defined(__FreeBSD__) || defined(_M_UNIX) || defined(__linux__) || defined(__dos__) +extern pthread_sighandler_t signal + _C_PROTOTYPE((int __sig, + pthread_sighandler_t handler)); +#else +#if defined(SOLARIS_NP) && defined(__cplusplus) +extern void (*signal(int, void (*)(int)))(int); +#else +extern void (*signal())(); +#endif +#endif + +/* yet to come... +extern unsigned int alarm _C_PROTOTYPE((unsigned int __seconds)); +*/ +extern int nanosleep _C_PROTOTYPE((const struct timespec *__rqtp, + struct timespec *__rmtp)); +extern unsigned int sleep _C_PROTOTYPE((unsigned int __seconds)); +extern int clock_gettime _C_PROTOTYPE((int __clock_id, + struct timespec *__tp)); + +/******************************/ +/* Low-Level Functions */ +/******************************/ + +#ifndef __dos__ +#ifdef setjmp +#undef setjmp +#endif +#ifdef longjmp +#undef longjmp +#endif +#ifdef sigsetjmp +#undef sigsetjmp +#endif +#ifdef siglongjmp +#undef siglongjmp +#endif +#endif + +extern int setjmp _C_PROTOTYPE((jmp_buf __env)); +extern void longjmp _C_PROTOTYPE((LONGJMP_CONST jmp_buf __env, + int __val)); +extern int sigsetjmp _C_PROTOTYPE((sigjmp_buf __env, + int __savemask)); +extern void siglongjmp _C_PROTOTYPE((SIGLONGJMP_CONST sigjmp_buf __env, + int __val)); + +/******************************/ +/* I/O Functions */ +/******************************/ + +#if !defined(__DJGPP__) +extern int read _C_PROTOTYPE((int __fd, + void *__buf, + IO_SIZE_T __nbytes)); +extern int write _C_PROTOTYPE((int __fd, + const void *__buf, + IO_SIZE_T __nbytes)); +#endif /* !__DJGPP__ */ + +#ifdef __cplusplus +} /* extern "C" */ +#endif + +#define PTHREAD_CANCELED (void *)-1 + +#endif /* !_pthread_pthread_h */ diff --git a/3rdparty/pthreads/include/pthread/asm.h b/3rdparty/pthreads/include/pthread/asm.h new file mode 100644 index 0000000..8bcbd59 --- /dev/null +++ b/3rdparty/pthreads/include/pthread/asm.h @@ -0,0 +1,111 @@ +/* Copyright (C) 1992-2000 the Florida State University + Distributed by the Florida State University under the terms of the + GNU Library General Public License. + +This file is part of Pthreads. + +Pthreads is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public +License as published by the Free Software Foundation (version 2). + +Pthreads is distributed "AS IS" 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 Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with Pthreads; see the file COPYING. If not, write +to the Free Software Foundation, 675 Mass Ave, Cambridge, +MA 02139, USA. + +Report problems and direct all questions to: + + pthreads-bugs@ada.cs.fsu.edu + + @(#)asm.h 3.14 11/8/00 + +*/ + +#ifndef _pthread_pthread_asm_h +#define _pthread_pthread_asm_h + +/* + * sched attribute values + */ +#ifdef _POSIX_THREAD_PRIORITY_SCHEDULING +#define SCHED_FIFO 0 +#define SCHED_RR 1 +#define SCHED_OTHER 2 +#endif + +/* + * If we have Priority Ceilings, then we have Priority Inheritance + */ +#ifdef _POSIX_THREADS_PRIO_PROTECT +#define _POSIX_THREADS_PRIO_INHERIT +#define PTHREAD_PRIO_NONE 0 +#define PTHREAD_PRIO_INHERIT 1 +#define PTHREAD_PRIO_PROTECT 2 +#endif + +#ifndef SIG_BLOCK +/* sigprocmask flags */ +#define SIG_BLOCK 0x0001 +#define SIG_UNBLOCK 0x0002 +#define SIG_SETMASK 0x0004 +#endif /* SIG_BLOCK */ + +#define HEAP_SIZE 1024*1024 + +/* + * timer set modes + */ +#define ABS_TIME 0x01 +#define REL_TIME 0x02 +#define SYNC_TIME (ABS_TIME | REL_TIME) +#define RR_TIME 0x04 +#define DL_TIME 0x08 +#define ANY_TIME (SYNC_TIME | RR_TIME | DL_TIME) +#define ALL_TIME (0x10 | ANY_TIME) + +/* + * Thread status bits. + */ +#define T_MAIN 0x1 +#define T_RETURNED 0x2 +#define T_DETACHED 0x4 +#define T_RUNNING 0x8 +#define T_BLOCKED 0x10 +#define T_CONDTIMER 0x20 +#define T_SIGWAIT 0x40 +#define T_SYNCTIMER 0x80 +#define T_SIGSUSPEND 0x100 +#define T_CONTROLLED 0x200 +#define T_INTR_POINT 0x400 +#define T_ASYNCTIMER 0x800 +#define T_LOCKED 0x1000 +#define T_IO_OVER 0x2000 +#define T_EXITING 0x4000 +#define T_SUSPEND 0x8000 + +#ifndef NULL +#define NULL 0 +#endif + +/* + * Offset added to pc by ret and retl instruction on SPARC. + * When a call instruction is issued, the address of the call + * instruction is saved as the return address. When executing + * a ret/retl a constant of 8 is added to the pc to return + * to the instruction following the call *and* the daley slot. + * Below, we want to jump to a function through the dispatcher + * which uses ret/retl. Thus, we need to provide the address of + * the function *minus* 8. + */ +#if defined(__FreeBSD__) || defined(_M_UNIX) || defined(__linux__) +#define RETURN_OFFSET 0 +#else +#define RETURN_OFFSET 8 +#endif + +#endif /* !_pthread_pthread_asm_h */ diff --git a/3rdparty/pthreads/include/pthread/config.h b/3rdparty/pthreads/include/pthread/config.h new file mode 100644 index 0000000..e12109f --- /dev/null +++ b/3rdparty/pthreads/include/pthread/config.h @@ -0,0 +1,64 @@ +/* Copyright (C) 1992, 1993, 1994, 1995, 1996 the Florida State University + Distributed by the Florida State University under the terms of the + GNU Library General Public License. + +This file is part of Pthreads. + +Pthreads is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public +License as published by the Free Software Foundation (version 2). + +Pthreads is distributed "AS IS" 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 Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with Pthreads; see the file COPYING. If not, write +to the Free Software Foundation, 675 Mass Ave, Cambridge, +MA 02139, USA. + +Report problems and direct all questions to: + + pthreads-bugs@ada.cs.fsu.edu + + %@(#)config_header.c 3.14% %11/8/00% +*/ + +/* + * configuration header file to identify compile options + */ + +#ifndef C_INTERFACE_NP +#define C_INTERFACE_NP +#endif + +#ifndef SRP_NP +#define SRP_NP +#endif + +#ifndef __dos__ +#define __dos__ +#endif + +#ifndef _POSIX +#define _POSIX +#endif + +#ifndef CLEANUP_HEAP_NP +#define CLEANUP_HEAP_NP +#endif + +#ifndef C_CONTEXT_SWITCH_NP +#define C_CONTEXT_SWITCH_NP +#endif + +#ifndef RELEASE_NP +#define RELEASE_NP 5 +#endif + +#ifndef _M_UNIX +#if defined(M_UNIX) || defined(__M_UNIX) +#define _M_UNIX +#endif +#endif diff --git a/3rdparty/pthreads/include/pthread/dos_setjmp.h b/3rdparty/pthreads/include/pthread/dos_setjmp.h new file mode 100644 index 0000000..30a0870 --- /dev/null +++ b/3rdparty/pthreads/include/pthread/dos_setjmp.h @@ -0,0 +1,73 @@ +/* Copyright (C) 1992-2000 the Florida State University + Distributed by the Florida State University under the terms of the + GNU Library General Public License. + +This file is part of Pthreads. + +Pthreads is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public +License as published by the Free Software Foundation (version 2). + +Pthreads is distributed "AS IS" 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 Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with Pthreads; see the file COPYING. If not, write +to the Free Software Foundation, 675 Mass Ave, Cambridge, +MA 02139, USA. + +Report problems and direct all questions to: + + pthreads-bugs@ada.cs.fsu.edu + + @(#)dos_setjmp.h 1.1 07 Apr 1996 + +*/ + +#ifndef _PTHREAD_SETJMP_H_ +#define _PTHREAD_SETJMP_H_ + +#ifndef _ANSI_SOURCE +#include + +typedef int pthread_sigjmp_buf[44]; /* see src/dos/setjmp.S */ +#endif /* !_ANSI_SOURCE */ + +#include + +__BEGIN_DECLS + + +/* For DOS, the libc exception handler routine must be free to call its + own setjmp/longjmp functions--which use a different format buffer from + ours. Thus, we must not replace the existing setjmp/longjmp functions. + + We kludge our way around this by using macros to redefine the function + names. This is a potential sore spot if linked with binaries that do + not include pthread.h! + + These macros are also defined in src/dos/setjmp.S. */ + +#undef setjmp +#undef longjmp +#undef sigsetjmp +#undef siglongjmp +#define setjmp pt_setjmp +#define longjmp pt_longjmp +#define sigsetjmp pt_sigsetjmp +#define siglongjmp pt_siglongjmp + + +#ifndef _ANSI_SOURCE +int pthread_sigsetjmp __P((pthread_sigjmp_buf, int, int)); +void pthread_siglongjmp __P((pthread_sigjmp_buf, int, int)); + +#define SYS_SIGSETJMP pthread_sigsetjmp +#define SYS_SIGLONGJMP pthread_siglongjmp +#define SYS_SIGJMP_BUF pthread_sigjmp_buf +#endif /* !_ANSI_SOURCE */ +__END_DECLS + +#endif /* !_PTHREAD_SETJMP_H_ */ diff --git a/3rdparty/pthreads/include/pthread/errno.h b/3rdparty/pthreads/include/pthread/errno.h new file mode 100644 index 0000000..4a1c1d8 --- /dev/null +++ b/3rdparty/pthreads/include/pthread/errno.h @@ -0,0 +1,41 @@ +/* Copyright (C) 1992-2000 the Florida State University + Distributed by the Florida State University under the terms of the + GNU Library General Public License. + +This file is part of Pthreads. + +Pthreads is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public +License as published by the Free Software Foundation (version 2). + +Pthreads is distributed "AS IS" 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 Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with Pthreads; see the file COPYING. If not, write +to the Free Software Foundation, 675 Mass Ave, Cambridge, +MA 02139, USA. + +Report problems and direct all questions to: + + pthreads-bugs@ada.cs.fsu.edu + + @(#)errno.h 3.14 11/8/00 +*/ + +#ifndef _pthread_errno_h +#define _pthread_errno_h + +#include + +#ifndef ENOSYS +#define ENOSYS EREMCHG+1 +#endif + +#ifndef ENOTSUP +#define ENOTSUP ENOSYS+1 +#endif + +#endif /* !__pthread_errno_h */ diff --git a/3rdparty/pthreads/include/pthread/fre_setjmp.h b/3rdparty/pthreads/include/pthread/fre_setjmp.h new file mode 100644 index 0000000..8ce41e2 --- /dev/null +++ b/3rdparty/pthreads/include/pthread/fre_setjmp.h @@ -0,0 +1,49 @@ +/* Copyright (C) 1992-2000 the Florida State University + Distributed by the Florida State University under the terms of the + GNU Library General Public License. + +This file is part of Pthreads. + +Pthreads is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public +License as published by the Free Software Foundation (version 2). + +Pthreads is distributed "AS IS" 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 Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with Pthreads; see the file COPYING. If not, write +to the Free Software Foundation, 675 Mass Ave, Cambridge, +MA 02139, USA. + +Report problems and direct all questions to: + + pthreads-bugs@ada.cs.fsu.edu + + @(#)fre_setjmp.h 3.14 11/8/00 + +*/ + +#ifndef _PTHREAD_SETJMP_H_ +#define _PTHREAD_SETJMP_H_ + +#ifndef _ANSI_SOURCE +typedef int pthread_sigjmp_buf[68]; +#endif /* !_ANSI_SOURCE */ + +#include + +__BEGIN_DECLS +#ifndef _ANSI_SOURCE +int pthread_sigsetjmp __P((pthread_sigjmp_buf, int, int)); +void pthread_siglongjmp __P((pthread_sigjmp_buf, int, int)); + +#define SYS_SIGSETJMP pthread_sigsetjmp +#define SYS_SIGLONGJMP pthread_siglongjmp +#define SYS_SIGJMP_BUF pthread_sigjmp_buf +#endif /* !_ANSI_SOURCE */ +__END_DECLS + +#endif /* !_PTHREAD_SETJMP_H_ */ diff --git a/3rdparty/pthreads/include/pthread/int_types.h b/3rdparty/pthreads/include/pthread/int_types.h new file mode 100644 index 0000000..4e8bfa0 --- /dev/null +++ b/3rdparty/pthreads/include/pthread/int_types.h @@ -0,0 +1,150 @@ +/* + * Copyright (c) 1996 by Sun Microsystems, Inc. + * All rights reserved. + */ + +#ifndef _SYS_INT_TYPES_H +#define _SYS_INT_TYPES_H + +#pragma ident "@(#)int_types.h 1.4 96/09/25 SMI" + +/* + * This file, , is part of the Sun Microsystems implementation + * of as proposed in the ISO/JTC1/SC22/WG14 C committee's working + * draft for the revision of the current ISO C standard, ISO/IEC 9899:1990 + * Programming language - C. + * + * Programs/Modules should not directly include this file. Access to the + * types defined in this file should be through the inclusion of one of the + * following files: + * + * Provides only the "_t" types defined in this + * file which is a subset of the contents of + * . (This can be appropriate for + * all programs/modules except those claiming + * ANSI-C conformance.) + * + * Provides the Kernel and Driver appropriate + * components of . + * + * For use by applications. + * + * See these files for more details. + * + * Use at your own risk. As of February 1996, the committee is squarely + * behind the fixed sized types; the "least" and "fast" types are still being + * discussed. The probability that the "fast" types may be removed before + * the standard is finalized is high enough that they are not currently + * implemented. The unimplemented "fast" types are of the form + * [u]int_fast[0-9]*_t and [u]intfast_t. + */ + +#include + +#ifdef __cplusplus +extern "C" { +#endif + +/* + * Basic / Extended integer types + * + * The following defines the basic fixed-size integer types. + * + * Implementations are free to typedef them to Standard C integer types or + * extensions that they support. If an implementation does not support one + * of the particular integer data types below, then it should not define the + * typedefs and macros corresponding to that data type. Note that int8_t + * is not defined in -Xs mode on ISAs for which the ABI specifies "char" + * as an unsigned entity because there is not way to defined an eight bit + * signed integral. + */ +#if defined(_CHAR_IS_SIGNED) +typedef char int8_t; +#else +#if defined(__STDC__) +typedef signed char int8_t; +#endif +#endif +typedef short int16_t; +typedef int int32_t; +#ifdef _LP64 +typedef long int64_t; +#else /* _ILP32 */ +#if __STDC__ - 0 == 0 && !defined(_NO_LONGLONG) +typedef long long int64_t; +#endif +#endif + +typedef unsigned char uint8_t; +typedef unsigned short uint16_t; +typedef unsigned int uint32_t; +#ifdef _LP64 +typedef unsigned long uint64_t; +#else /* _ILP32 */ +#if __STDC__ - 0 == 0 && !defined(_NO_LONGLONG) +typedef unsigned long long uint64_t; +#endif +#endif + +/* + * intmax_t and uintmax_t are to be the longest (in number of bits) signed + * and unsigned integer types supported by the implementation. + */ +#if defined(_LP64) || (__STDC__ - 0 == 0 && !defined(_NO_LONGLONG)) +typedef int64_t intmax_t; +typedef uint64_t uintmax_t; +#else +typedef int32_t intmax_t; +typedef uint32_t uintmax_t; +#endif + +/* + * intptr_t and uintptr_t are signed and unsigned integer types large enough + * to hold any data pointer; that is, data pointers can be assigned into or + * from these integer types without losing precision. + */ +#ifdef _LP64 +typedef long intptr_t; +typedef unsigned long uintptr_t; +#else +typedef int intptr_t; +typedef unsigned int uintptr_t; +#endif /* _LP64 */ + +/* + * The following define the smallest integer types that can hold the + * specified number of bits. + */ +#if defined(_CHAR_IS_SIGNED) +typedef char int_least8_t; +#else +#if defined(__STDC__) +typedef signed char int_least8_t; +#endif +#endif +typedef short int_least16_t; +typedef int int_least32_t; +#ifdef _LP64 +typedef long int_least64_t; +#else /* _ILP32 */ +#if __STDC__ - 0 == 0 && !defined(_NO_LONGLONG) +typedef long long int_least64_t; +#endif +#endif + +typedef unsigned char uint_least8_t; +typedef unsigned short uint_least16_t; +typedef unsigned int uint_least32_t; +#ifdef _LP64 +typedef unsigned long uint_least64_t; +#else /* _ILP32 */ +#if __STDC__ - 0 == 0 && !defined(_NO_LONGLONG) +typedef unsigned long long uint_least64_t; +#endif +#endif + +#ifdef __cplusplus +} +#endif + +#endif /* _SYS_INT_TYPES_H */ diff --git a/3rdparty/pthreads/include/pthread/limits.h b/3rdparty/pthreads/include/pthread/limits.h new file mode 100644 index 0000000..3b43961 --- /dev/null +++ b/3rdparty/pthreads/include/pthread/limits.h @@ -0,0 +1,44 @@ +/* Copyright (C) 1992-2000 the Florida State University + Distributed by the Florida State University under the terms of the + GNU Library General Public License. + +This file is part of Pthreads. + +Pthreads is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public +License as published by the Free Software Foundation (version 2). + +Pthreads is distributed "AS IS" 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 Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with Pthreads; see the file COPYING. If not, write +to the Free Software Foundation, 675 Mass Ave, Cambridge, +MA 02139, USA. + +Report problems and direct all questions to: + + pthreads-bugs@ada.cs.fsu.edu + + @(#)limits.h 3.14 11/8/00 + +*/ + +#ifndef _pthread_limits_h +#define _pthread_limits_h + +#ifndef STAND_ALONE_NP +#include +#endif /* STAND_ALONE_NP */ + +#ifndef _POSIX_DATAKEYS_MAX +#define _POSIX_DATAKEYS_MAX 256 +#endif + +#ifndef _POSIX_TIMER_MAX +#define _POSIX_TIMER_MAX 128 +#endif + +#endif /* !_pthread_limits_h */ diff --git a/3rdparty/pthreads/include/pthread/lin_setjmp.h b/3rdparty/pthreads/include/pthread/lin_setjmp.h new file mode 100644 index 0000000..93df98f --- /dev/null +++ b/3rdparty/pthreads/include/pthread/lin_setjmp.h @@ -0,0 +1,49 @@ +/* Copyright (C) 1992-2000 the Florida State University + Distributed by the Florida State University under the terms of the + GNU Library General Public License. + +This file is part of Pthreads. + +Pthreads is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public +License as published by the Free Software Foundation (version 2). + +Pthreads is distributed "AS IS" 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 Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with Pthreads; see the file COPYING. If not, write +to the Free Software Foundation, 675 Mass Ave, Cambridge, +MA 02139, USA. + +Report problems and direct all questions to: + + pthreads-bugs@ada.cs.fsu.edu + + @(#)lin_setjmp.h 3.14 11/8/00 + +*/ + +#ifndef _PTHREAD_SETJMP_H_ +#define _PTHREAD_SETJMP_H_ + +#ifndef _ANSI_SOURCE +typedef int pthread_sigjmp_buf[68]; +#endif /* !_ANSI_SOURCE */ + +#include + +__BEGIN_DECLS +#ifndef _ANSI_SOURCE +int pthread_sigsetjmp __P((pthread_sigjmp_buf, int, int)); +void pthread_siglongjmp __P((pthread_sigjmp_buf, int, int)); + +#define SYS_SIGSETJMP pthread_sigsetjmp +#define SYS_SIGLONGJMP pthread_siglongjmp +#define SYS_SIGJMP_BUF pthread_sigjmp_buf +#endif /* !_ANSI_SOURCE */ +__END_DECLS + +#endif /* !_PTHREAD_SETJMP_H_ */ diff --git a/3rdparty/pthreads/include/pthread/sco_setjmp.h b/3rdparty/pthreads/include/pthread/sco_setjmp.h new file mode 100644 index 0000000..8a2ffe9 --- /dev/null +++ b/3rdparty/pthreads/include/pthread/sco_setjmp.h @@ -0,0 +1,53 @@ +/* Copyright (C) 1992-2000 the Florida State University + Distributed by the Florida State University under the terms of the + GNU Library General Public License. + +This file is part of Pthreads. + +Pthreads is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public +License as published by the Free Software Foundation (version 2). + +Pthreads is distributed "AS IS" 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 Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with Pthreads; see the file COPYING. If not, write +to the Free Software Foundation, 675 Mass Ave, Cambridge, +MA 02139, USA. + +Report problems and direct all questions to: + + pthreads-bugs@ada.cs.fsu.edu + + @(#)sco_setjmp.h 3.14 11/8/00 + +*/ + +#ifndef _PTHREAD_SETJMP_H_ +#define _PTHREAD_SETJMP_H_ + +#ifdef __cplusplus +extern "C" { +#endif + +#if defined(SCO5_NP) +typedef int pthread_sigjmp_buf[_SIGJBLEN]; +#else +typedef int pthread_sigjmp_buf[68]; +#endif + +extern int pthread_sigsetjmp (pthread_sigjmp_buf, int, int); +extern void pthread_siglongjmp (pthread_sigjmp_buf, int, int); + +#define SYS_SIGSETJMP pthread_sigsetjmp +#define SYS_SIGLONGJMP pthread_siglongjmp +#define SYS_SIGJMP_BUF pthread_sigjmp_buf + +#ifdef __cplusplus +}; +#endif + +#endif /* !_PTHREAD_SETJMP_H_ */ diff --git a/3rdparty/pthreads/include/pthread/signal-sol.h b/3rdparty/pthreads/include/pthread/signal-sol.h new file mode 100644 index 0000000..197d9bd --- /dev/null +++ b/3rdparty/pthreads/include/pthread/signal-sol.h @@ -0,0 +1,251 @@ +/* Copyright (c) 1988 AT&T */ +/* All Rights Reserved */ + +/* THIS IS UNPUBLISHED PROPRIETARY SOURCE CODE OF AT&T */ +/* The copyright notice above does not evidence any */ +/* actual or intended publication of such source code. */ + +/* + * Copyright (c) 1998-1999, by Sun Microsystems, Inc. + * All rights reserved. + */ + + +#ifndef _SIGNAL_H +#define _SIGNAL_H + +#pragma ident "@(#)signal.h 1.38 99/08/10 SMI" /* SVr4.0 1.5.3.4 */ + +#include + +#if defined(__EXTENSIONS__) || __STDC__ == 0 || \ + defined(_POSIX_C_SOURCE) || defined(_XOPEN_SOURCE) +#include /* need pid_t/uid_t/size_t/clock_t/caddr_t/pthread_t */ +#endif + +#if defined(SOLARIS_NP) && #if RELEASE_NP > 56 +#include +#endif +#include + +/* + * Allow global visibility for symbols defined in + * C++ "std" namespace in . + */ +#if __cplusplus >= 199711L +using std::sig_atomic_t; +using std::signal; +using std::raise; +#endif + +#ifdef __cplusplus +extern "C" { +#endif + + +#if defined(__STDC__) + +extern const char **_sys_siglistp; /* signal descriptions */ +extern const int _sys_siglistn; /* # of signal descriptions */ + +#if defined(__EXTENSIONS__) || \ + (!defined(_POSIX_C_SOURCE) && !defined(_XOPEN_SOURCE)) +#define _sys_siglist _sys_siglistp +#define _sys_nsig _sys_siglistn +#endif + +#if defined(__EXTENSIONS__) || __STDC__ == 0 || \ + defined(_POSIX_C_SOURCE) || defined(_XOPEN_SOURCE) +extern int kill(pid_t, int); +extern int sigaction(int, const struct sigaction *, struct sigaction *); +#ifndef _KERNEL +extern int sigaddset(sigset_t *, int); +extern int sigdelset(sigset_t *, int); +extern int sigemptyset(sigset_t *); +extern int sigfillset(sigset_t *); +extern int sigismember(const sigset_t *, int); +#endif +extern int sigpending(sigset_t *); +extern int sigprocmask(int, const sigset_t *, sigset_t *); +extern int sigsuspend(const sigset_t *); +#endif /* defined(__EXTENSIONS__) || __STDC__ == 0 ... */ + +#if defined(__EXTENSIONS__) || (__STDC__ == 0 && \ + !defined(_POSIX_C_SOURCE) && !defined(_XOPEN_SOURCE)) +#include +extern int gsignal(int); +extern int (*ssignal(int, int (*)(int)))(int); +extern int sigsend(idtype_t, id_t, int); +extern int sigsendset(const procset_t *, int); +extern int sig2str(int, char *); +extern int str2sig(const char *, int *); +#define SIG2STR_MAX 32 +#endif /* defined(__EXTENSIONS__) || (__STDC__ == 0 ... */ + +#if defined(__EXTENSIONS__) || (__STDC__ == 0 && \ + !defined(_POSIX_C_SOURCE) && !defined(_XOPEN_SOURCE)) || \ + defined(_XPG4_2) +extern void (*bsd_signal(int, void (*)(int)))(int); +extern int killpg(pid_t, int); +extern int siginterrupt(int, int); +extern int sigaltstack(const stack_t *, stack_t *); +extern int sighold(int); +extern int sigignore(int); +extern int sigpause(int); +extern int sigrelse(int); +extern void (*sigset(int, void (*)(int)))(int); +extern int sigstack(struct sigstack *, struct sigstack *); +#endif /* defined(__EXTENSIONS__) || (__STDC__ == 0 ... */ + +#if defined(__EXTENSIONS__) || (__STDC__ == 0 && \ + !defined(_POSIX_C_SOURCE) && !defined(_XOPEN_SOURCE)) || \ + (_POSIX_C_SOURCE > 2) +#include +#include +#ifndef SOLARIS_NP +extern int pthread_kill(pthread_t, int); +#endif +extern int pthread_sigmask(int, const sigset_t *, sigset_t *); +extern int sigwaitinfo(const sigset_t *, siginfo_t *); +extern int sigtimedwait(const sigset_t *, siginfo_t *, const struct timespec *); +extern int sigqueue(pid_t, int, const union sigval); +#endif /* defined(__EXTENSIONS__) || (__STDC__ == 0 ... */ + +#else /* __STDC__ */ + +extern char **_sys_siglistp; /* signal descriptions */ +extern int _sys_siglistn; /* # of signal descriptions */ + +#define _sys_siglist _sys_siglistp +#define _sys_nsig _sys_siglistn + +extern void(*signal())(); +extern int raise(); + +extern int kill(); +extern int sigaction(); +#ifndef _KERNEL +extern int sigaddset(); +extern int sigdelset(); +extern int sigemptyset(); +extern int sigfillset(); +extern int sigismember(); +#endif +extern int sigpending(); +extern int sigprocmask(); +extern int sigsuspend(); + +#if defined(__EXTENSIONS__) || \ + (!defined(_POSIX_C_SOURCE) && !defined(_XOPEN_SOURCE)) || \ + defined(_XPG4_2) +extern void (*bsd_signal())(); +extern int killpg(); +extern int siginterrupt(); +extern int sigstack(); +#endif /* defined(__EXTENSIONS__) ... */ + +#if defined(__EXTENSIONS__) || \ + (!defined(_POSIX_C_SOURCE) && !defined(_XOPEN_SOURCE)) +extern int gsignal(); +extern int (*ssignal)(); +extern int sigsend(); +extern int sigsendset(); +extern int sig2str(); +extern int str2sig(); +#define SIG2STR_MAX 32 +#endif + +#if defined(__EXTENSIONS__) || \ + (!defined(_POSIX_C_SOURCE) && !defined(_XOPEN_SOURCE)) || \ + defined(_XPG4_2) +extern int sigaltstack(); +extern int sighold(); +extern int sigignore(); +extern int sigpause(); +extern int sigrelse(); +extern void (*sigset())(); +#endif + +#if defined(__EXTENSIONS__) || \ + (!defined(_POSIX_C_SOURCE) && !defined(_XOPEN_SOURCE)) || \ + (_POSIX_C_SOURCE > 2) +#include +#include +extern int pthread_kill(); +extern int pthread_sigmask(); +extern int sigwaitinfo(); +extern int sigtimedwait(); +extern int sigqueue(); +#endif + +#endif /* __STDC__ */ + +/* + * sigwait() prototype is defined here. + */ + +#if defined(__EXTENSIONS__) || (__STDC__ == 0 && \ + !defined(_POSIX_C_SOURCE) && !defined(_XOPEN_SOURCE)) || \ + (_POSIX_C_SOURCE - 0 >= 199506L) || defined(_POSIX_PTHREAD_SEMANTICS) + +#if defined(__STDC__) + +#if (_POSIX_C_SOURCE - 0 >= 199506L) || defined(_POSIX_PTHREAD_SEMANTICS) + +#ifdef __PRAGMA_REDEFINE_EXTNAME +extern int sigwait(const sigset_t *, int *); +#pragma redefine_extname sigwait __posix_sigwait +#else /* __PRAGMA_REDEFINE_EXTNAME */ + +static int +sigwait(const sigset_t *__setp, int *__signo) +{ + extern int __posix_sigwait(const sigset_t *, int *); + return (__posix_sigwait(__setp, __signo)); +} +#endif /* __PRAGMA_REDEFINE_EXTNAME */ + +#else /* (_POSIX_C_SOURCE - 0 >= 199506L) || ... */ + +#ifndef SOLARIS_NP +extern int sigwait(sigset_t *); +#endif + +#endif /* (_POSIX_C_SOURCE - 0 >= 199506L) || ... */ + + +#else /* __STDC__ */ + + +#if (_POSIX_C_SOURCE - 0 >= 199506L) || defined(_POSIX_PTHREAD_SEMANTICS) + +#ifdef __PRAGMA_REDEFINE_EXTNAME +extern int sigwait(); +#pragma redefine_extname sigwait __posix_sigwait +#else /* __PRAGMA_REDEFINE_EXTNAME */ + +static int +sigwait(__setp, __signo) + sigset_t *__setp; + int *__signo; +{ + extern int __posix_sigwait(); + return (__posix_sigwait(__setp, __signo)); +} +#endif /* __PRAGMA_REDEFINE_EXTNAME */ + +#else /* (_POSIX_C_SOURCE - 0 >= 199506L) || ... */ + +extern int sigwait(); + +#endif /* (_POSIX_C_SOURCE - 0 >= 199506L) || ... */ + +#endif /* __STDC__ */ + +#endif /* defined(__EXTENSIONS__) || (__STDC__ == 0 ... */ + +#ifdef __cplusplus +} +#endif + +#endif /* _SIGNAL_H */ diff --git a/3rdparty/pthreads/include/pthread/signal.h b/3rdparty/pthreads/include/pthread/signal.h new file mode 100644 index 0000000..faeb164 --- /dev/null +++ b/3rdparty/pthreads/include/pthread/signal.h @@ -0,0 +1,493 @@ +/* Copyright (C) 1992-2000 the Florida State University + Distributed by the Florida State University under the terms of the + GNU Library General Public License. + +This file is part of Pthreads. + +Pthreads is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public +License as published by the Free Software Foundation (version 2). + +Pthreads is distributed "AS IS" 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 Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with Pthreads; see the file COPYING. If not, write +to the Free Software Foundation, 675 Mass Ave, Cambridge, +MA 02139, USA. + +Report problems and direct all questions to: + + pthreads-bugs@ada.cs.fsu.edu + + @(#)signal.h 3.14 11/8/00 + +*/ + +#ifndef _pthread_signal_h +#define _pthread_signal_h + +#ifndef __signal_h + +#ifdef LOCORE +#undef LOCORE +#endif + +#ifdef SVR4_NP +#ifdef SOLARIS_NP + +#include + +#if RELEASE_NP > 56 + +/* from */ +#ifndef _SIGVAL +#define _SIGVAL +union sigval { + int sival_int; /* integer value */ + void *sival_ptr; /* pointer value */ +}; +#endif /* _SIGVAL */ + +/* from : */ +#ifndef _SIGEVENT +#define _SIGEVENT +struct sigevent { + int sigev_notify; /* notification mode */ + int sigev_signo; /* signal number */ + union sigval sigev_value; /* signal value */ + void (*sigev_notify_function)(union sigval); +#ifdef SOLARIS_NP + int *sigev_notify_attributes; +#else + pthread_attr_t *sigev_notify_attributes; +#endif + int __sigev_pad2; +}; +#endif /* _SIGEVENT */ + +#define pthread_t int + +#endif + +#include "types.h" +#undef pthread_t + +#endif +#include +#include +#else /* !SVR4_NP */ +#include "stdtypes.h" +#endif /* !SVR4_NP */ + +#ifdef _M_UNIX +#if defined(__STDC__) && !defined(SCO5_NP) +#undef __STDC__ +#include +#define __STDC__ 10 +#else /* !__STDC__ */ +#include +#endif /* !__STDC__ */ +#else /* !_M_UNIX */ +#if defined(__linux__) && !defined(__KERNEL__) +#define _BITS_PTHREADTYPES_H +#define _BITS_SIGTHREAD_H +#include +#include +#include +#ifndef __USE_BSD +#define __USE_BSD +#endif +#include +#else /* !__KERNEL__ */ +#ifdef SOLARIS_NP +#include "signal-sol.h" +#else +#include +#endif +#endif /* !__KERNEL__ */ +#endif /* !_M_UNIX */ + +#if !defined(__FreeBSD__) && !defined(_M_UNIX) && !defined(__linux__) && !defined (__dos__) +#if !defined(__signal_h) && !defined(_SIGNAL_H) && !defined(__SIGNAL_H) +#define __signal_h + +#ifndef _sys_signal_h +typedef unsigned int sigset_t; +#endif + +struct sigaction { + void (*sa_handler)(); + sigset_t sa_mask; + int sa_flags; +}; +#endif /* __signal_h */ +#endif /* __FreeBSD__ */ + +#ifndef SIGMAX +#define SIGMAX _NSIG +#endif + +#ifndef NSIG +#define NSIG SIGMAX +#endif + +#ifdef __SIGRTMIN +#define NNSIG __SIGRTMIN+1 +#else +#define NNSIG NSIG+1 +#endif + +#ifndef CLOCK_REALTIME +#define CLOCK_REALTIME 0 + +#if defined(__FreeBSD__) +#include +#elif defined(__dos__) +#include +#endif + +#define ts_sec tv_sec +#define ts_nsec tv_nsec + +#ifndef TIMEVAL_TO_TIMESPEC +#if !defined(__linux__) || !defined(_ASMi386_SIGCONTEXT_H) +#if !defined(__SI_MAX_SIZE) && !defined(_STRUCT_TIMESPEC) +struct timespec { + time_t tv_sec; + long tv_nsec; +}; +#endif +#endif /* !defined(__linux) || !defined(SIGCLD) */ +#endif /* !TIMEVAL_TO_TIMESPEC */ +#else /* CLOCK_REALTIME */ +#define ts_sec tv_sec +#define ts_nsec tv_nsec +#endif /* CLOCK_REALTIME */ + +#define PTHREAD_CANCEL_ENABLE SIG_UNBLOCK +#define PTHREAD_CANCEL_DISABLE SIG_BLOCK +#define PTHREAD_CANCEL_DEFERRED 0 +#define PTHREAD_CANCEL_ASYNCHRONOUS 1 + +#ifdef si_value +#undef si_value +#endif + +union p_sigval { + int sigval_int; + void *sigval_ptr; +}; + +struct p_siginfo { + int si_signo; + int si_code; + union p_sigval si_value; +}; + +/* + * This defines the implementation-dependent context structure provided + * as the third parameter to user handlers installed by sigaction(). + * It should be a copy of the first part of the BSD sigcontext structure. + * The second half should not be accessed since it is only present if + * a _sigtramp instance is present right below the user handler on the + * thread's stack. For SVR4, we will have to build this structure from scratch. + */ + +#ifdef SOLARIS_NP + +#include + +struct context_t { + u_long sc_flags; + struct ucontext *sc_link; + sigset_t sc_mask; /* per-thread signal mask to be restored */ + stack_t sc_stack; + int sc_filler; + greg_t sc_psr; + greg_t sc_pc; /* program counter to be restored */ + greg_t sc_npc; /* next pc (see below) */ + greg_t sc_y; + greg_t sc_g1; + greg_t sc_g2; + greg_t sc_g3; + greg_t sc_g4; + greg_t sc_g5; + greg_t sc_g6; + greg_t sc_g7; + greg_t sc_o0; + greg_t sc_o1; + greg_t sc_o2; + greg_t sc_o3; + greg_t sc_o4; + greg_t sc_o5; + greg_t sc_sp; /* stack pointer to be restored */ + greg_t sc_o7; +}; + +#else /* !SOLARIS_NP */ +#ifdef SVR4_NP + +struct context_t { + This needs to be defined ! + }; + +#else /* !SVR4_NP */ + +#if defined(__FreeBSD__) +#define context_t sigcontext +/*typedef struct sigcontext context_t;*/ +#define p_sigval sigval +#define siginfo p_siginfo +typedef struct siginfo siginfo_t; +#elif defined(__dos__) +#include + +#ifdef __DJGPP__ +#include +#endif +#if __DJGPP__ >= 2 && __DJGPP_MINOR__ >= 5 +#define context_t __jmp_buf +#else +/* Hate to do it, but since the dj include files don't define this + in such a way that I can get to it, I have to repeat the definition + here. This is subject to change with new releases of djgpp; let the + buyer beware. */ + +struct context_t { + unsigned long __eax, __ebx, __ecx, __edx, __esi; + unsigned long __edi, __ebp, __esp, __eip, __eflags; + unsigned short __cs, __ds, __es, __fs, __gs, __ss; + unsigned long __sigmask; /* for POSIX signals only */ + unsigned long __signum; /* for expansion */ + unsigned char __fpu_state[108]; /* for future use */ +}; +#endif + +#define sc_sp __esp +#define sc_fp __ebp +#define sc_pc __eip +#define sc_ps __eflags +#define sc_mask __sigmask +#define p_sigval sigval +#define siginfo p_siginfo +typedef struct siginfo siginfo_t; +#elif defined(__linux__) +#define context_t sigcontext_struct +#define sc_sp esp +#define sc_fp ebp +#define sc_pc eip +#define sc_ps eflags +#define sc_mask oldmask +#define p_sigval sigval +#define siginfo p_siginfo +#define FC_PROT 1 +#ifndef __SI_MAX_SIZE +typedef struct siginfo siginfo_t; +#endif + +#undef sigmask +#undef sigemptyset +#undef sigfillset +#undef sigaddset +#undef sigdelset +#undef sigismember + +/* + * Linux 2.x already defines these operations but we need to + * use the __sig*() interface instead of the system call since e.g. + * __sigaction() uses POSIX sigset_t (32 ints) while + * SYS_sigaction uses Linux kernel sigset_t (4 bytes). + */ +#if defined(_SIGSET_NWORDS) +#define sigmask(n) ((unsigned int)1 << ((n) - 1)) +#define sigemptyset(set) ((set)->__val[0] = 0) +#define sigfillset(set) ((set)->__val[0] = -1) +#define sigaddset(set, signo) ((set)->__val[0] |= sigmask(signo)) +#define sigdelset(set, signo) ((set)->__val[0] &= ~sigmask(signo)) +#define sigismember(set, signo) ((set)->__val[0] & sigmask(signo)) +#else +#include +#define sigmask(n) ((unsigned int)1 << ((n) - 1)) +#define sigemptyset(set) (*(set) = (sigset_t) 0) +#define sigfillset(set) (*(set) = (sigset_t) -1) +#define sigaddset(set, signo) (*(set) |= sigmask(signo)) +#define sigdelset(set, signo) (*(set) &= ~sigmask(signo)) +#define sigismember(set, signo) (*(set) & sigmask(signo)) +#endif + +#elif defined (_M_UNIX) +#if defined(SCO5_NP) +#include +#include +#ifndef _GREG_T +#define _GREG_T +typedef long greg_t; +#endif + +#define context_t ucontext + +# define sc_sp uc_mcontext.regs[UESP] +# define sc_fp uc_mcontext.regs[EBP] +# define sc_pc uc_mcontext.regs[EIP] +# define sc_ps uc_flags +# define sc_mask uc_sigmask +#else /* !SCO5_NP */ +#define context_t sigcontext +#define p_sigval sigval +#define siginfo p_siginfo +#ifndef __M_UNIX__ +typedef struct siginfo siginfo_t; +#endif + +#if 0 +/* + * Location of the users' stored registers relative to EAX. + * Usage is u.u_ar0[XX]. + * + * NOTE: ERR is the error code. + */ + +#define SCO_SS 18 +#define SCO_UESP 17 +#define SCO_EFL 16 +#define SCO_CS 15 +#define SCO_EIP 14 +#define SCO_ERR 13 +#define SCO_TRAPNO 12 +#define SCO_EAX 11 +#define SCO_ECX 10 +#define SCO_EDX 9 +#define SCO_EBX 8 +#define SCO_ESP 7 +#define SCO_EBP 6 +#define SCO_ESI 5 +#define SCO_EDI 4 +#define SCO_DS 3 +#define SCO_ES 2 +#define SCO_FS 1 +#define SCO_GS 0 +#endif + + +# define sc_sp sc_UESP +# define sc_fp sc_EBP +# define sc_pc sc_EIP +# define sc_ps sc_EFL + +struct sigcontext { + unsigned int sc_GS; /* 0 */ + unsigned int sc_FS; /* 1 */ + unsigned int sc_ES; /* 2 */ + unsigned int sc_DS; /* 3 */ + unsigned int sc_EDI; /* 4 */ + unsigned int sc_ESI; /* 5 */ + unsigned int sc_EBP; /* 6 */ + unsigned int sc_ESP; /* 7 */ + unsigned int sc_EBX; /* 8 */ + unsigned int sc_EDX; /* 9 */ + unsigned int sc_ECX; /* 10 */ + unsigned int sc_EAX; /* 11 */ + unsigned int sc_TRAPNO; /* 12 */ + unsigned int sc_ERR; /* 13 */ + unsigned int sc_EIP; /* 14 */ + unsigned int sc_CS; /* 15 */ + unsigned int sc_EFL; /* 16 */ + unsigned int sc_UESP; /* 17 */ + unsigned int sc_SS; /* 18 */ + unsigned int *piFPContext; + unsigned int *weitek_offset; + sigset_t sc_mask; +}; + + +/* + * Send an interrupt to process. + * Pass the signal number as an argument to the user signal handler. + * Also pushed on the user stack is all of the user registers and the + * ret addr for a piece of code that calls back into the kernel + * (after executing the user's signal handler) to restore the + * context of the user process. Last, but not least, put on the + * signal mask, for sigclean to restore after execution of the + * handler (POSIX). + */ + +/* +** From Nick Logan: +** After the signal number, the register are pushed onto the user stack in an +** array indexed by the values in /usr/include/sys/reg.h. If an FPU is used the +** FP registers are then saved on the stack. +** +** Note that this could change in any release, and you use it at your +** own risk! +*/ +#endif /* !SCO5_NP */ +#else +#define p_sigval sigval +#define siginfo p_siginfo +typedef struct siginfo siginfo_t; + +typedef int greg_t; + +struct context_t { + greg_t sc_onstack;/* ignored */ + sigset_t sc_mask; /* per-thread signal mask to be restored */ + greg_t sc_sp; /* stack pointer to be restored */ + greg_t sc_pc; /* program counter to be restored */ + greg_t sc_npc; /* next pc, only used if _sigtramp present + * on thread's stack, ignored o.w. + * should usually be pc+4 + */ + greg_t sc_g1; + greg_t sc_o0; +}; + +#endif +#endif /* !SVR4_NP */ +#endif /* !SOLARIS_NP */ + +#ifndef SA_SIGINFO +#define SA_SIGINFO 0 +#endif + +#ifndef SA_ONESHOT +#define SA_ONESHOT 0 +#endif + +#ifndef SA_NOMASK +#define SA_NOMASK 0 +#endif + +#ifndef SA_ONSTACK +#define SA_ONSTACK SV_ONSTACK +#endif /* !SA_ONSTACK */ + +#ifndef BUS_OBJERR +#define BUS_OBJERR FC_OBJERR +#endif /* !BUS_OBJERR */ + +#ifndef BUS_CODE +#define BUS_CODE(x) FC_CODE(x) +#endif + +#ifndef SIGCANCEL +#ifdef __dos__ +#define SIGCANCEL NSIG-1 /* To get arounnd a bug in djgpp v2 */ +#else +#ifdef SCO5 +#define SIGCANCEL (NSIG-1) +#else +#ifdef __SIGRTMIN +#define SIGCANCEL __SIGRTMIN +#else +#define SIGCANCEL NSIG +#endif +#endif +#endif +#endif + +#endif /* __signal_h */ + +#endif /* !_pthread_signal_h */ diff --git a/3rdparty/pthreads/include/pthread/stdtypes.h b/3rdparty/pthreads/include/pthread/stdtypes.h new file mode 100644 index 0000000..488ed53 --- /dev/null +++ b/3rdparty/pthreads/include/pthread/stdtypes.h @@ -0,0 +1,82 @@ +/* Copyright (C) 1992-2000 the Florida State University + Distributed by the Florida State University under the terms of the + GNU Library General Public License. + +This file is part of Pthreads. + +Pthreads is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public +License as published by the Free Software Foundation (version 2). + +Pthreads is distributed "AS IS" 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 Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with Pthreads; see the file COPYING. If not, write +to the Free Software Foundation, 675 Mass Ave, Cambridge, +MA 02139, USA. + +Report problems and direct all questions to: + + pthreads-bugs@ada.cs.fsu.edu + + @(#)stdtypes.h 3.14 11/8/00 + +*/ + +#ifndef _pthread_stdtypes_h +#define _pthread_stdtypes_h + +#if !defined(__FreeBSD__) && !defined(_M_UNIX) && !defined(__linux__) && !defined(__dos__) +/* we don't want any of these since we already have them + */ +#ifndef __sys_stdtypes_h +#define __sys_stdtypes_h + +#if defined(STAND_ALONE_NP) && defined(sun4e_NP) +#ifndef _TYPES_ +#define _TYPES_ + +/* + * substitutes used types of #include + * to avoid wrong definition of size_t + */ + +typedef unsigned char u_char; +typedef unsigned int u_int; +typedef char * caddr_t; +typedef char * addr_t; + +#endif /* !_TYPES_ */ +#endif /* STAND_ALONE_NP && sun4e_NP */ + +typedef unsigned int sigset_t; /* signal mask - may change */ + +typedef unsigned int speed_t; /* tty speeds */ +typedef unsigned long tcflag_t; /* tty line disc modes */ +typedef unsigned char cc_t; /* tty control char */ +typedef int pid_t; /* process id */ + +typedef unsigned short mode_t; /* file mode bits */ +typedef short nlink_t; /* links to a file */ + +typedef long clock_t; /* units=ticks (typically 60/sec) */ +typedef long time_t; /* value = secs since epoch */ + +#ifdef __GNUC__ +typedef long unsigned int size_t; /* ??? */ +#else /* !__GNUC__ */ +typedef unsigned int size_t; /* ??? */ +#endif /* !__GNUC__ */ +typedef int ptrdiff_t; /* result of subtracting two pointers */ + +typedef unsigned short wchar_t; /* big enough for biggest char set */ + +#endif /* !__sys_stdtypes_h */ +#elif defined (_M_UNIX) +#include +#endif + +#endif /* !_pthread_stdtypes_h */ diff --git a/3rdparty/pthreads/include/pthread/types.h b/3rdparty/pthreads/include/pthread/types.h new file mode 100644 index 0000000..2edffff --- /dev/null +++ b/3rdparty/pthreads/include/pthread/types.h @@ -0,0 +1,577 @@ +/* Copyright (c) 1984, 1986, 1987, 1988, 1989 AT&T */ +/* All Rights Reserved */ + +/* THIS IS UNPUBLISHED PROPRIETARY SOURCE CODE OF AT&T */ +/* The copyright notice above does not evidence any */ +/* actual or intended publication of such source code. */ + +/* + * Copyright (c) 1996-1999 by Sun Microsystems, Inc. + * All rights reserved. + */ + +#ifndef _SYS_TYPES_H +#define _SYS_TYPES_H + +#pragma ident "@(#)types.h 1.65 99/11/15 SMI" + +#include +#include + +/* + * Machine dependent definitions moved to . + */ +#include + +/* + * Include fixed width type declarations proposed by the ISO/JTC1/SC22/WG14 C + * committee's working draft for the revision of the current ISO C standard, + * ISO/IEC 9899:1990 Programming language - C. These are not currently + * required by any standard but constitute a useful, general purpose set + * of type definitions which is namespace clean with respect to all standards. + */ +#ifdef SOLARIS_NP +#ifdef __STDC__ +#undef __STDC__ +#define __STDC__NP +#endif +#define __STDC__ 0 +#include "int_types.h" +#ifndef _FILE_OFFSET_BITS +#define _FILE_OFFSET_BITS 32 +#endif +#else +#ifdef _KERNEL +#include +#else /* _KERNEL */ +#include +#endif /* _KERNEL */ +#endif + +#if defined(_KERNEL) || defined(_SYSCALL32) +#include +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +/* + * The following protects users who use other than Sun compilers + * (eg, GNU C) that don't support long long, and need to include + * this header file. + */ +#if __STDC__ - 0 == 0 && !defined(_NO_LONGLONG) +typedef long long longlong_t; +typedef unsigned long long u_longlong_t; +#else +/* used to reserve space and generate alignment */ +typedef union { + double _d; + int32_t _l[2]; +} longlong_t; +typedef union { + double _d; + uint32_t _l[2]; +} u_longlong_t; +#endif /* __STDC__ - 0 == 0 && !defined(_NO_LONGLONG) */ + +/* + * These types (t_{u}scalar_t) exist because the XTI/TPI/DLPI standards had + * to use them instead of int32_t and uint32_t because DEC had + * shipped 64-bit wide. + */ +#if defined(_LP64) || defined(_I32LPx) +typedef int32_t t_scalar_t; +typedef uint32_t t_uscalar_t; +#else +typedef long t_scalar_t; /* historical versions */ +typedef unsigned long t_uscalar_t; +#endif /* defined(_LP64) || defined(_I32LPx) */ + +/* + * POSIX Extensions + */ +typedef unsigned char uchar_t; +typedef unsigned short ushort_t; +typedef unsigned int uint_t; +typedef unsigned long ulong_t; + +typedef char *caddr_t; /* ? type */ +typedef long daddr_t; /* type */ +typedef short cnt_t; /* ? type */ + +#if defined(_ILP32) /* only used in i386 code */ +typedef ulong_t paddr_t; /* type */ +#elif defined(__ia64) /* XXX Fix me */ +typedef uint_t paddr_t; +#endif + +#ifndef _PTRDIFF_T +#define _PTRDIFF_T +#if defined(_LP64) || defined(_I32LPx) +typedef long ptrdiff_t; /* pointer difference */ +#else +typedef int ptrdiff_t; /* (historical version) */ +#endif +#endif + +/* + * VM-related types + */ +typedef ulong_t pfn_t; /* page frame number */ +typedef ulong_t pgcnt_t; /* number of pages */ +typedef long spgcnt_t; /* signed number of pages */ + +typedef uchar_t use_t; /* use count for swap. */ +typedef short sysid_t; +typedef short index_t; +typedef void *timeout_id_t; /* opaque handle from timeout(9F) */ +typedef void *bufcall_id_t; /* opaque handle from bufcall(9F) */ + +/* + * The size of off_t and related types depends on the setting of + * _FILE_OFFSET_BITS. (Note that other system headers define other types + * related to those defined here.) + * + * If _LARGEFILE64_SOURCE is defined, variants of these types that are + * explicitly 64 bits wide become available. + */ +#ifndef _OFF_T +#define _OFF_T + +#if defined(_LP64) || _FILE_OFFSET_BITS == 32 +typedef long off_t; /* offsets within files */ +#elif _FILE_OFFSET_BITS == 64 +typedef longlong_t off_t; /* offsets within files */ +#endif + +#if defined(_LARGEFILE64_SOURCE) +#ifdef _LP64 +typedef off_t off64_t; /* offsets within files */ +#else +typedef longlong_t off64_t; /* offsets within files */ +#endif +#endif /* _LARGEFILE64_SOURCE */ + +#endif /* _OFF_T */ + +#if defined(_LP64) || _FILE_OFFSET_BITS == 32 +typedef ulong_t ino_t; /* expanded inode type */ +typedef long blkcnt_t; /* count of file blocks */ +typedef ulong_t fsblkcnt_t; /* count of file system blocks */ +typedef ulong_t fsfilcnt_t; /* count of files */ +#elif _FILE_OFFSET_BITS == 64 +typedef u_longlong_t ino_t; /* expanded inode type */ +typedef longlong_t blkcnt_t; /* count of file blocks */ +typedef u_longlong_t fsblkcnt_t; /* count of file system blocks */ +typedef u_longlong_t fsfilcnt_t; /* count of files */ +#endif + +#if defined(_LARGEFILE64_SOURCE) +#ifdef _LP64 +typedef ino_t ino64_t; /* expanded inode type */ +typedef blkcnt_t blkcnt64_t; /* count of file blocks */ +typedef fsblkcnt_t fsblkcnt64_t; /* count of file system blocks */ +typedef fsfilcnt_t fsfilcnt64_t; /* count of files */ +#else +typedef u_longlong_t ino64_t; /* expanded inode type */ +typedef longlong_t blkcnt64_t; /* count of file blocks */ +typedef u_longlong_t fsblkcnt64_t; /* count of file system blocks */ +typedef u_longlong_t fsfilcnt64_t; /* count of files */ +#endif +#endif /* _LARGEFILE64_SOURCE */ + +#ifdef _LP64 +typedef int blksize_t; /* used for block sizes */ +#else +typedef long blksize_t; /* used for block sizes */ +#endif + +#if defined(_POSIX_C_SOURCE) || defined(_XOPEN_SOURCE) +typedef enum { _B_FALSE, _B_TRUE } boolean_t; +#else +typedef enum { B_FALSE, B_TRUE } boolean_t; +#endif /* defined(_POSIX_C_SOURCE) || defined(_XOPEN_SOURCE) */ + +/* + * The [u]pad64_t is to be used in structures such that those structures + * may be accessed by code produced by compilation environments which don't + * support a 64 bit integral datatype. This intention is not to allow + * use of these fields in such environments, but to maintain the alignment + * and offsets of the structure. + */ +#if __STDC__ - 0 == 0 && !defined(_NO_LONGLONG) +typedef int64_t pad64_t; +typedef uint64_t upad64_t; +#else +typedef union { + double _d; + int32_t _l[2]; +} pad64_t; +typedef union { + double _d; + uint32_t _l[2]; +} upad64_t; +#endif + +typedef longlong_t offset_t; +typedef u_longlong_t u_offset_t; +typedef u_longlong_t len_t; +typedef longlong_t diskaddr_t; + +/* + * Definitions remaining from previous partial support for 64-bit file + * offsets. This partial support for devices greater than 2gb requires + * compiler support for long long. + */ +#ifdef _LONG_LONG_LTOH +typedef union { + offset_t _f; /* Full 64 bit offset value */ + struct { + int32_t _l; /* lower 32 bits of offset value */ + int32_t _u; /* upper 32 bits of offset value */ + } _p; +} lloff_t; +#endif + +#ifdef _LONG_LONG_HTOL +typedef union { + offset_t _f; /* Full 64 bit offset value */ + struct { + int32_t _u; /* upper 32 bits of offset value */ + int32_t _l; /* lower 32 bits of offset value */ + } _p; +} lloff_t; +#endif + +#ifdef _LONG_LONG_LTOH +typedef union { + diskaddr_t _f; /* Full 64 bit disk address value */ + struct { + int32_t _l; /* lower 32 bits of disk address value */ + int32_t _u; /* upper 32 bits of disk address value */ + } _p; +} lldaddr_t; +#endif + +#ifdef _LONG_LONG_HTOL +typedef union { + diskaddr_t _f; /* Full 64 bit disk address value */ + struct { + int32_t _u; /* upper 32 bits of disk address value */ + int32_t _l; /* lower 32 bits of disk address value */ + } _p; +} lldaddr_t; +#endif + +typedef uint_t k_fltset_t; /* kernel fault set type */ + +/* + * The following type is for various kinds of identifiers. The + * actual type must be the same for all since some system calls + * (such as sigsend) take arguments that may be any of these + * types. The enumeration type idtype_t defined in sys/procset.h + * is used to indicate what type of id is being specified. + */ +#if defined(_LP64) || defined(_I32LPx) +typedef int id_t; /* A process id, */ + /* process group id, */ + /* session id, */ + /* scheduling class id, */ + /* user id, or group id */ +#else +typedef long id_t; /* (historical version) */ +#endif + +/* + * Type useconds_t is an unsigned integral type capable of storing + * values at least in the range of zero to 1,000,000. + */ +typedef uint_t useconds_t; /* Time, in microseconds */ + +#ifndef _SUSECONDS_T +#define _SUSECONDS_T +typedef long suseconds_t; /* signed # of microseconds */ +#endif /* _SUSECONDS_T */ + +/* + * Typedefs for dev_t components. + */ +#if defined(_LP64) || defined(_I32LPx) +typedef uint_t major_t; /* major part of device number */ +typedef uint_t minor_t; /* minor part of device number */ +#else +typedef ulong_t major_t; /* (historical version) */ +typedef ulong_t minor_t; /* (historical version) */ +#endif + +/* + * The data type of a thread priority. + */ +typedef short pri_t; + +/* + * For compatibility reasons the following typedefs (prefixed o_) + * can't grow regardless of the EFT definition. Although, + * applications should not explicitly use these typedefs + * they may be included via a system header definition. + * WARNING: These typedefs may be removed in a future + * release. + * ex. the definitions in s5inode.h remain small + * to preserve compatibility in the S5 + * file system type. + */ +typedef ushort_t o_mode_t; /* old file attribute type */ +typedef short o_dev_t; /* old device type */ +typedef ushort_t o_uid_t; /* old UID type */ +typedef o_uid_t o_gid_t; /* old GID type */ +typedef short o_nlink_t; /* old file link type */ +typedef short o_pid_t; /* old process id type */ +typedef ushort_t o_ino_t; /* old inode type */ + + +/* + * POSIX and XOPEN Declarations + */ +typedef int key_t; /* IPC key type */ +#if defined(_LP64) || defined(_I32LPx) +typedef uint_t mode_t; /* file attribute type */ +#else +typedef ulong_t mode_t; /* (historical version) */ +#endif + +#ifndef _UID_T +#define _UID_T +#if defined(_LP64) || defined(_I32LPx) +typedef int uid_t; /* UID type */ +#else +typedef long uid_t; /* (historical version) */ +#endif +#endif /* _UID_T */ + +typedef uid_t gid_t; /* GID type */ + +/* + * POSIX definitions are same as defined in thread.h and synch.h. + * Any changes made to here should be reflected in corresponding + * files as described in comments. + */ +#ifndef SOLARIS_NP +typedef unsigned int pthread_t; /* = thread_t in thread.h */ +typedef unsigned int pthread_key_t; /* = thread_key_t in thread.h */ + +typedef struct _pthread_mutex { /* = mutex_t in synch.h */ + struct { + uint16_t __pthread_mutex_flag1; + uint8_t __pthread_mutex_flag2; + uint8_t __pthread_mutex_ceiling; + uint32_t __pthread_mutex_type; + } __pthread_mutex_flags; + union { + struct { + uint8_t __pthread_mutex_pad[8]; + } __pthread_mutex_lock64; + upad64_t __pthread_mutex_owner64; + } __pthread_mutex_lock; + upad64_t __pthread_mutex_data; +} pthread_mutex_t; + +typedef struct _pthread_cond { /* = cond_t in synch.h */ + struct { + uint8_t __pthread_cond_flag[4]; + uint32_t __pthread_cond_type; + } __pthread_cond_flags; + upad64_t __pthread_cond_data; +} pthread_cond_t; + +/* + * UNIX 98 Extension + */ +typedef struct _pthread_rwlock { /* = rwlock_t in synch.h */ + int32_t __pthread_rwlock_readers; + uint16_t __pthread_rwlock_type; + uint16_t __pthread_rwlock_magic; + upad64_t __pthread_rwlock_pad1[3]; + upad64_t __pthread_rwlock_pad2[2]; + upad64_t __pthread_rwlock_pad3[2]; +} pthread_rwlock_t; + +/* + * attributes for threads, dynamically allocated by library + */ +typedef struct _pthread_attr { + void *__pthread_attrp; +} pthread_attr_t; + + +/* + * attributes for mutex, dynamically allocated by library + */ +typedef struct _pthread_mutexattr { + void *__pthread_mutexattrp; +} pthread_mutexattr_t; + + +/* + * attributes for cond, dynamically allocated by library + */ +typedef struct _pthread_condattr { + void *__pthread_condattrp; +} pthread_condattr_t; + +/* + * pthread_once + */ +typedef struct _once { + upad64_t __pthread_once_pad[4]; +} pthread_once_t; + +/* + * UNIX 98 Extensions + * attributes for rwlock, dynamically allocated by library + */ +typedef struct _pthread_rwlockattr { + void *__pthread_rwlockattrp; +} pthread_rwlockattr_t; +#endif + +typedef ulong_t dev_t; /* expanded device type */ + +#if defined(_LP64) || defined(_I32LPx) +typedef uint_t nlink_t; /* file link type */ +typedef int pid_t; /* process id type */ +#else +typedef ulong_t nlink_t; /* (historical version) */ +typedef long pid_t; /* (historical version) */ +#endif + +#ifndef _SIZE_T +#define _SIZE_T +#if defined(_LP64) || defined(_I32LPx) +typedef ulong_t size_t; /* size of something in bytes */ +#else +typedef uint_t size_t; /* (historical version) */ +#endif +#endif /* _SIZE_T */ + +#ifndef _SSIZE_T +#define _SSIZE_T +#if defined(_LP64) || defined(_I32LPx) +typedef long ssize_t; /* size of something in bytes or -1 */ +#else +typedef int ssize_t; /* (historical version) */ +#endif +#endif /* _SSIZE_T */ + +#ifndef _TIME_T +#define _TIME_T +typedef long time_t; /* time of day in seconds */ +#endif /* _TIME_T */ + +#ifndef _CLOCK_T +#define _CLOCK_T +typedef long clock_t; /* relative time in a specified resolution */ +#endif /* ifndef _CLOCK_T */ + +#ifndef _CLOCKID_T +#define _CLOCKID_T +typedef int clockid_t; /* clock identifier type */ +#endif /* ifndef _CLOCKID_T */ + +#ifndef _TIMER_T +#define _TIMER_T +typedef int timer_t; /* timer identifier type */ +#endif /* ifndef _TIMER_T */ + +#if defined(__EXTENSIONS__) || \ + (!defined(_POSIX_C_SOURCE) && !defined(_XOPEN_SOURCE)) + +/* BEGIN CSTYLED */ +typedef unsigned char unchar; +typedef unsigned short ushort; +typedef unsigned int uint; +typedef unsigned long ulong; +/* END CSTYLED */ + +#if defined(_KERNEL) + +#define SHRT_MIN (-32768) /* min value of a "short int" */ +#define SHRT_MAX 32767 /* max value of a "short int" */ +#define USHRT_MAX 65535 /* max of "unsigned short int" */ +#define INT_MIN (-2147483647-1) /* min value of an "int" */ +#define INT_MAX 2147483647 /* max value of an "int" */ +#define UINT_MAX 4294967295U /* max value of an "unsigned int" */ +#if defined(_LP64) +#define LONG_MIN (-9223372036854775807L-1L) + /* min value of a "long int" */ +#define LONG_MAX 9223372036854775807L + /* max value of a "long int" */ +#define ULONG_MAX 18446744073709551615UL + /* max of "unsigned long int" */ +#else /* _ILP32 */ +#define LONG_MIN (-2147483647L-1L) + /* min value of a "long int" */ +#define LONG_MAX 2147483647L /* max value of a "long int" */ +#define ULONG_MAX 4294967295UL /* max of "unsigned long int" */ +#endif + +#endif /* defined(_KERNEL) */ + +#define P_MYPID ((pid_t)0) + +/* + * The following is the value of type id_t to use to indicate the + * caller's current id. See procset.h for the type idtype_t + * which defines which kind of id is being specified. + */ +#define P_MYID (-1) +#define NOPID (pid_t)(-1) + +#ifndef NODEV +#define NODEV (dev_t)(-1l) +#ifdef _SYSCALL32 +#define NODEV32 (dev32_t)(-1) +#endif /* _SYSCALL32 */ +#endif /* NODEV */ + +/* + * The following value of type pfn_t is used to indicate + * invalid page frame number. + */ +#define PFN_INVALID ((pfn_t)-1) + +/* BEGIN CSTYLED */ +typedef unsigned char u_char; +typedef unsigned short u_short; +typedef unsigned int u_int; +typedef unsigned long u_long; +typedef struct _quad { int val[2]; } quad_t; /* used by UFS */ +typedef quad_t quad; /* used by UFS */ +/* END CSTYLED */ + +/* + * Nested include for BSD/sockets source compatibility. + * (The select macros used to be defined here). + */ +#include + +#endif /* defined(__EXTENSIONS__) || (!defined(_POSIX_C_SOURCE) && ... */ + +/* + * _VOID was defined to be either void or char but this is not + * required because previous SunOS compilers have accepted the void + * type. However, because many system header and source files use the + * void keyword, the volatile keyword, and ANSI C function prototypes, + * non-ANSI compilers cannot compile the system anyway. The _VOID macro + * should therefore not be used and remains for source compatibility + * only. + */ +/* CSTYLED */ +#define _VOID void + +#ifdef __cplusplus +} +#endif + +#endif /* _SYS_TYPES_H */ diff --git a/3rdparty/pthreads/include/pthread/unistd.h b/3rdparty/pthreads/include/pthread/unistd.h new file mode 100644 index 0000000..f8c9eb6 --- /dev/null +++ b/3rdparty/pthreads/include/pthread/unistd.h @@ -0,0 +1,83 @@ +/* Copyright (C) 1992-2000 the Florida State University + Distributed by the Florida State University under the terms of the + GNU Library General Public License. + +This file is part of Pthreads. + +Pthreads is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public +License as published by the Free Software Foundation (version 2). + +Pthreads is distributed "AS IS" 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 Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with Pthreads; see the file COPYING. If not, write +to the Free Software Foundation, 675 Mass Ave, Cambridge, +MA 02139, USA. + +Report problems and direct all questions to: + + pthreads-bugs@ada.cs.fsu.edu + + @(#)unistd.h 3.14 11/8/00 + +*/ + +#ifndef _pthread_unistd_h +#define _pthread_unistd_h + +#ifdef LOCORE + +#if defined(__linux__) || defined(__dos__) +#ifndef __ELF__ +#define NAME(x) _/**/x +#else +#define NAME(x) x +#endif +#endif /* __linux__ */ + +#else /* !LOCORE */ + +#include + +/* + * We need to figure out (from header files) what system we are running on. + */ +#if !defined(_M_UNIX) && !defined(__linux__) && !defined(__dos__) +#ifdef _UNISTD_H +#ifndef SOLARIS_NP +#define SOLARIS_NP +#define SVR4_NP +#endif +#elif defined(SOLARIS_NP) || defined(SVR4_NP) +#undef SOLARIS_NP +#undef SVR4_NP +#endif /* _UNISTD_H */ +#endif /* !_M_UNIX && !__linux__ && !__dos__ */ + +#endif /* !LOCORE */ + +#ifndef _POSIX_THREADS +#define _POSIX_THREADS +#endif /* !_POSIX_THREADS */ +#ifndef _POSIX_THREAD_PRIORITY_SCHEDULING +#define _POSIX_THREAD_PRIORITY_SCHEDULING +#endif /* !_POSIX_THREAD_PRIORITY_SCHEDULING */ +#ifndef _POSIX_THREAD_ATTR_STACKSIZE +#define _POSIX_THREAD_ATTR_STACKSIZE +#endif /* !_POSIX_THREAD_ATTR_STACKSIZE */ +#ifdef SRP_NP +#ifndef _POSIX_THREADS_PRIO_PROTECT +#define _POSIX_THREADS_PRIO_PROTECT +#endif /* !_POSIX_THREADS_PRIO_PROTECT */ +#else /* !SRP_NP */ +#undef _POSIX_THREADS_PRIO_PROTECT +#endif /* !SRP_NP */ +#undef _POSIX_THREADS_PROCESS_SHARED +#undef _POSIX_THREADS_PRIO_INHERIT +#undef _POSIX_REENTRANT_FUNCTIONS + +#endif /* !_pthread_unistd_h */ diff --git a/3rdparty/pthreads/lib/libgthreads.a b/3rdparty/pthreads/lib/libgthreads.a new file mode 100644 index 0000000..40a257d --- /dev/null +++ b/3rdparty/pthreads/lib/libgthreads.a @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9a7fe7becc1048479fb97851bfa2804e099962ac082ddb7d6a134a33c278c6b5 +size 90728 diff --git a/3rdparty/stb_ds.h b/3rdparty/stb_ds.h new file mode 100644 index 0000000..d69a682 --- /dev/null +++ b/3rdparty/stb_ds.h @@ -0,0 +1,1895 @@ +/* stb_ds.h - v0.67 - public domain data structures - Sean Barrett 2019 + + This is a single-header-file library that provides easy-to-use + dynamic arrays and hash tables for C (also works in C++). + + For a gentle introduction: + http://nothings.org/stb_ds + + To use this library, do this in *one* C or C++ file: + #define STB_DS_IMPLEMENTATION + #include "stb_ds.h" + +TABLE OF CONTENTS + + Table of Contents + Compile-time options + License + Documentation + Notes + Notes - Dynamic arrays + Notes - Hash maps + Credits + +COMPILE-TIME OPTIONS + + #define STBDS_NO_SHORT_NAMES + + This flag needs to be set globally. + + By default stb_ds exposes shorter function names that are not qualified + with the "stbds_" prefix. If these names conflict with the names in your + code, define this flag. + + #define STBDS_SIPHASH_2_4 + + This flag only needs to be set in the file containing #define STB_DS_IMPLEMENTATION. + + By default stb_ds.h hashes using a weaker variant of SipHash and a custom hash for + 4- and 8-byte keys. On 64-bit platforms, you can define the above flag to force + stb_ds.h to use specification-compliant SipHash-2-4 for all keys. Doing so makes + hash table insertion about 20% slower on 4- and 8-byte keys, 5% slower on + 64-byte keys, and 10% slower on 256-byte keys on my test computer. + + #define STBDS_REALLOC(context,ptr,size) better_realloc + #define STBDS_FREE(context,ptr) better_free + + These defines only need to be set in the file containing #define STB_DS_IMPLEMENTATION. + + By default stb_ds uses stdlib realloc() and free() for memory management. You can + substitute your own functions instead by defining these symbols. You must either + define both, or neither. Note that at the moment, 'context' will always be NULL. + @TODO add an array/hash initialization function that takes a memory context pointer. + + #define STBDS_UNIT_TESTS + + Defines a function stbds_unit_tests() that checks the functioning of the data structures. + + Note that on older versions of gcc (e.g. 5.x.x) you may need to build with '-std=c++0x' + (or equivalentally '-std=c++11') when using anonymous structures as seen on the web + page or in STBDS_UNIT_TESTS. + +LICENSE + + Placed in the public domain and also MIT licensed. + See end of file for detailed license information. + +DOCUMENTATION + + Dynamic Arrays + + Non-function interface: + + Declare an empty dynamic array of type T + T* foo = NULL; + + Access the i'th item of a dynamic array 'foo' of type T, T* foo: + foo[i] + + Functions (actually macros) + + arrfree: + void arrfree(T*); + Frees the array. + + arrlen: + ptrdiff_t arrlen(T*); + Returns the number of elements in the array. + + arrlenu: + size_t arrlenu(T*); + Returns the number of elements in the array as an unsigned type. + + arrpop: + T arrpop(T* a) + Removes the final element of the array and returns it. + + arrput: + T arrput(T* a, T b); + Appends the item b to the end of array a. Returns b. + + arrins: + T arrins(T* a, int p, T b); + Inserts the item b into the middle of array a, into a[p], + moving the rest of the array over. Returns b. + + arrinsn: + void arrinsn(T* a, int p, int n); + Inserts n uninitialized items into array a starting at a[p], + moving the rest of the array over. + + arraddnptr: + T* arraddnptr(T* a, int n) + Appends n uninitialized items onto array at the end. + Returns a pointer to the first uninitialized item added. + + arraddnindex: + size_t arraddnindex(T* a, int n) + Appends n uninitialized items onto array at the end. + Returns the index of the first uninitialized item added. + + arrdel: + void arrdel(T* a, int p); + Deletes the element at a[p], moving the rest of the array over. + + arrdeln: + void arrdeln(T* a, int p, int n); + Deletes n elements starting at a[p], moving the rest of the array over. + + arrdelswap: + void arrdelswap(T* a, int p); + Deletes the element at a[p], replacing it with the element from + the end of the array. O(1) performance. + + arrsetlen: + void arrsetlen(T* a, int n); + Changes the length of the array to n. Allocates uninitialized + slots at the end if necessary. + + arrsetcap: + size_t arrsetcap(T* a, int n); + Sets the length of allocated storage to at least n. It will not + change the length of the array. + + arrcap: + size_t arrcap(T* a); + Returns the number of total elements the array can contain without + needing to be reallocated. + + Hash maps & String hash maps + + Given T is a structure type: struct { TK key; TV value; }. Note that some + functions do not require TV value and can have other fields. For string + hash maps, TK must be 'char *'. + + Special interface: + + stbds_rand_seed: + void stbds_rand_seed(size_t seed); + For security against adversarially chosen data, you should seed the + library with a strong random number. Or at least seed it with time(). + + stbds_hash_string: + size_t stbds_hash_string(char *str, size_t seed); + Returns a hash value for a string. + + stbds_hash_bytes: + size_t stbds_hash_bytes(void *p, size_t len, size_t seed); + These functions hash an arbitrary number of bytes. The function + uses a custom hash for 4- and 8-byte data, and a weakened version + of SipHash for everything else. On 64-bit platforms you can get + specification-compliant SipHash-2-4 on all data by defining + STBDS_SIPHASH_2_4, at a significant cost in speed. + + Non-function interface: + + Declare an empty hash map of type T + T* foo = NULL; + + Access the i'th entry in a hash table T* foo: + foo[i] + + Function interface (actually macros): + + hmfree + shfree + void hmfree(T*); + void shfree(T*); + Frees the hashmap and sets the pointer to NULL. + + hmlen + shlen + ptrdiff_t hmlen(T*) + ptrdiff_t shlen(T*) + Returns the number of elements in the hashmap. + + hmlenu + shlenu + size_t hmlenu(T*) + size_t shlenu(T*) + Returns the number of elements in the hashmap. + + hmgeti + shgeti + hmgeti_ts + ptrdiff_t hmgeti(T*, TK key) + ptrdiff_t shgeti(T*, char* key) + ptrdiff_t hmgeti_ts(T*, TK key, ptrdiff_t tempvar) + Returns the index in the hashmap which has the key 'key', or -1 + if the key is not present. + + hmget + hmget_ts + shget + TV hmget(T*, TK key) + TV shget(T*, char* key) + TV hmget_ts(T*, TK key, ptrdiff_t tempvar) + Returns the value corresponding to 'key' in the hashmap. + The structure must have a 'value' field + + hmgets + shgets + T hmgets(T*, TK key) + T shgets(T*, char* key) + Returns the structure corresponding to 'key' in the hashmap. + + hmgetp + shgetp + hmgetp_ts + hmgetp_null + shgetp_null + T* hmgetp(T*, TK key) + T* shgetp(T*, char* key) + T* hmgetp_ts(T*, TK key, ptrdiff_t tempvar) + T* hmgetp_null(T*, TK key) + T* shgetp_null(T*, char *key) + Returns a pointer to the structure corresponding to 'key' in + the hashmap. Functions ending in "_null" return NULL if the key + is not present in the hashmap; the others return a pointer to a + structure holding the default value (but not the searched-for key). + + hmdefault + shdefault + TV hmdefault(T*, TV value) + TV shdefault(T*, TV value) + Sets the default value for the hashmap, the value which will be + returned by hmget/shget if the key is not present. + + hmdefaults + shdefaults + TV hmdefaults(T*, T item) + TV shdefaults(T*, T item) + Sets the default struct for the hashmap, the contents which will be + returned by hmgets/shgets if the key is not present. + + hmput + shput + TV hmput(T*, TK key, TV value) + TV shput(T*, char* key, TV value) + Inserts a pair into the hashmap. If the key is already + present in the hashmap, updates its value. + + hmputs + shputs + T hmputs(T*, T item) + T shputs(T*, T item) + Inserts a struct with T.key into the hashmap. If the struct is already + present in the hashmap, updates it. + + hmdel + shdel + int hmdel(T*, TK key) + int shdel(T*, char* key) + If 'key' is in the hashmap, deletes its entry and returns 1. + Otherwise returns 0. + + Function interface (actually macros) for strings only: + + sh_new_strdup + void sh_new_strdup(T*); + Overwrites the existing pointer with a newly allocated + string hashmap which will automatically allocate and free + each string key using realloc/free + + sh_new_arena + void sh_new_arena(T*); + Overwrites the existing pointer with a newly allocated + string hashmap which will automatically allocate each string + key to a string arena. Every string key ever used by this + hash table remains in the arena until the arena is freed. + Additionally, any key which is deleted and reinserted will + be allocated multiple times in the string arena. + +NOTES + + * These data structures are realloc'd when they grow, and the macro + "functions" write to the provided pointer. This means: (a) the pointer + must be an lvalue, and (b) the pointer to the data structure is not + stable, and you must maintain it the same as you would a realloc'd + pointer. For example, if you pass a pointer to a dynamic array to a + function which updates it, the function must return back the new + pointer to the caller. This is the price of trying to do this in C. + + * The following are the only functions that are thread-safe on a single data + structure, i.e. can be run in multiple threads simultaneously on the same + data structure + hmlen shlen + hmlenu shlenu + hmget_ts shget_ts + hmgeti_ts shgeti_ts + hmgets_ts shgets_ts + + * You iterate over the contents of a dynamic array and a hashmap in exactly + the same way, using arrlen/hmlen/shlen: + + for (i=0; i < arrlen(foo); ++i) + ... foo[i] ... + + * All operations except arrins/arrdel are O(1) amortized, but individual + operations can be slow, so these data structures may not be suitable + for real time use. Dynamic arrays double in capacity as needed, so + elements are copied an average of once. Hash tables double/halve + their size as needed, with appropriate hysteresis to maintain O(1) + performance. + +NOTES - DYNAMIC ARRAY + + * If you know how long a dynamic array is going to be in advance, you can avoid + extra memory allocations by using arrsetlen to allocate it to that length in + advance and use foo[n] while filling it out, or arrsetcap to allocate the memory + for that length and use arrput/arrpush as normal. + + * Unlike some other versions of the dynamic array, this version should + be safe to use with strict-aliasing optimizations. + +NOTES - HASH MAP + + * For compilers other than GCC and clang (e.g. Visual Studio), for hmput/hmget/hmdel + and variants, the key must be an lvalue (so the macro can take the address of it). + Extensions are used that eliminate this requirement if you're using C99 and later + in GCC or clang, or if you're using C++ in GCC. But note that this can make your + code less portable. + + * To test for presence of a key in a hashmap, just do 'hmgeti(foo,key) >= 0'. + + * The iteration order of your data in the hashmap is determined solely by the + order of insertions and deletions. In particular, if you never delete, new + keys are always added at the end of the array. This will be consistent + across all platforms and versions of the library. However, you should not + attempt to serialize the internal hash table, as the hash is not consistent + between different platforms, and may change with future versions of the library. + + * Use sh_new_arena() for string hashmaps that you never delete from. Initialize + with NULL if you're managing the memory for your strings, or your strings are + never freed (at least until the hashmap is freed). Otherwise, use sh_new_strdup(). + @TODO: make an arena variant that garbage collects the strings with a trivial + copy collector into a new arena whenever the table shrinks / rebuilds. Since + current arena recommendation is to only use arena if it never deletes, then + this can just replace current arena implementation. + + * If adversarial input is a serious concern and you're on a 64-bit platform, + enable STBDS_SIPHASH_2_4 (see the 'Compile-time options' section), and pass + a strong random number to stbds_rand_seed. + + * The default value for the hash table is stored in foo[-1], so if you + use code like 'hmget(T,k)->value = 5' you can accidentally overwrite + the value stored by hmdefault if 'k' is not present. + +CREDITS + + Sean Barrett -- library, idea for dynamic array API/implementation + Per Vognsen -- idea for hash table API/implementation + Rafael Sachetto -- arrpop() + github:HeroicKatora -- arraddn() reworking + + Bugfixes: + Andy Durdin + Shane Liesegang + Vinh Truong + Andreas Molzer + github:hashitaku + github:srdjanstipic + Macoy Madson + Andreas Vennstrom + Tobias Mansfield-Williams +*/ + +#ifdef STBDS_UNIT_TESTS +#define _CRT_SECURE_NO_WARNINGS +#endif + +#ifndef INCLUDE_STB_DS_H +#define INCLUDE_STB_DS_H + +#include +#include + +#ifndef STBDS_NO_SHORT_NAMES +#define arrlen stbds_arrlen +#define arrlenu stbds_arrlenu +#define arrput stbds_arrput +#define arrpush stbds_arrput +#define arrpop stbds_arrpop +#define arrfree stbds_arrfree +#define arraddn stbds_arraddn // deprecated, use one of the following instead: +#define arraddnptr stbds_arraddnptr +#define arraddnindex stbds_arraddnindex +#define arrsetlen stbds_arrsetlen +#define arrlast stbds_arrlast +#define arrins stbds_arrins +#define arrinsn stbds_arrinsn +#define arrdel stbds_arrdel +#define arrdeln stbds_arrdeln +#define arrdelswap stbds_arrdelswap +#define arrcap stbds_arrcap +#define arrsetcap stbds_arrsetcap + +#define hmput stbds_hmput +#define hmputs stbds_hmputs +#define hmget stbds_hmget +#define hmget_ts stbds_hmget_ts +#define hmgets stbds_hmgets +#define hmgetp stbds_hmgetp +#define hmgetp_ts stbds_hmgetp_ts +#define hmgetp_null stbds_hmgetp_null +#define hmgeti stbds_hmgeti +#define hmgeti_ts stbds_hmgeti_ts +#define hmdel stbds_hmdel +#define hmlen stbds_hmlen +#define hmlenu stbds_hmlenu +#define hmfree stbds_hmfree +#define hmdefault stbds_hmdefault +#define hmdefaults stbds_hmdefaults + +#define shput stbds_shput +#define shputi stbds_shputi +#define shputs stbds_shputs +#define shget stbds_shget +#define shgeti stbds_shgeti +#define shgets stbds_shgets +#define shgetp stbds_shgetp +#define shgetp_null stbds_shgetp_null +#define shdel stbds_shdel +#define shlen stbds_shlen +#define shlenu stbds_shlenu +#define shfree stbds_shfree +#define shdefault stbds_shdefault +#define shdefaults stbds_shdefaults +#define sh_new_arena stbds_sh_new_arena +#define sh_new_strdup stbds_sh_new_strdup + +#define stralloc stbds_stralloc +#define strreset stbds_strreset +#endif + +#if defined(STBDS_REALLOC) && !defined(STBDS_FREE) || !defined(STBDS_REALLOC) && defined(STBDS_FREE) +#error "You must define both STBDS_REALLOC and STBDS_FREE, or neither." +#endif +#if !defined(STBDS_REALLOC) && !defined(STBDS_FREE) +#include +#define STBDS_REALLOC(c,p,s) realloc(p,s) +#define STBDS_FREE(c,p) free(p) +#endif + +#ifdef _MSC_VER +#define STBDS_NOTUSED(v) (void)(v) +#else +#define STBDS_NOTUSED(v) (void)sizeof(v) +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +// for security against attackers, seed the library with a random number, at least time() but stronger is better +extern void stbds_rand_seed(size_t seed); + +// these are the hash functions used internally if you want to test them or use them for other purposes +extern size_t stbds_hash_bytes(void *p, size_t len, size_t seed); +extern size_t stbds_hash_string(char *str, size_t seed); + +// this is a simple string arena allocator, initialize with e.g. 'stbds_string_arena my_arena={0}'. +typedef struct stbds_string_arena stbds_string_arena; +extern char * stbds_stralloc(stbds_string_arena *a, char *str); +extern void stbds_strreset(stbds_string_arena *a); + +// have to #define STBDS_UNIT_TESTS to call this +extern void stbds_unit_tests(void); + +/////////////// +// +// Everything below here is implementation details +// + +extern void * stbds_arrgrowf(void *a, size_t elemsize, size_t addlen, size_t min_cap); +extern void stbds_arrfreef(void *a); +extern void stbds_hmfree_func(void *p, size_t elemsize); +extern void * stbds_hmget_key(void *a, size_t elemsize, void *key, size_t keysize, int mode); +extern void * stbds_hmget_key_ts(void *a, size_t elemsize, void *key, size_t keysize, ptrdiff_t *temp, int mode); +extern void * stbds_hmput_default(void *a, size_t elemsize); +extern void * stbds_hmput_key(void *a, size_t elemsize, void *key, size_t keysize, int mode); +extern void * stbds_hmdel_key(void *a, size_t elemsize, void *key, size_t keysize, size_t keyoffset, int mode); +extern void * stbds_shmode_func(size_t elemsize, int mode); + +#ifdef __cplusplus +} +#endif + +#if defined(__GNUC__) || defined(__clang__) +#define STBDS_HAS_TYPEOF +#ifdef __cplusplus +//#define STBDS_HAS_LITERAL_ARRAY // this is currently broken for clang +#endif +#endif + +#if !defined(__cplusplus) +#if defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L +#define STBDS_HAS_LITERAL_ARRAY +#endif +#endif + +// this macro takes the address of the argument, but on gcc/clang can accept rvalues +#if defined(STBDS_HAS_LITERAL_ARRAY) && defined(STBDS_HAS_TYPEOF) + #if __clang__ + #define STBDS_ADDRESSOF(typevar, value) ((__typeof__(typevar)[1]){value}) // literal array decays to pointer to value + #else + #define STBDS_ADDRESSOF(typevar, value) ((typeof(typevar)[1]){value}) // literal array decays to pointer to value + #endif +#else +#define STBDS_ADDRESSOF(typevar, value) &(value) +#endif + +#define STBDS_OFFSETOF(var,field) ((char *) &(var)->field - (char *) (var)) + +#define stbds_header(t) ((stbds_array_header *) (t) - 1) +#define stbds_temp(t) stbds_header(t)->temp +#define stbds_temp_key(t) (*(char **) stbds_header(t)->hash_table) + +#define stbds_arrsetcap(a,n) (stbds_arrgrow(a,0,n)) +#define stbds_arrsetlen(a,n) ((stbds_arrcap(a) < (size_t) (n) ? stbds_arrsetcap((a),(size_t)(n)),0 : 0), (a) ? stbds_header(a)->length = (size_t) (n) : 0) +#define stbds_arrcap(a) ((a) ? stbds_header(a)->capacity : 0) +#define stbds_arrlen(a) ((a) ? (ptrdiff_t) stbds_header(a)->length : 0) +#define stbds_arrlenu(a) ((a) ? stbds_header(a)->length : 0) +#define stbds_arrput(a,v) (stbds_arrmaybegrow(a,1), (a)[stbds_header(a)->length++] = (v)) +#define stbds_arrpush stbds_arrput // synonym +#define stbds_arrpop(a) (stbds_header(a)->length--, (a)[stbds_header(a)->length]) +#define stbds_arraddn(a,n) ((void)(stbds_arraddnindex(a, n))) // deprecated, use one of the following instead: +#define stbds_arraddnptr(a,n) (stbds_arrmaybegrow(a,n), (n) ? (stbds_header(a)->length += (n), &(a)[stbds_header(a)->length-(n)]) : (a)) +#define stbds_arraddnindex(a,n)(stbds_arrmaybegrow(a,n), (n) ? (stbds_header(a)->length += (n), stbds_header(a)->length-(n)) : stbds_arrlen(a)) +#define stbds_arraddnoff stbds_arraddnindex +#define stbds_arrlast(a) ((a)[stbds_header(a)->length-1]) +#define stbds_arrfree(a) ((void) ((a) ? STBDS_FREE(NULL,stbds_header(a)) : (void)0), (a)=NULL) +#define stbds_arrdel(a,i) stbds_arrdeln(a,i,1) +#define stbds_arrdeln(a,i,n) (memmove(&(a)[i], &(a)[(i)+(n)], sizeof *(a) * (stbds_header(a)->length-(n)-(i))), stbds_header(a)->length -= (n)) +#define stbds_arrdelswap(a,i) ((a)[i] = stbds_arrlast(a), stbds_header(a)->length -= 1) +#define stbds_arrinsn(a,i,n) (stbds_arraddn((a),(n)), memmove(&(a)[(i)+(n)], &(a)[i], sizeof *(a) * (stbds_header(a)->length-(n)-(i)))) +#define stbds_arrins(a,i,v) (stbds_arrinsn((a),(i),1), (a)[i]=(v)) + +#define stbds_arrmaybegrow(a,n) ((!(a) || stbds_header(a)->length + (n) > stbds_header(a)->capacity) \ + ? (stbds_arrgrow(a,n,0),0) : 0) + +#define stbds_arrgrow(a,b,c) ((a) = stbds_arrgrowf_wrapper((a), sizeof *(a), (b), (c))) + +#define stbds_hmput(t, k, v) \ + ((t) = stbds_hmput_key_wrapper((t), sizeof *(t), (void*) STBDS_ADDRESSOF((t)->key, (k)), sizeof (t)->key, 0), \ + (t)[stbds_temp((t)-1)].key = (k), \ + (t)[stbds_temp((t)-1)].value = (v)) + +#define stbds_hmputs(t, s) \ + ((t) = stbds_hmput_key_wrapper((t), sizeof *(t), &(s).key, sizeof (s).key, STBDS_HM_BINARY), \ + (t)[stbds_temp((t)-1)] = (s)) + +#define stbds_hmgeti(t,k) \ + ((t) = stbds_hmget_key_wrapper((t), sizeof *(t), (void*) STBDS_ADDRESSOF((t)->key, (k)), sizeof (t)->key, STBDS_HM_BINARY), \ + stbds_temp((t)-1)) + +#define stbds_hmgeti_ts(t,k,temp) \ + ((t) = stbds_hmget_key_ts_wrapper((t), sizeof *(t), (void*) STBDS_ADDRESSOF((t)->key, (k)), sizeof (t)->key, &(temp), STBDS_HM_BINARY), \ + (temp)) + +#define stbds_hmgetp(t, k) \ + ((void) stbds_hmgeti(t,k), &(t)[stbds_temp((t)-1)]) + +#define stbds_hmgetp_ts(t, k, temp) \ + ((void) stbds_hmgeti_ts(t,k,temp), &(t)[temp]) + +#define stbds_hmdel(t,k) \ + (((t) = stbds_hmdel_key_wrapper((t),sizeof *(t), (void*) STBDS_ADDRESSOF((t)->key, (k)), sizeof (t)->key, STBDS_OFFSETOF((t),key), STBDS_HM_BINARY)),(t)?stbds_temp((t)-1):0) + +#define stbds_hmdefault(t, v) \ + ((t) = stbds_hmput_default_wrapper((t), sizeof *(t)), (t)[-1].value = (v)) + +#define stbds_hmdefaults(t, s) \ + ((t) = stbds_hmput_default_wrapper((t), sizeof *(t)), (t)[-1] = (s)) + +#define stbds_hmfree(p) \ + ((void) ((p) != NULL ? stbds_hmfree_func((p)-1,sizeof*(p)),0 : 0),(p)=NULL) + +#define stbds_hmgets(t, k) (*stbds_hmgetp(t,k)) +#define stbds_hmget(t, k) (stbds_hmgetp(t,k)->value) +#define stbds_hmget_ts(t, k, temp) (stbds_hmgetp_ts(t,k,temp)->value) +#define stbds_hmlen(t) ((t) ? (ptrdiff_t) stbds_header((t)-1)->length-1 : 0) +#define stbds_hmlenu(t) ((t) ? stbds_header((t)-1)->length-1 : 0) +#define stbds_hmgetp_null(t,k) (stbds_hmgeti(t,k) == -1 ? NULL : &(t)[stbds_temp((t)-1)]) + +#define stbds_shput(t, k, v) \ + ((t) = stbds_hmput_key_wrapper((t), sizeof *(t), (void*) (k), sizeof (t)->key, STBDS_HM_STRING), \ + (t)[stbds_temp((t)-1)].value = (v)) + +#define stbds_shputi(t, k, v) \ + ((t) = stbds_hmput_key_wrapper((t), sizeof *(t), (void*) (k), sizeof (t)->key, STBDS_HM_STRING), \ + (t)[stbds_temp((t)-1)].value = (v), stbds_temp((t)-1)) + +#define stbds_shputs(t, s) \ + ((t) = stbds_hmput_key_wrapper((t), sizeof *(t), (void*) (s).key, sizeof (s).key, STBDS_HM_STRING), \ + (t)[stbds_temp((t)-1)] = (s), \ + (t)[stbds_temp((t)-1)].key = stbds_temp_key((t)-1)) // above line overwrites whole structure, so must rewrite key here if it was allocated internally + +#define stbds_pshput(t, p) \ + ((t) = stbds_hmput_key_wrapper((t), sizeof *(t), (void*) (p)->key, sizeof (p)->key, STBDS_HM_PTR_TO_STRING), \ + (t)[stbds_temp((t)-1)] = (p)) + +#define stbds_shgeti(t,k) \ + ((t) = stbds_hmget_key_wrapper((t), sizeof *(t), (void*) (k), sizeof (t)->key, STBDS_HM_STRING), \ + stbds_temp((t)-1)) + +#define stbds_pshgeti(t,k) \ + ((t) = stbds_hmget_key_wrapper((t), sizeof *(t), (void*) (k), sizeof (*(t))->key, STBDS_HM_PTR_TO_STRING), \ + stbds_temp((t)-1)) + +#define stbds_shgetp(t, k) \ + ((void) stbds_shgeti(t,k), &(t)[stbds_temp((t)-1)]) + +#define stbds_pshget(t, k) \ + ((void) stbds_pshgeti(t,k), (t)[stbds_temp((t)-1)]) + +#define stbds_shdel(t,k) \ + (((t) = stbds_hmdel_key_wrapper((t),sizeof *(t), (void*) (k), sizeof (t)->key, STBDS_OFFSETOF((t),key), STBDS_HM_STRING)),(t)?stbds_temp((t)-1):0) +#define stbds_pshdel(t,k) \ + (((t) = stbds_hmdel_key_wrapper((t),sizeof *(t), (void*) (k), sizeof (*(t))->key, STBDS_OFFSETOF(*(t),key), STBDS_HM_PTR_TO_STRING)),(t)?stbds_temp((t)-1):0) + +#define stbds_sh_new_arena(t) \ + ((t) = stbds_shmode_func_wrapper(t, sizeof *(t), STBDS_SH_ARENA)) +#define stbds_sh_new_strdup(t) \ + ((t) = stbds_shmode_func_wrapper(t, sizeof *(t), STBDS_SH_STRDUP)) + +#define stbds_shdefault(t, v) stbds_hmdefault(t,v) +#define stbds_shdefaults(t, s) stbds_hmdefaults(t,s) + +#define stbds_shfree stbds_hmfree +#define stbds_shlenu stbds_hmlenu + +#define stbds_shgets(t, k) (*stbds_shgetp(t,k)) +#define stbds_shget(t, k) (stbds_shgetp(t,k)->value) +#define stbds_shgetp_null(t,k) (stbds_shgeti(t,k) == -1 ? NULL : &(t)[stbds_temp((t)-1)]) +#define stbds_shlen stbds_hmlen + +typedef struct +{ + size_t length; + size_t capacity; + void * hash_table; + ptrdiff_t temp; +} stbds_array_header; + +typedef struct stbds_string_block +{ + struct stbds_string_block *next; + char storage[8]; +} stbds_string_block; + +struct stbds_string_arena +{ + stbds_string_block *storage; + size_t remaining; + unsigned char block; + unsigned char mode; // this isn't used by the string arena itself +}; + +#define STBDS_HM_BINARY 0 +#define STBDS_HM_STRING 1 + +enum +{ + STBDS_SH_NONE, + STBDS_SH_DEFAULT, + STBDS_SH_STRDUP, + STBDS_SH_ARENA +}; + +#ifdef __cplusplus +// in C we use implicit assignment from these void*-returning functions to T*. +// in C++ these templates make the same code work +template static T * stbds_arrgrowf_wrapper(T *a, size_t elemsize, size_t addlen, size_t min_cap) { + return (T*)stbds_arrgrowf((void *)a, elemsize, addlen, min_cap); +} +template static T * stbds_hmget_key_wrapper(T *a, size_t elemsize, void *key, size_t keysize, int mode) { + return (T*)stbds_hmget_key((void*)a, elemsize, key, keysize, mode); +} +template static T * stbds_hmget_key_ts_wrapper(T *a, size_t elemsize, void *key, size_t keysize, ptrdiff_t *temp, int mode) { + return (T*)stbds_hmget_key_ts((void*)a, elemsize, key, keysize, temp, mode); +} +template static T * stbds_hmput_default_wrapper(T *a, size_t elemsize) { + return (T*)stbds_hmput_default((void *)a, elemsize); +} +template static T * stbds_hmput_key_wrapper(T *a, size_t elemsize, void *key, size_t keysize, int mode) { + return (T*)stbds_hmput_key((void*)a, elemsize, key, keysize, mode); +} +template static T * stbds_hmdel_key_wrapper(T *a, size_t elemsize, void *key, size_t keysize, size_t keyoffset, int mode){ + return (T*)stbds_hmdel_key((void*)a, elemsize, key, keysize, keyoffset, mode); +} +template static T * stbds_shmode_func_wrapper(T *, size_t elemsize, int mode) { + return (T*)stbds_shmode_func(elemsize, mode); +} +#else +#define stbds_arrgrowf_wrapper stbds_arrgrowf +#define stbds_hmget_key_wrapper stbds_hmget_key +#define stbds_hmget_key_ts_wrapper stbds_hmget_key_ts +#define stbds_hmput_default_wrapper stbds_hmput_default +#define stbds_hmput_key_wrapper stbds_hmput_key +#define stbds_hmdel_key_wrapper stbds_hmdel_key +#define stbds_shmode_func_wrapper(t,e,m) stbds_shmode_func(e,m) +#endif + +#endif // INCLUDE_STB_DS_H + + +////////////////////////////////////////////////////////////////////////////// +// +// IMPLEMENTATION +// + +#ifdef STB_DS_IMPLEMENTATION +#include +#include + +#ifndef STBDS_ASSERT +#define STBDS_ASSERT_WAS_UNDEFINED +#define STBDS_ASSERT(x) ((void) 0) +#endif + +#ifdef STBDS_STATISTICS +#define STBDS_STATS(x) x +size_t stbds_array_grow; +size_t stbds_hash_grow; +size_t stbds_hash_shrink; +size_t stbds_hash_rebuild; +size_t stbds_hash_probes; +size_t stbds_hash_alloc; +size_t stbds_rehash_probes; +size_t stbds_rehash_items; +#else +#define STBDS_STATS(x) +#endif + +// +// stbds_arr implementation +// + +//int *prev_allocs[65536]; +//int num_prev; + +void *stbds_arrgrowf(void *a, size_t elemsize, size_t addlen, size_t min_cap) +{ + stbds_array_header temp={0}; // force debugging + void *b; + size_t min_len = stbds_arrlen(a) + addlen; + (void) sizeof(temp); + + // compute the minimum capacity needed + if (min_len > min_cap) + min_cap = min_len; + + if (min_cap <= stbds_arrcap(a)) + return a; + + // increase needed capacity to guarantee O(1) amortized + if (min_cap < 2 * stbds_arrcap(a)) + min_cap = 2 * stbds_arrcap(a); + else if (min_cap < 4) + min_cap = 4; + + //if (num_prev < 65536) if (a) prev_allocs[num_prev++] = (int *) ((char *) a+1); + //if (num_prev == 2201) + // num_prev = num_prev; + b = STBDS_REALLOC(NULL, (a) ? stbds_header(a) : 0, elemsize * min_cap + sizeof(stbds_array_header)); + //if (num_prev < 65536) prev_allocs[num_prev++] = (int *) (char *) b; + b = (char *) b + sizeof(stbds_array_header); + if (a == NULL) { + stbds_header(b)->length = 0; + stbds_header(b)->hash_table = 0; + stbds_header(b)->temp = 0; + } else { + STBDS_STATS(++stbds_array_grow); + } + stbds_header(b)->capacity = min_cap; + + return b; +} + +void stbds_arrfreef(void *a) +{ + STBDS_FREE(NULL, stbds_header(a)); +} + +// +// stbds_hm hash table implementation +// + +#ifdef STBDS_INTERNAL_SMALL_BUCKET +#define STBDS_BUCKET_LENGTH 4 +#else +#define STBDS_BUCKET_LENGTH 8 +#endif + +#define STBDS_BUCKET_SHIFT (STBDS_BUCKET_LENGTH == 8 ? 3 : 2) +#define STBDS_BUCKET_MASK (STBDS_BUCKET_LENGTH-1) +#define STBDS_CACHE_LINE_SIZE 64 + +#define STBDS_ALIGN_FWD(n,a) (((n) + (a) - 1) & ~((a)-1)) + +typedef struct +{ + size_t hash [STBDS_BUCKET_LENGTH]; + ptrdiff_t index[STBDS_BUCKET_LENGTH]; +} stbds_hash_bucket; // in 32-bit, this is one 64-byte cache line; in 64-bit, each array is one 64-byte cache line + +typedef struct +{ + char * temp_key; // this MUST be the first field of the hash table + size_t slot_count; + size_t used_count; + size_t used_count_threshold; + size_t used_count_shrink_threshold; + size_t tombstone_count; + size_t tombstone_count_threshold; + size_t seed; + size_t slot_count_log2; + stbds_string_arena string; + stbds_hash_bucket *storage; // not a separate allocation, just 64-byte aligned storage after this struct +} stbds_hash_index; + +#define STBDS_INDEX_EMPTY -1 +#define STBDS_INDEX_DELETED -2 +#define STBDS_INDEX_IN_USE(x) ((x) >= 0) + +#define STBDS_HASH_EMPTY 0 +#define STBDS_HASH_DELETED 1 + +static size_t stbds_hash_seed=0x31415926; + +void stbds_rand_seed(size_t seed) +{ + stbds_hash_seed = seed; +} + +#define stbds_load_32_or_64(var, temp, v32, v64_hi, v64_lo) \ + temp = v64_lo ^ v32, temp <<= 16, temp <<= 16, temp >>= 16, temp >>= 16, /* discard if 32-bit */ \ + var = v64_hi, var <<= 16, var <<= 16, /* discard if 32-bit */ \ + var ^= temp ^ v32 + +#define STBDS_SIZE_T_BITS ((sizeof (size_t)) * 8) + +static size_t stbds_probe_position(size_t hash, size_t slot_count, size_t slot_log2) +{ + size_t pos; + STBDS_NOTUSED(slot_log2); + pos = hash & (slot_count-1); + #ifdef STBDS_INTERNAL_BUCKET_START + pos &= ~STBDS_BUCKET_MASK; + #endif + return pos; +} + +static size_t stbds_log2(size_t slot_count) +{ + size_t n=0; + while (slot_count > 1) { + slot_count >>= 1; + ++n; + } + return n; +} + +static stbds_hash_index *stbds_make_hash_index(size_t slot_count, stbds_hash_index *ot) +{ + stbds_hash_index *t; + t = (stbds_hash_index *) STBDS_REALLOC(NULL,0,(slot_count >> STBDS_BUCKET_SHIFT) * sizeof(stbds_hash_bucket) + sizeof(stbds_hash_index) + STBDS_CACHE_LINE_SIZE-1); + t->storage = (stbds_hash_bucket *) STBDS_ALIGN_FWD((size_t) (t+1), STBDS_CACHE_LINE_SIZE); + t->slot_count = slot_count; + t->slot_count_log2 = stbds_log2(slot_count); + t->tombstone_count = 0; + t->used_count = 0; + + #if 0 // A1 + t->used_count_threshold = slot_count*12/16; // if 12/16th of table is occupied, grow + t->tombstone_count_threshold = slot_count* 2/16; // if tombstones are 2/16th of table, rebuild + t->used_count_shrink_threshold = slot_count* 4/16; // if table is only 4/16th full, shrink + #elif 1 // A2 + //t->used_count_threshold = slot_count*12/16; // if 12/16th of table is occupied, grow + //t->tombstone_count_threshold = slot_count* 3/16; // if tombstones are 3/16th of table, rebuild + //t->used_count_shrink_threshold = slot_count* 4/16; // if table is only 4/16th full, shrink + + // compute without overflowing + t->used_count_threshold = slot_count - (slot_count>>2); + t->tombstone_count_threshold = (slot_count>>3) + (slot_count>>4); + t->used_count_shrink_threshold = slot_count >> 2; + + #elif 0 // B1 + t->used_count_threshold = slot_count*13/16; // if 13/16th of table is occupied, grow + t->tombstone_count_threshold = slot_count* 2/16; // if tombstones are 2/16th of table, rebuild + t->used_count_shrink_threshold = slot_count* 5/16; // if table is only 5/16th full, shrink + #else // C1 + t->used_count_threshold = slot_count*14/16; // if 14/16th of table is occupied, grow + t->tombstone_count_threshold = slot_count* 2/16; // if tombstones are 2/16th of table, rebuild + t->used_count_shrink_threshold = slot_count* 6/16; // if table is only 6/16th full, shrink + #endif + // Following statistics were measured on a Core i7-6700 @ 4.00Ghz, compiled with clang 7.0.1 -O2 + // Note that the larger tables have high variance as they were run fewer times + // A1 A2 B1 C1 + // 0.10ms : 0.10ms : 0.10ms : 0.11ms : 2,000 inserts creating 2K table + // 0.96ms : 0.95ms : 0.97ms : 1.04ms : 20,000 inserts creating 20K table + // 14.48ms : 14.46ms : 10.63ms : 11.00ms : 200,000 inserts creating 200K table + // 195.74ms : 196.35ms : 203.69ms : 214.92ms : 2,000,000 inserts creating 2M table + // 2193.88ms : 2209.22ms : 2285.54ms : 2437.17ms : 20,000,000 inserts creating 20M table + // 65.27ms : 53.77ms : 65.33ms : 65.47ms : 500,000 inserts & deletes in 2K table + // 72.78ms : 62.45ms : 71.95ms : 72.85ms : 500,000 inserts & deletes in 20K table + // 89.47ms : 77.72ms : 96.49ms : 96.75ms : 500,000 inserts & deletes in 200K table + // 97.58ms : 98.14ms : 97.18ms : 97.53ms : 500,000 inserts & deletes in 2M table + // 118.61ms : 119.62ms : 120.16ms : 118.86ms : 500,000 inserts & deletes in 20M table + // 192.11ms : 194.39ms : 196.38ms : 195.73ms : 500,000 inserts & deletes in 200M table + + if (slot_count <= STBDS_BUCKET_LENGTH) + t->used_count_shrink_threshold = 0; + // to avoid infinite loop, we need to guarantee that at least one slot is empty and will terminate probes + STBDS_ASSERT(t->used_count_threshold + t->tombstone_count_threshold < t->slot_count); + STBDS_STATS(++stbds_hash_alloc); + if (ot) { + t->string = ot->string; + // reuse old seed so we can reuse old hashes so below "copy out old data" doesn't do any hashing + t->seed = ot->seed; + } else { + size_t a,b,temp; + memset(&t->string, 0, sizeof(t->string)); + t->seed = stbds_hash_seed; + // LCG + // in 32-bit, a = 2147001325 b = 715136305 + // in 64-bit, a = 2862933555777941757 b = 3037000493 + stbds_load_32_or_64(a,temp, 2147001325, 0x27bb2ee6, 0x87b0b0fd); + stbds_load_32_or_64(b,temp, 715136305, 0, 0xb504f32d); + stbds_hash_seed = stbds_hash_seed * a + b; + } + + { + size_t i,j; + for (i=0; i < slot_count >> STBDS_BUCKET_SHIFT; ++i) { + stbds_hash_bucket *b = &t->storage[i]; + for (j=0; j < STBDS_BUCKET_LENGTH; ++j) + b->hash[j] = STBDS_HASH_EMPTY; + for (j=0; j < STBDS_BUCKET_LENGTH; ++j) + b->index[j] = STBDS_INDEX_EMPTY; + } + } + + // copy out the old data, if any + if (ot) { + size_t i,j; + t->used_count = ot->used_count; + for (i=0; i < ot->slot_count >> STBDS_BUCKET_SHIFT; ++i) { + stbds_hash_bucket *ob = &ot->storage[i]; + for (j=0; j < STBDS_BUCKET_LENGTH; ++j) { + if (STBDS_INDEX_IN_USE(ob->index[j])) { + size_t hash = ob->hash[j]; + size_t pos = stbds_probe_position(hash, t->slot_count, t->slot_count_log2); + size_t step = STBDS_BUCKET_LENGTH; + STBDS_STATS(++stbds_rehash_items); + for (;;) { + size_t limit,z; + stbds_hash_bucket *bucket; + bucket = &t->storage[pos >> STBDS_BUCKET_SHIFT]; + STBDS_STATS(++stbds_rehash_probes); + + for (z=pos & STBDS_BUCKET_MASK; z < STBDS_BUCKET_LENGTH; ++z) { + if (bucket->hash[z] == 0) { + bucket->hash[z] = hash; + bucket->index[z] = ob->index[j]; + goto done; + } + } + + limit = pos & STBDS_BUCKET_MASK; + for (z = 0; z < limit; ++z) { + if (bucket->hash[z] == 0) { + bucket->hash[z] = hash; + bucket->index[z] = ob->index[j]; + goto done; + } + } + + pos += step; // quadratic probing + step += STBDS_BUCKET_LENGTH; + pos &= (t->slot_count-1); + } + } + done: + ; + } + } + } + + return t; +} + +#define STBDS_ROTATE_LEFT(val, n) (((val) << (n)) | ((val) >> (STBDS_SIZE_T_BITS - (n)))) +#define STBDS_ROTATE_RIGHT(val, n) (((val) >> (n)) | ((val) << (STBDS_SIZE_T_BITS - (n)))) + +size_t stbds_hash_string(char *str, size_t seed) +{ + size_t hash = seed; + while (*str) + hash = STBDS_ROTATE_LEFT(hash, 9) + (unsigned char) *str++; + + // Thomas Wang 64-to-32 bit mix function, hopefully also works in 32 bits + hash ^= seed; + hash = (~hash) + (hash << 18); + hash ^= hash ^ STBDS_ROTATE_RIGHT(hash,31); + hash = hash * 21; + hash ^= hash ^ STBDS_ROTATE_RIGHT(hash,11); + hash += (hash << 6); + hash ^= STBDS_ROTATE_RIGHT(hash,22); + return hash+seed; +} + +#ifdef STBDS_SIPHASH_2_4 +#define STBDS_SIPHASH_C_ROUNDS 2 +#define STBDS_SIPHASH_D_ROUNDS 4 +typedef int STBDS_SIPHASH_2_4_can_only_be_used_in_64_bit_builds[sizeof(size_t) == 8 ? 1 : -1]; +#endif + +#ifndef STBDS_SIPHASH_C_ROUNDS +#define STBDS_SIPHASH_C_ROUNDS 1 +#endif +#ifndef STBDS_SIPHASH_D_ROUNDS +#define STBDS_SIPHASH_D_ROUNDS 1 +#endif + +#ifdef _MSC_VER +#pragma warning(push) +#pragma warning(disable:4127) // conditional expression is constant, for do..while(0) and sizeof()== +#endif + +static size_t stbds_siphash_bytes(void *p, size_t len, size_t seed) +{ + unsigned char *d = (unsigned char *) p; + size_t i,j; + size_t v0,v1,v2,v3, data; + + // hash that works on 32- or 64-bit registers without knowing which we have + // (computes different results on 32-bit and 64-bit platform) + // derived from siphash, but on 32-bit platforms very different as it uses 4 32-bit state not 4 64-bit + v0 = ((((size_t) 0x736f6d65 << 16) << 16) + 0x70736575) ^ seed; + v1 = ((((size_t) 0x646f7261 << 16) << 16) + 0x6e646f6d) ^ ~seed; + v2 = ((((size_t) 0x6c796765 << 16) << 16) + 0x6e657261) ^ seed; + v3 = ((((size_t) 0x74656462 << 16) << 16) + 0x79746573) ^ ~seed; + + #ifdef STBDS_TEST_SIPHASH_2_4 + // hardcoded with key material in the siphash test vectors + v0 ^= 0x0706050403020100ull ^ seed; + v1 ^= 0x0f0e0d0c0b0a0908ull ^ ~seed; + v2 ^= 0x0706050403020100ull ^ seed; + v3 ^= 0x0f0e0d0c0b0a0908ull ^ ~seed; + #endif + + #define STBDS_SIPROUND() \ + do { \ + v0 += v1; v1 = STBDS_ROTATE_LEFT(v1, 13); v1 ^= v0; v0 = STBDS_ROTATE_LEFT(v0,STBDS_SIZE_T_BITS/2); \ + v2 += v3; v3 = STBDS_ROTATE_LEFT(v3, 16); v3 ^= v2; \ + v2 += v1; v1 = STBDS_ROTATE_LEFT(v1, 17); v1 ^= v2; v2 = STBDS_ROTATE_LEFT(v2,STBDS_SIZE_T_BITS/2); \ + v0 += v3; v3 = STBDS_ROTATE_LEFT(v3, 21); v3 ^= v0; \ + } while (0) + + for (i=0; i+sizeof(size_t) <= len; i += sizeof(size_t), d += sizeof(size_t)) { + data = d[0] | (d[1] << 8) | (d[2] << 16) | (d[3] << 24); + data |= (size_t) (d[4] | (d[5] << 8) | (d[6] << 16) | (d[7] << 24)) << 16 << 16; // discarded if size_t == 4 + + v3 ^= data; + for (j=0; j < STBDS_SIPHASH_C_ROUNDS; ++j) + STBDS_SIPROUND(); + v0 ^= data; + } + data = len << (STBDS_SIZE_T_BITS-8); + switch (len - i) { + case 7: data |= ((size_t) d[6] << 24) << 24; // fall through + case 6: data |= ((size_t) d[5] << 20) << 20; // fall through + case 5: data |= ((size_t) d[4] << 16) << 16; // fall through + case 4: data |= (d[3] << 24); // fall through + case 3: data |= (d[2] << 16); // fall through + case 2: data |= (d[1] << 8); // fall through + case 1: data |= d[0]; // fall through + case 0: break; + } + v3 ^= data; + for (j=0; j < STBDS_SIPHASH_C_ROUNDS; ++j) + STBDS_SIPROUND(); + v0 ^= data; + v2 ^= 0xff; + for (j=0; j < STBDS_SIPHASH_D_ROUNDS; ++j) + STBDS_SIPROUND(); + +#ifdef STBDS_SIPHASH_2_4 + return v0^v1^v2^v3; +#else + return v1^v2^v3; // slightly stronger since v0^v3 in above cancels out final round operation? I tweeted at the authors of SipHash about this but they didn't reply +#endif +} + +size_t stbds_hash_bytes(void *p, size_t len, size_t seed) +{ +#ifdef STBDS_SIPHASH_2_4 + return stbds_siphash_bytes(p,len,seed); +#else + unsigned char *d = (unsigned char *) p; + + if (len == 4) { + unsigned int hash = d[0] | (d[1] << 8) | (d[2] << 16) | (d[3] << 24); + #if 0 + // HASH32-A Bob Jenkin's hash function w/o large constants + hash ^= seed; + hash -= (hash<<6); + hash ^= (hash>>17); + hash -= (hash<<9); + hash ^= seed; + hash ^= (hash<<4); + hash -= (hash<<3); + hash ^= (hash<<10); + hash ^= (hash>>15); + #elif 1 + // HASH32-BB Bob Jenkin's presumably-accidental version of Thomas Wang hash with rotates turned into shifts. + // Note that converting these back to rotates makes it run a lot slower, presumably due to collisions, so I'm + // not really sure what's going on. + hash ^= seed; + hash = (hash ^ 61) ^ (hash >> 16); + hash = hash + (hash << 3); + hash = hash ^ (hash >> 4); + hash = hash * 0x27d4eb2d; + hash ^= seed; + hash = hash ^ (hash >> 15); + #else // HASH32-C - Murmur3 + hash ^= seed; + hash *= 0xcc9e2d51; + hash = (hash << 17) | (hash >> 15); + hash *= 0x1b873593; + hash ^= seed; + hash = (hash << 19) | (hash >> 13); + hash = hash*5 + 0xe6546b64; + hash ^= hash >> 16; + hash *= 0x85ebca6b; + hash ^= seed; + hash ^= hash >> 13; + hash *= 0xc2b2ae35; + hash ^= hash >> 16; + #endif + // Following statistics were measured on a Core i7-6700 @ 4.00Ghz, compiled with clang 7.0.1 -O2 + // Note that the larger tables have high variance as they were run fewer times + // HASH32-A // HASH32-BB // HASH32-C + // 0.10ms // 0.10ms // 0.10ms : 2,000 inserts creating 2K table + // 0.96ms // 0.95ms // 0.99ms : 20,000 inserts creating 20K table + // 14.69ms // 14.43ms // 14.97ms : 200,000 inserts creating 200K table + // 199.99ms // 195.36ms // 202.05ms : 2,000,000 inserts creating 2M table + // 2234.84ms // 2187.74ms // 2240.38ms : 20,000,000 inserts creating 20M table + // 55.68ms // 53.72ms // 57.31ms : 500,000 inserts & deletes in 2K table + // 63.43ms // 61.99ms // 65.73ms : 500,000 inserts & deletes in 20K table + // 80.04ms // 77.96ms // 81.83ms : 500,000 inserts & deletes in 200K table + // 100.42ms // 97.40ms // 102.39ms : 500,000 inserts & deletes in 2M table + // 119.71ms // 120.59ms // 121.63ms : 500,000 inserts & deletes in 20M table + // 185.28ms // 195.15ms // 187.74ms : 500,000 inserts & deletes in 200M table + // 15.58ms // 14.79ms // 15.52ms : 200,000 inserts creating 200K table with varying key spacing + + return (((size_t) hash << 16 << 16) | hash) ^ seed; + } else if (len == 8 && sizeof(size_t) == 8) { + size_t hash = d[0] | (d[1] << 8) | (d[2] << 16) | (d[3] << 24); + hash |= (size_t) (d[4] | (d[5] << 8) | (d[6] << 16) | (d[7] << 24)) << 16 << 16; // avoid warning if size_t == 4 + hash ^= seed; + hash = (~hash) + (hash << 21); + hash ^= STBDS_ROTATE_RIGHT(hash,24); + hash *= 265; + hash ^= STBDS_ROTATE_RIGHT(hash,14); + hash ^= seed; + hash *= 21; + hash ^= STBDS_ROTATE_RIGHT(hash,28); + hash += (hash << 31); + hash = (~hash) + (hash << 18); + return hash; + } else { + return stbds_siphash_bytes(p,len,seed); + } +#endif +} +#ifdef _MSC_VER +#pragma warning(pop) +#endif + + +static int stbds_is_key_equal(void *a, size_t elemsize, void *key, size_t keysize, size_t keyoffset, int mode, size_t i) +{ + if (mode >= STBDS_HM_STRING) + return 0==strcmp((char *) key, * (char **) ((char *) a + elemsize*i + keyoffset)); + else + return 0==memcmp(key, (char *) a + elemsize*i + keyoffset, keysize); +} + +#define STBDS_HASH_TO_ARR(x,elemsize) ((char*) (x) - (elemsize)) +#define STBDS_ARR_TO_HASH(x,elemsize) ((char*) (x) + (elemsize)) + +#define stbds_hash_table(a) ((stbds_hash_index *) stbds_header(a)->hash_table) + +void stbds_hmfree_func(void *a, size_t elemsize) +{ + if (a == NULL) return; + if (stbds_hash_table(a) != NULL) { + if (stbds_hash_table(a)->string.mode == STBDS_SH_STRDUP) { + size_t i; + // skip 0th element, which is default + for (i=1; i < stbds_header(a)->length; ++i) + STBDS_FREE(NULL, *(char**) ((char *) a + elemsize*i)); + } + stbds_strreset(&stbds_hash_table(a)->string); + } + STBDS_FREE(NULL, stbds_header(a)->hash_table); + STBDS_FREE(NULL, stbds_header(a)); +} + +static ptrdiff_t stbds_hm_find_slot(void *a, size_t elemsize, void *key, size_t keysize, size_t keyoffset, int mode) +{ + void *raw_a = STBDS_HASH_TO_ARR(a,elemsize); + stbds_hash_index *table = stbds_hash_table(raw_a); + size_t hash = mode >= STBDS_HM_STRING ? stbds_hash_string((char*)key,table->seed) : stbds_hash_bytes(key, keysize,table->seed); + size_t step = STBDS_BUCKET_LENGTH; + size_t limit,i; + size_t pos; + stbds_hash_bucket *bucket; + + if (hash < 2) hash += 2; // stored hash values are forbidden from being 0, so we can detect empty slots + + pos = stbds_probe_position(hash, table->slot_count, table->slot_count_log2); + + for (;;) { + STBDS_STATS(++stbds_hash_probes); + bucket = &table->storage[pos >> STBDS_BUCKET_SHIFT]; + + // start searching from pos to end of bucket, this should help performance on small hash tables that fit in cache + for (i=pos & STBDS_BUCKET_MASK; i < STBDS_BUCKET_LENGTH; ++i) { + if (bucket->hash[i] == hash) { + if (stbds_is_key_equal(a, elemsize, key, keysize, keyoffset, mode, bucket->index[i])) { + return (pos & ~STBDS_BUCKET_MASK)+i; + } + } else if (bucket->hash[i] == STBDS_HASH_EMPTY) { + return -1; + } + } + + // search from beginning of bucket to pos + limit = pos & STBDS_BUCKET_MASK; + for (i = 0; i < limit; ++i) { + if (bucket->hash[i] == hash) { + if (stbds_is_key_equal(a, elemsize, key, keysize, keyoffset, mode, bucket->index[i])) { + return (pos & ~STBDS_BUCKET_MASK)+i; + } + } else if (bucket->hash[i] == STBDS_HASH_EMPTY) { + return -1; + } + } + + // quadratic probing + pos += step; + step += STBDS_BUCKET_LENGTH; + pos &= (table->slot_count-1); + } + /* NOTREACHED */ +} + +void * stbds_hmget_key_ts(void *a, size_t elemsize, void *key, size_t keysize, ptrdiff_t *temp, int mode) +{ + size_t keyoffset = 0; + if (a == NULL) { + // make it non-empty so we can return a temp + a = stbds_arrgrowf(0, elemsize, 0, 1); + stbds_header(a)->length += 1; + memset(a, 0, elemsize); + *temp = STBDS_INDEX_EMPTY; + // adjust a to point after the default element + return STBDS_ARR_TO_HASH(a,elemsize); + } else { + stbds_hash_index *table; + void *raw_a = STBDS_HASH_TO_ARR(a,elemsize); + // adjust a to point to the default element + table = (stbds_hash_index *) stbds_header(raw_a)->hash_table; + if (table == 0) { + *temp = -1; + } else { + ptrdiff_t slot = stbds_hm_find_slot(a, elemsize, key, keysize, keyoffset, mode); + if (slot < 0) { + *temp = STBDS_INDEX_EMPTY; + } else { + stbds_hash_bucket *b = &table->storage[slot >> STBDS_BUCKET_SHIFT]; + *temp = b->index[slot & STBDS_BUCKET_MASK]; + } + } + return a; + } +} + +void * stbds_hmget_key(void *a, size_t elemsize, void *key, size_t keysize, int mode) +{ + ptrdiff_t temp; + void *p = stbds_hmget_key_ts(a, elemsize, key, keysize, &temp, mode); + stbds_temp(STBDS_HASH_TO_ARR(p,elemsize)) = temp; + return p; +} + +void * stbds_hmput_default(void *a, size_t elemsize) +{ + // three cases: + // a is NULL <- allocate + // a has a hash table but no entries, because of shmode <- grow + // a has entries <- do nothing + if (a == NULL || stbds_header(STBDS_HASH_TO_ARR(a,elemsize))->length == 0) { + a = stbds_arrgrowf(a ? STBDS_HASH_TO_ARR(a,elemsize) : NULL, elemsize, 0, 1); + stbds_header(a)->length += 1; + memset(a, 0, elemsize); + a=STBDS_ARR_TO_HASH(a,elemsize); + } + return a; +} + +static char *stbds_strdup(char *str); + +void *stbds_hmput_key(void *a, size_t elemsize, void *key, size_t keysize, int mode) +{ + size_t keyoffset=0; + void *raw_a; + stbds_hash_index *table; + + if (a == NULL) { + a = stbds_arrgrowf(0, elemsize, 0, 1); + memset(a, 0, elemsize); + stbds_header(a)->length += 1; + // adjust a to point AFTER the default element + a = STBDS_ARR_TO_HASH(a,elemsize); + } + + // adjust a to point to the default element + raw_a = a; + a = STBDS_HASH_TO_ARR(a,elemsize); + + table = (stbds_hash_index *) stbds_header(a)->hash_table; + + if (table == NULL || table->used_count >= table->used_count_threshold) { + stbds_hash_index *nt; + size_t slot_count; + + slot_count = (table == NULL) ? STBDS_BUCKET_LENGTH : table->slot_count*2; + nt = stbds_make_hash_index(slot_count, table); + if (table) + STBDS_FREE(NULL, table); + else + nt->string.mode = mode >= STBDS_HM_STRING ? STBDS_SH_DEFAULT : 0; + stbds_header(a)->hash_table = table = nt; + STBDS_STATS(++stbds_hash_grow); + } + + // we iterate hash table explicitly because we want to track if we saw a tombstone + { + size_t hash = mode >= STBDS_HM_STRING ? stbds_hash_string((char*)key,table->seed) : stbds_hash_bytes(key, keysize,table->seed); + size_t step = STBDS_BUCKET_LENGTH; + size_t pos; + ptrdiff_t tombstone = -1; + stbds_hash_bucket *bucket; + + // stored hash values are forbidden from being 0, so we can detect empty slots to early out quickly + if (hash < 2) hash += 2; + + pos = stbds_probe_position(hash, table->slot_count, table->slot_count_log2); + + for (;;) { + size_t limit, i; + STBDS_STATS(++stbds_hash_probes); + bucket = &table->storage[pos >> STBDS_BUCKET_SHIFT]; + + // start searching from pos to end of bucket + for (i=pos & STBDS_BUCKET_MASK; i < STBDS_BUCKET_LENGTH; ++i) { + if (bucket->hash[i] == hash) { + if (stbds_is_key_equal(raw_a, elemsize, key, keysize, keyoffset, mode, bucket->index[i])) { + stbds_temp(a) = bucket->index[i]; + if (mode >= STBDS_HM_STRING) + stbds_temp_key(a) = * (char **) ((char *) raw_a + elemsize*bucket->index[i] + keyoffset); + return STBDS_ARR_TO_HASH(a,elemsize); + } + } else if (bucket->hash[i] == 0) { + pos = (pos & ~STBDS_BUCKET_MASK) + i; + goto found_empty_slot; + } else if (tombstone < 0) { + if (bucket->index[i] == STBDS_INDEX_DELETED) + tombstone = (ptrdiff_t) ((pos & ~STBDS_BUCKET_MASK) + i); + } + } + + // search from beginning of bucket to pos + limit = pos & STBDS_BUCKET_MASK; + for (i = 0; i < limit; ++i) { + if (bucket->hash[i] == hash) { + if (stbds_is_key_equal(raw_a, elemsize, key, keysize, keyoffset, mode, bucket->index[i])) { + stbds_temp(a) = bucket->index[i]; + return STBDS_ARR_TO_HASH(a,elemsize); + } + } else if (bucket->hash[i] == 0) { + pos = (pos & ~STBDS_BUCKET_MASK) + i; + goto found_empty_slot; + } else if (tombstone < 0) { + if (bucket->index[i] == STBDS_INDEX_DELETED) + tombstone = (ptrdiff_t) ((pos & ~STBDS_BUCKET_MASK) + i); + } + } + + // quadratic probing + pos += step; + step += STBDS_BUCKET_LENGTH; + pos &= (table->slot_count-1); + } + found_empty_slot: + if (tombstone >= 0) { + pos = tombstone; + --table->tombstone_count; + } + ++table->used_count; + + { + ptrdiff_t i = (ptrdiff_t) stbds_arrlen(a); + // we want to do stbds_arraddn(1), but we can't use the macros since we don't have something of the right type + if ((size_t) i+1 > stbds_arrcap(a)) + *(void **) &a = stbds_arrgrowf(a, elemsize, 1, 0); + raw_a = STBDS_ARR_TO_HASH(a,elemsize); + + STBDS_ASSERT((size_t) i+1 <= stbds_arrcap(a)); + stbds_header(a)->length = i+1; + bucket = &table->storage[pos >> STBDS_BUCKET_SHIFT]; + bucket->hash[pos & STBDS_BUCKET_MASK] = hash; + bucket->index[pos & STBDS_BUCKET_MASK] = i-1; + stbds_temp(a) = i-1; + + switch (table->string.mode) { + case STBDS_SH_STRDUP: stbds_temp_key(a) = *(char **) ((char *) a + elemsize*i) = stbds_strdup((char*) key); break; + case STBDS_SH_ARENA: stbds_temp_key(a) = *(char **) ((char *) a + elemsize*i) = stbds_stralloc(&table->string, (char*)key); break; + case STBDS_SH_DEFAULT: stbds_temp_key(a) = *(char **) ((char *) a + elemsize*i) = (char *) key; break; + default: memcpy((char *) a + elemsize*i, key, keysize); break; + } + } + return STBDS_ARR_TO_HASH(a,elemsize); + } +} + +void * stbds_shmode_func(size_t elemsize, int mode) +{ + void *a = stbds_arrgrowf(0, elemsize, 0, 1); + stbds_hash_index *h; + memset(a, 0, elemsize); + stbds_header(a)->length = 1; + stbds_header(a)->hash_table = h = (stbds_hash_index *) stbds_make_hash_index(STBDS_BUCKET_LENGTH, NULL); + h->string.mode = (unsigned char) mode; + return STBDS_ARR_TO_HASH(a,elemsize); +} + +void * stbds_hmdel_key(void *a, size_t elemsize, void *key, size_t keysize, size_t keyoffset, int mode) +{ + if (a == NULL) { + return 0; + } else { + stbds_hash_index *table; + void *raw_a = STBDS_HASH_TO_ARR(a,elemsize); + table = (stbds_hash_index *) stbds_header(raw_a)->hash_table; + stbds_temp(raw_a) = 0; + if (table == 0) { + return a; + } else { + ptrdiff_t slot; + slot = stbds_hm_find_slot(a, elemsize, key, keysize, keyoffset, mode); + if (slot < 0) + return a; + else { + stbds_hash_bucket *b = &table->storage[slot >> STBDS_BUCKET_SHIFT]; + int i = slot & STBDS_BUCKET_MASK; + ptrdiff_t old_index = b->index[i]; + ptrdiff_t final_index = (ptrdiff_t) stbds_arrlen(raw_a)-1-1; // minus one for the raw_a vs a, and minus one for 'last' + STBDS_ASSERT(slot < (ptrdiff_t) table->slot_count); + --table->used_count; + ++table->tombstone_count; + stbds_temp(raw_a) = 1; + STBDS_ASSERT(table->used_count >= 0); + //STBDS_ASSERT(table->tombstone_count < table->slot_count/4); + b->hash[i] = STBDS_HASH_DELETED; + b->index[i] = STBDS_INDEX_DELETED; + + if (mode == STBDS_HM_STRING && table->string.mode == STBDS_SH_STRDUP) + STBDS_FREE(NULL, *(char**) ((char *) a+elemsize*old_index)); + + // if indices are the same, memcpy is a no-op, but back-pointer-fixup will fail, so skip + if (old_index != final_index) { + // swap delete + memmove((char*) a + elemsize*old_index, (char*) a + elemsize*final_index, elemsize); + + // now find the slot for the last element + if (mode == STBDS_HM_STRING) + slot = stbds_hm_find_slot(a, elemsize, *(char**) ((char *) a+elemsize*old_index + keyoffset), keysize, keyoffset, mode); + else + slot = stbds_hm_find_slot(a, elemsize, (char* ) a+elemsize*old_index + keyoffset, keysize, keyoffset, mode); + STBDS_ASSERT(slot >= 0); + b = &table->storage[slot >> STBDS_BUCKET_SHIFT]; + i = slot & STBDS_BUCKET_MASK; + STBDS_ASSERT(b->index[i] == final_index); + b->index[i] = old_index; + } + stbds_header(raw_a)->length -= 1; + + if (table->used_count < table->used_count_shrink_threshold && table->slot_count > STBDS_BUCKET_LENGTH) { + stbds_header(raw_a)->hash_table = stbds_make_hash_index(table->slot_count>>1, table); + STBDS_FREE(NULL, table); + STBDS_STATS(++stbds_hash_shrink); + } else if (table->tombstone_count > table->tombstone_count_threshold) { + stbds_header(raw_a)->hash_table = stbds_make_hash_index(table->slot_count , table); + STBDS_FREE(NULL, table); + STBDS_STATS(++stbds_hash_rebuild); + } + + return a; + } + } + } + /* NOTREACHED */ +} + +static char *stbds_strdup(char *str) +{ + // to keep replaceable allocator simple, we don't want to use strdup. + // rolling our own also avoids problem of strdup vs _strdup + size_t len = strlen(str)+1; + char *p = (char*) STBDS_REALLOC(NULL, 0, len); + memmove(p, str, len); + return p; +} + +#ifndef STBDS_STRING_ARENA_BLOCKSIZE_MIN +#define STBDS_STRING_ARENA_BLOCKSIZE_MIN 512u +#endif +#ifndef STBDS_STRING_ARENA_BLOCKSIZE_MAX +#define STBDS_STRING_ARENA_BLOCKSIZE_MAX (1u<<20) +#endif + +char *stbds_stralloc(stbds_string_arena *a, char *str) +{ + char *p; + size_t len = strlen(str)+1; + if (len > a->remaining) { + // compute the next blocksize + size_t blocksize = a->block; + + // size is 512, 512, 1024, 1024, 2048, 2048, 4096, 4096, etc., so that + // there are log(SIZE) allocations to free when we destroy the table + blocksize = (size_t) (STBDS_STRING_ARENA_BLOCKSIZE_MIN) << (blocksize>>1); + + // if size is under 1M, advance to next blocktype + if (blocksize < (size_t)(STBDS_STRING_ARENA_BLOCKSIZE_MAX)) + ++a->block; + + if (len > blocksize) { + // if string is larger than blocksize, then just allocate the full size. + // note that we still advance string_block so block size will continue + // increasing, so e.g. if somebody only calls this with 1000-long strings, + // eventually the arena will start doubling and handling those as well + stbds_string_block *sb = (stbds_string_block *) STBDS_REALLOC(NULL, 0, sizeof(*sb)-8 + len); + memmove(sb->storage, str, len); + if (a->storage) { + // insert it after the first element, so that we don't waste the space there + sb->next = a->storage->next; + a->storage->next = sb; + } else { + sb->next = 0; + a->storage = sb; + a->remaining = 0; // this is redundant, but good for clarity + } + return sb->storage; + } else { + stbds_string_block *sb = (stbds_string_block *) STBDS_REALLOC(NULL, 0, sizeof(*sb)-8 + blocksize); + sb->next = a->storage; + a->storage = sb; + a->remaining = blocksize; + } + } + + STBDS_ASSERT(len <= a->remaining); + p = a->storage->storage + a->remaining - len; + a->remaining -= len; + memmove(p, str, len); + return p; +} + +void stbds_strreset(stbds_string_arena *a) +{ + stbds_string_block *x,*y; + x = a->storage; + while (x) { + y = x->next; + STBDS_FREE(NULL, x); + x = y; + } + memset(a, 0, sizeof(*a)); +} + +#endif + +////////////////////////////////////////////////////////////////////////////// +// +// UNIT TESTS +// + +#ifdef STBDS_UNIT_TESTS +#include +#ifdef STBDS_ASSERT_WAS_UNDEFINED +#undef STBDS_ASSERT +#endif +#ifndef STBDS_ASSERT +#define STBDS_ASSERT assert +#include +#endif + +typedef struct { int key,b,c,d; } stbds_struct; +typedef struct { int key[2],b,c,d; } stbds_struct2; + +static char buffer[256]; +char *strkey(int n) +{ +#if defined(_WIN32) && defined(__STDC_WANT_SECURE_LIB__) + sprintf_s(buffer, sizeof(buffer), "test_%d", n); +#else + sprintf(buffer, "test_%d", n); +#endif + return buffer; +} + +void stbds_unit_tests(void) +{ +#if defined(_MSC_VER) && _MSC_VER <= 1200 && defined(__cplusplus) + // VC6 C++ doesn't like the template<> trick on unnamed structures, so do nothing! + STBDS_ASSERT(0); +#else + const int testsize = 100000; + const int testsize2 = testsize/20; + int *arr=NULL; + struct { int key; int value; } *intmap = NULL; + struct { char *key; int value; } *strmap = NULL, s; + struct { stbds_struct key; int value; } *map = NULL; + stbds_struct *map2 = NULL; + stbds_struct2 *map3 = NULL; + stbds_string_arena sa = { 0 }; + int key3[2] = { 1,2 }; + ptrdiff_t temp; + + int i,j; + + STBDS_ASSERT(arrlen(arr)==0); + for (i=0; i < 20000; i += 50) { + for (j=0; j < i; ++j) + arrpush(arr,j); + arrfree(arr); + } + + for (i=0; i < 4; ++i) { + arrpush(arr,1); arrpush(arr,2); arrpush(arr,3); arrpush(arr,4); + arrdel(arr,i); + arrfree(arr); + arrpush(arr,1); arrpush(arr,2); arrpush(arr,3); arrpush(arr,4); + arrdelswap(arr,i); + arrfree(arr); + } + + for (i=0; i < 5; ++i) { + arrpush(arr,1); arrpush(arr,2); arrpush(arr,3); arrpush(arr,4); + stbds_arrins(arr,i,5); + STBDS_ASSERT(arr[i] == 5); + if (i < 4) + STBDS_ASSERT(arr[4] == 4); + arrfree(arr); + } + + i = 1; + STBDS_ASSERT(hmgeti(intmap,i) == -1); + hmdefault(intmap, -2); + STBDS_ASSERT(hmgeti(intmap, i) == -1); + STBDS_ASSERT(hmget (intmap, i) == -2); + for (i=0; i < testsize; i+=2) + hmput(intmap, i, i*5); + for (i=0; i < testsize; i+=1) { + if (i & 1) STBDS_ASSERT(hmget(intmap, i) == -2 ); + else STBDS_ASSERT(hmget(intmap, i) == i*5); + if (i & 1) STBDS_ASSERT(hmget_ts(intmap, i, temp) == -2 ); + else STBDS_ASSERT(hmget_ts(intmap, i, temp) == i*5); + } + for (i=0; i < testsize; i+=2) + hmput(intmap, i, i*3); + for (i=0; i < testsize; i+=1) + if (i & 1) STBDS_ASSERT(hmget(intmap, i) == -2 ); + else STBDS_ASSERT(hmget(intmap, i) == i*3); + for (i=2; i < testsize; i+=4) + hmdel(intmap, i); // delete half the entries + for (i=0; i < testsize; i+=1) + if (i & 3) STBDS_ASSERT(hmget(intmap, i) == -2 ); + else STBDS_ASSERT(hmget(intmap, i) == i*3); + for (i=0; i < testsize; i+=1) + hmdel(intmap, i); // delete the rest of the entries + for (i=0; i < testsize; i+=1) + STBDS_ASSERT(hmget(intmap, i) == -2 ); + hmfree(intmap); + for (i=0; i < testsize; i+=2) + hmput(intmap, i, i*3); + hmfree(intmap); + + #if defined(__clang__) || defined(__GNUC__) + #ifndef __cplusplus + intmap = NULL; + hmput(intmap, 15, 7); + hmput(intmap, 11, 3); + hmput(intmap, 9, 5); + STBDS_ASSERT(hmget(intmap, 9) == 5); + STBDS_ASSERT(hmget(intmap, 11) == 3); + STBDS_ASSERT(hmget(intmap, 15) == 7); + #endif + #endif + + for (i=0; i < testsize; ++i) + stralloc(&sa, strkey(i)); + strreset(&sa); + + { + s.key = "a", s.value = 1; + shputs(strmap, s); + STBDS_ASSERT(*strmap[0].key == 'a'); + STBDS_ASSERT(strmap[0].key == s.key); + STBDS_ASSERT(strmap[0].value == s.value); + shfree(strmap); + } + + { + s.key = "a", s.value = 1; + sh_new_strdup(strmap); + shputs(strmap, s); + STBDS_ASSERT(*strmap[0].key == 'a'); + STBDS_ASSERT(strmap[0].key != s.key); + STBDS_ASSERT(strmap[0].value == s.value); + shfree(strmap); + } + + { + s.key = "a", s.value = 1; + sh_new_arena(strmap); + shputs(strmap, s); + STBDS_ASSERT(*strmap[0].key == 'a'); + STBDS_ASSERT(strmap[0].key != s.key); + STBDS_ASSERT(strmap[0].value == s.value); + shfree(strmap); + } + + for (j=0; j < 2; ++j) { + STBDS_ASSERT(shgeti(strmap,"foo") == -1); + if (j == 0) + sh_new_strdup(strmap); + else + sh_new_arena(strmap); + STBDS_ASSERT(shgeti(strmap,"foo") == -1); + shdefault(strmap, -2); + STBDS_ASSERT(shgeti(strmap,"foo") == -1); + for (i=0; i < testsize; i+=2) + shput(strmap, strkey(i), i*3); + for (i=0; i < testsize; i+=1) + if (i & 1) STBDS_ASSERT(shget(strmap, strkey(i)) == -2 ); + else STBDS_ASSERT(shget(strmap, strkey(i)) == i*3); + for (i=2; i < testsize; i+=4) + shdel(strmap, strkey(i)); // delete half the entries + for (i=0; i < testsize; i+=1) + if (i & 3) STBDS_ASSERT(shget(strmap, strkey(i)) == -2 ); + else STBDS_ASSERT(shget(strmap, strkey(i)) == i*3); + for (i=0; i < testsize; i+=1) + shdel(strmap, strkey(i)); // delete the rest of the entries + for (i=0; i < testsize; i+=1) + STBDS_ASSERT(shget(strmap, strkey(i)) == -2 ); + shfree(strmap); + } + + { + struct { char *key; char value; } *hash = NULL; + char name[4] = "jen"; + shput(hash, "bob" , 'h'); + shput(hash, "sally" , 'e'); + shput(hash, "fred" , 'l'); + shput(hash, "jen" , 'x'); + shput(hash, "doug" , 'o'); + + shput(hash, name , 'l'); + shfree(hash); + } + + for (i=0; i < testsize; i += 2) { + stbds_struct s = { i,i*2,i*3,i*4 }; + hmput(map, s, i*5); + } + + for (i=0; i < testsize; i += 1) { + stbds_struct s = { i,i*2,i*3 ,i*4 }; + stbds_struct t = { i,i*2,i*3+1,i*4 }; + if (i & 1) STBDS_ASSERT(hmget(map, s) == 0); + else STBDS_ASSERT(hmget(map, s) == i*5); + if (i & 1) STBDS_ASSERT(hmget_ts(map, s, temp) == 0); + else STBDS_ASSERT(hmget_ts(map, s, temp) == i*5); + //STBDS_ASSERT(hmget(map, t.key) == 0); + } + + for (i=0; i < testsize; i += 2) { + stbds_struct s = { i,i*2,i*3,i*4 }; + hmputs(map2, s); + } + hmfree(map); + + for (i=0; i < testsize; i += 1) { + stbds_struct s = { i,i*2,i*3,i*4 }; + stbds_struct t = { i,i*2,i*3+1,i*4 }; + if (i & 1) STBDS_ASSERT(hmgets(map2, s.key).d == 0); + else STBDS_ASSERT(hmgets(map2, s.key).d == i*4); + //STBDS_ASSERT(hmgetp(map2, t.key) == 0); + } + hmfree(map2); + + for (i=0; i < testsize; i += 2) { + stbds_struct2 s = { { i,i*2 }, i*3,i*4, i*5 }; + hmputs(map3, s); + } + for (i=0; i < testsize; i += 1) { + stbds_struct2 s = { { i,i*2}, i*3, i*4, i*5 }; + stbds_struct2 t = { { i,i*2}, i*3+1, i*4, i*5 }; + if (i & 1) STBDS_ASSERT(hmgets(map3, s.key).d == 0); + else STBDS_ASSERT(hmgets(map3, s.key).d == i*5); + //STBDS_ASSERT(hmgetp(map3, t.key) == 0); + } +#endif +} +#endif + + +/* +------------------------------------------------------------------------------ +This software is available under 2 licenses -- choose whichever you prefer. +------------------------------------------------------------------------------ +ALTERNATIVE A - MIT License +Copyright (c) 2019 Sean Barrett +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. +------------------------------------------------------------------------------ +ALTERNATIVE B - Public Domain (www.unlicense.org) +This is free and unencumbered software released into the public domain. +Anyone is free to copy, modify, publish, use, compile, sell, or distribute this +software, either in source code form or as a compiled binary, for any purpose, +commercial or non-commercial, and by any means. +In jurisdictions that recognize copyright laws, the author or authors of this +software dedicate any and all copyright interest in the software to the public +domain. We make this dedication for the benefit of the public at large and to +the detriment of our heirs and successors. We intend this dedication to be an +overt act of relinquishment in perpetuity of all present and future rights to +this software under copyright law. +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 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. +------------------------------------------------------------------------------ +*/ diff --git a/3rdparty/stb_image.h b/3rdparty/stb_image.h new file mode 100644 index 0000000..b53e6e3 --- /dev/null +++ b/3rdparty/stb_image.h @@ -0,0 +1,7897 @@ +/* stb_image - v2.27 - public domain image loader - http://nothings.org/stb + no warranty implied; use at your own risk + + Do this: + #define STB_IMAGE_IMPLEMENTATION + before you include this file in *one* C or C++ file to create the implementation. + + // i.e. it should look like this: + #include ... + #include ... + #include ... + #define STB_IMAGE_IMPLEMENTATION + #include "stb_image.h" + + You can #define STBI_ASSERT(x) before the #include to avoid using assert.h. + And #define STBI_MALLOC, STBI_REALLOC, and STBI_FREE to avoid using malloc,realloc,free + + + QUICK NOTES: + Primarily of interest to game developers and other people who can + avoid problematic images and only need the trivial interface + + JPEG baseline & progressive (12 bpc/arithmetic not supported, same as stock IJG lib) + PNG 1/2/4/8/16-bit-per-channel + + TGA (not sure what subset, if a subset) + BMP non-1bpp, non-RLE + PSD (composited view only, no extra channels, 8/16 bit-per-channel) + + GIF (*comp always reports as 4-channel) + HDR (radiance rgbE format) + PIC (Softimage PIC) + PNM (PPM and PGM binary only) + + Animated GIF still needs a proper API, but here's one way to do it: + http://gist.github.com/urraka/685d9a6340b26b830d49 + + - decode from memory or through FILE (define STBI_NO_STDIO to remove code) + - decode from arbitrary I/O callbacks + - SIMD acceleration on x86/x64 (SSE2) and ARM (NEON) + + Full documentation under "DOCUMENTATION" below. + + +LICENSE + + See end of file for license information. + +RECENT REVISION HISTORY: + + 2.27 (2021-07-11) document stbi_info better, 16-bit PNM support, bug fixes + 2.26 (2020-07-13) many minor fixes + 2.25 (2020-02-02) fix warnings + 2.24 (2020-02-02) fix warnings; thread-local failure_reason and flip_vertically + 2.23 (2019-08-11) fix clang static analysis warning + 2.22 (2019-03-04) gif fixes, fix warnings + 2.21 (2019-02-25) fix typo in comment + 2.20 (2019-02-07) support utf8 filenames in Windows; fix warnings and platform ifdefs + 2.19 (2018-02-11) fix warning + 2.18 (2018-01-30) fix warnings + 2.17 (2018-01-29) bugfix, 1-bit BMP, 16-bitness query, fix warnings + 2.16 (2017-07-23) all functions have 16-bit variants; optimizations; bugfixes + 2.15 (2017-03-18) fix png-1,2,4; all Imagenet JPGs; no runtime SSE detection on GCC + 2.14 (2017-03-03) remove deprecated STBI_JPEG_OLD; fixes for Imagenet JPGs + 2.13 (2016-12-04) experimental 16-bit API, only for PNG so far; fixes + 2.12 (2016-04-02) fix typo in 2.11 PSD fix that caused crashes + 2.11 (2016-04-02) 16-bit PNGS; enable SSE2 in non-gcc x64 + RGB-format JPEG; remove white matting in PSD; + allocate large structures on the stack; + correct channel count for PNG & BMP + 2.10 (2016-01-22) avoid warning introduced in 2.09 + 2.09 (2016-01-16) 16-bit TGA; comments in PNM files; STBI_REALLOC_SIZED + + See end of file for full revision history. + + + ============================ Contributors ========================= + + Image formats Extensions, features + Sean Barrett (jpeg, png, bmp) Jetro Lauha (stbi_info) + Nicolas Schulz (hdr, psd) Martin "SpartanJ" Golini (stbi_info) + Jonathan Dummer (tga) James "moose2000" Brown (iPhone PNG) + Jean-Marc Lienher (gif) Ben "Disch" Wenger (io callbacks) + Tom Seddon (pic) Omar Cornut (1/2/4-bit PNG) + Thatcher Ulrich (psd) Nicolas Guillemot (vertical flip) + Ken Miller (pgm, ppm) Richard Mitton (16-bit PSD) + github:urraka (animated gif) Junggon Kim (PNM comments) + Christopher Forseth (animated gif) Daniel Gibson (16-bit TGA) + socks-the-fox (16-bit PNG) + Jeremy Sawicki (handle all ImageNet JPGs) + Optimizations & bugfixes Mikhail Morozov (1-bit BMP) + Fabian "ryg" Giesen Anael Seghezzi (is-16-bit query) + Arseny Kapoulkine Simon Breuss (16-bit PNM) + John-Mark Allen + Carmelo J Fdez-Aguera + + Bug & warning fixes + Marc LeBlanc David Woo Guillaume George Martins Mozeiko + Christpher Lloyd Jerry Jansson Joseph Thomson Blazej Dariusz Roszkowski + Phil Jordan Dave Moore Roy Eltham + Hayaki Saito Nathan Reed Won Chun + Luke Graham Johan Duparc Nick Verigakis the Horde3D community + Thomas Ruf Ronny Chevalier github:rlyeh + Janez Zemva John Bartholomew Michal Cichon github:romigrou + Jonathan Blow Ken Hamada Tero Hanninen github:svdijk + Eugene Golushkov Laurent Gomila Cort Stratton github:snagar + Aruelien Pocheville Sergio Gonzalez Thibault Reuille github:Zelex + Cass Everitt Ryamond Barbiero github:grim210 + Paul Du Bois Engin Manap Aldo Culquicondor github:sammyhw + Philipp Wiesemann Dale Weiler Oriol Ferrer Mesia github:phprus + Josh Tobin Matthew Gregan github:poppolopoppo + Julian Raschke Gregory Mullen Christian Floisand github:darealshinji + Baldur Karlsson Kevin Schmidt JR Smith github:Michaelangel007 + Brad Weinberger Matvey Cherevko github:mosra + Luca Sas Alexander Veselov Zack Middleton [reserved] + Ryan C. Gordon [reserved] [reserved] + DO NOT ADD YOUR NAME HERE + + Jacko Dirks + + To add your name to the credits, pick a random blank space in the middle and fill it. + 80% of merge conflicts on stb PRs are due to people adding their name at the end + of the credits. +*/ + +#ifndef STBI_INCLUDE_STB_IMAGE_H +#define STBI_INCLUDE_STB_IMAGE_H + +// DOCUMENTATION +// +// Limitations: +// - no 12-bit-per-channel JPEG +// - no JPEGs with arithmetic coding +// - GIF always returns *comp=4 +// +// Basic usage (see HDR discussion below for HDR usage): +// int x,y,n; +// unsigned char *data = stbi_load(filename, &x, &y, &n, 0); +// // ... process data if not NULL ... +// // ... x = width, y = height, n = # 8-bit components per pixel ... +// // ... replace '0' with '1'..'4' to force that many components per pixel +// // ... but 'n' will always be the number that it would have been if you said 0 +// stbi_image_free(data) +// +// Standard parameters: +// int *x -- outputs image width in pixels +// int *y -- outputs image height in pixels +// int *channels_in_file -- outputs # of image components in image file +// int desired_channels -- if non-zero, # of image components requested in result +// +// The return value from an image loader is an 'unsigned char *' which points +// to the pixel data, or NULL on an allocation failure or if the image is +// corrupt or invalid. The pixel data consists of *y scanlines of *x pixels, +// with each pixel consisting of N interleaved 8-bit components; the first +// pixel pointed to is top-left-most in the image. There is no padding between +// image scanlines or between pixels, regardless of format. The number of +// components N is 'desired_channels' if desired_channels is non-zero, or +// *channels_in_file otherwise. If desired_channels is non-zero, +// *channels_in_file has the number of components that _would_ have been +// output otherwise. E.g. if you set desired_channels to 4, you will always +// get RGBA output, but you can check *channels_in_file to see if it's trivially +// opaque because e.g. there were only 3 channels in the source image. +// +// An output image with N components has the following components interleaved +// in this order in each pixel: +// +// N=#comp components +// 1 grey +// 2 grey, alpha +// 3 red, green, blue +// 4 red, green, blue, alpha +// +// If image loading fails for any reason, the return value will be NULL, +// and *x, *y, *channels_in_file will be unchanged. The function +// stbi_failure_reason() can be queried for an extremely brief, end-user +// unfriendly explanation of why the load failed. Define STBI_NO_FAILURE_STRINGS +// to avoid compiling these strings at all, and STBI_FAILURE_USERMSG to get slightly +// more user-friendly ones. +// +// Paletted PNG, BMP, GIF, and PIC images are automatically depalettized. +// +// To query the width, height and component count of an image without having to +// decode the full file, you can use the stbi_info family of functions: +// +// int x,y,n,ok; +// ok = stbi_info(filename, &x, &y, &n); +// // returns ok=1 and sets x, y, n if image is a supported format, +// // 0 otherwise. +// +// Note that stb_image pervasively uses ints in its public API for sizes, +// including sizes of memory buffers. This is now part of the API and thus +// hard to change without causing breakage. As a result, the various image +// loaders all have certain limits on image size; these differ somewhat +// by format but generally boil down to either just under 2GB or just under +// 1GB. When the decoded image would be larger than this, stb_image decoding +// will fail. +// +// Additionally, stb_image will reject image files that have any of their +// dimensions set to a larger value than the configurable STBI_MAX_DIMENSIONS, +// which defaults to 2**24 = 16777216 pixels. Due to the above memory limit, +// the only way to have an image with such dimensions load correctly +// is for it to have a rather extreme aspect ratio. Either way, the +// assumption here is that such larger images are likely to be malformed +// or malicious. If you do need to load an image with individual dimensions +// larger than that, and it still fits in the overall size limit, you can +// #define STBI_MAX_DIMENSIONS on your own to be something larger. +// +// =========================================================================== +// +// UNICODE: +// +// If compiling for Windows and you wish to use Unicode filenames, compile +// with +// #define STBI_WINDOWS_UTF8 +// and pass utf8-encoded filenames. Call stbi_convert_wchar_to_utf8 to convert +// Windows wchar_t filenames to utf8. +// +// =========================================================================== +// +// Philosophy +// +// stb libraries are designed with the following priorities: +// +// 1. easy to use +// 2. easy to maintain +// 3. good performance +// +// Sometimes I let "good performance" creep up in priority over "easy to maintain", +// and for best performance I may provide less-easy-to-use APIs that give higher +// performance, in addition to the easy-to-use ones. Nevertheless, it's important +// to keep in mind that from the standpoint of you, a client of this library, +// all you care about is #1 and #3, and stb libraries DO NOT emphasize #3 above all. +// +// Some secondary priorities arise directly from the first two, some of which +// provide more explicit reasons why performance can't be emphasized. +// +// - Portable ("ease of use") +// - Small source code footprint ("easy to maintain") +// - No dependencies ("ease of use") +// +// =========================================================================== +// +// I/O callbacks +// +// I/O callbacks allow you to read from arbitrary sources, like packaged +// files or some other source. Data read from callbacks are processed +// through a small internal buffer (currently 128 bytes) to try to reduce +// overhead. +// +// The three functions you must define are "read" (reads some bytes of data), +// "skip" (skips some bytes of data), "eof" (reports if the stream is at the end). +// +// =========================================================================== +// +// SIMD support +// +// The JPEG decoder will try to automatically use SIMD kernels on x86 when +// supported by the compiler. For ARM Neon support, you must explicitly +// request it. +// +// (The old do-it-yourself SIMD API is no longer supported in the current +// code.) +// +// On x86, SSE2 will automatically be used when available based on a run-time +// test; if not, the generic C versions are used as a fall-back. On ARM targets, +// the typical path is to have separate builds for NEON and non-NEON devices +// (at least this is true for iOS and Android). Therefore, the NEON support is +// toggled by a build flag: define STBI_NEON to get NEON loops. +// +// If for some reason you do not want to use any of SIMD code, or if +// you have issues compiling it, you can disable it entirely by +// defining STBI_NO_SIMD. +// +// =========================================================================== +// +// HDR image support (disable by defining STBI_NO_HDR) +// +// stb_image supports loading HDR images in general, and currently the Radiance +// .HDR file format specifically. You can still load any file through the existing +// interface; if you attempt to load an HDR file, it will be automatically remapped +// to LDR, assuming gamma 2.2 and an arbitrary scale factor defaulting to 1; +// both of these constants can be reconfigured through this interface: +// +// stbi_hdr_to_ldr_gamma(2.2f); +// stbi_hdr_to_ldr_scale(1.0f); +// +// (note, do not use _inverse_ constants; stbi_image will invert them +// appropriately). +// +// Additionally, there is a new, parallel interface for loading files as +// (linear) floats to preserve the full dynamic range: +// +// float *data = stbi_loadf(filename, &x, &y, &n, 0); +// +// If you load LDR images through this interface, those images will +// be promoted to floating point values, run through the inverse of +// constants corresponding to the above: +// +// stbi_ldr_to_hdr_scale(1.0f); +// stbi_ldr_to_hdr_gamma(2.2f); +// +// Finally, given a filename (or an open file or memory block--see header +// file for details) containing image data, you can query for the "most +// appropriate" interface to use (that is, whether the image is HDR or +// not), using: +// +// stbi_is_hdr(char *filename); +// +// =========================================================================== +// +// iPhone PNG support: +// +// We optionally support converting iPhone-formatted PNGs (which store +// premultiplied BGRA) back to RGB, even though they're internally encoded +// differently. To enable this conversion, call +// stbi_convert_iphone_png_to_rgb(1). +// +// Call stbi_set_unpremultiply_on_load(1) as well to force a divide per +// pixel to remove any premultiplied alpha *only* if the image file explicitly +// says there's premultiplied data (currently only happens in iPhone images, +// and only if iPhone convert-to-rgb processing is on). +// +// =========================================================================== +// +// ADDITIONAL CONFIGURATION +// +// - You can suppress implementation of any of the decoders to reduce +// your code footprint by #defining one or more of the following +// symbols before creating the implementation. +// +// STBI_NO_JPEG +// STBI_NO_PNG +// STBI_NO_BMP +// STBI_NO_PSD +// STBI_NO_TGA +// STBI_NO_GIF +// STBI_NO_HDR +// STBI_NO_PIC +// STBI_NO_PNM (.ppm and .pgm) +// +// - You can request *only* certain decoders and suppress all other ones +// (this will be more forward-compatible, as addition of new decoders +// doesn't require you to disable them explicitly): +// +// STBI_ONLY_JPEG +// STBI_ONLY_PNG +// STBI_ONLY_BMP +// STBI_ONLY_PSD +// STBI_ONLY_TGA +// STBI_ONLY_GIF +// STBI_ONLY_HDR +// STBI_ONLY_PIC +// STBI_ONLY_PNM (.ppm and .pgm) +// +// - If you use STBI_NO_PNG (or _ONLY_ without PNG), and you still +// want the zlib decoder to be available, #define STBI_SUPPORT_ZLIB +// +// - If you define STBI_MAX_DIMENSIONS, stb_image will reject images greater +// than that size (in either width or height) without further processing. +// This is to let programs in the wild set an upper bound to prevent +// denial-of-service attacks on untrusted data, as one could generate a +// valid image of gigantic dimensions and force stb_image to allocate a +// huge block of memory and spend disproportionate time decoding it. By +// default this is set to (1 << 24), which is 16777216, but that's still +// very big. + +#ifndef STBI_NO_STDIO +#include +#endif // STBI_NO_STDIO + +#define STBI_VERSION 1 + +enum +{ + STBI_default = 0, // only used for desired_channels + + STBI_grey = 1, + STBI_grey_alpha = 2, + STBI_rgb = 3, + STBI_rgb_alpha = 4 +}; + +#include +typedef unsigned char stbi_uc; +typedef unsigned short stbi_us; + +#ifdef __cplusplus +extern "C" { +#endif + +#ifndef STBIDEF +#ifdef STB_IMAGE_STATIC +#define STBIDEF static +#else +#define STBIDEF extern +#endif +#endif + +////////////////////////////////////////////////////////////////////////////// +// +// PRIMARY API - works on images of any type +// + +// +// load image by filename, open file, or memory buffer +// + +typedef struct +{ + int (*read) (void *user,char *data,int size); // fill 'data' with 'size' bytes. return number of bytes actually read + void (*skip) (void *user,int n); // skip the next 'n' bytes, or 'unget' the last -n bytes if negative + int (*eof) (void *user); // returns nonzero if we are at end of file/data +} stbi_io_callbacks; + +//////////////////////////////////// +// +// 8-bits-per-channel interface +// + +STBIDEF stbi_uc *stbi_load_from_memory (stbi_uc const *buffer, int len , int *x, int *y, int *channels_in_file, int desired_channels); +STBIDEF stbi_uc *stbi_load_from_callbacks(stbi_io_callbacks const *clbk , void *user, int *x, int *y, int *channels_in_file, int desired_channels); + +#ifndef STBI_NO_STDIO +STBIDEF stbi_uc *stbi_load (char const *filename, int *x, int *y, int *channels_in_file, int desired_channels); +STBIDEF stbi_uc *stbi_load_from_file (FILE *f, int *x, int *y, int *channels_in_file, int desired_channels); +// for stbi_load_from_file, file pointer is left pointing immediately after image +#endif + +#ifndef STBI_NO_GIF +STBIDEF stbi_uc *stbi_load_gif_from_memory(stbi_uc const *buffer, int len, int **delays, int *x, int *y, int *z, int *comp, int req_comp); +#endif + +#ifdef STBI_WINDOWS_UTF8 +STBIDEF int stbi_convert_wchar_to_utf8(char *buffer, size_t bufferlen, const wchar_t* input); +#endif + +//////////////////////////////////// +// +// 16-bits-per-channel interface +// + +STBIDEF stbi_us *stbi_load_16_from_memory (stbi_uc const *buffer, int len, int *x, int *y, int *channels_in_file, int desired_channels); +STBIDEF stbi_us *stbi_load_16_from_callbacks(stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *channels_in_file, int desired_channels); + +#ifndef STBI_NO_STDIO +STBIDEF stbi_us *stbi_load_16 (char const *filename, int *x, int *y, int *channels_in_file, int desired_channels); +STBIDEF stbi_us *stbi_load_from_file_16(FILE *f, int *x, int *y, int *channels_in_file, int desired_channels); +#endif + +//////////////////////////////////// +// +// float-per-channel interface +// +#ifndef STBI_NO_LINEAR + STBIDEF float *stbi_loadf_from_memory (stbi_uc const *buffer, int len, int *x, int *y, int *channels_in_file, int desired_channels); + STBIDEF float *stbi_loadf_from_callbacks (stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *channels_in_file, int desired_channels); + + #ifndef STBI_NO_STDIO + STBIDEF float *stbi_loadf (char const *filename, int *x, int *y, int *channels_in_file, int desired_channels); + STBIDEF float *stbi_loadf_from_file (FILE *f, int *x, int *y, int *channels_in_file, int desired_channels); + #endif +#endif + +#ifndef STBI_NO_HDR + STBIDEF void stbi_hdr_to_ldr_gamma(float gamma); + STBIDEF void stbi_hdr_to_ldr_scale(float scale); +#endif // STBI_NO_HDR + +#ifndef STBI_NO_LINEAR + STBIDEF void stbi_ldr_to_hdr_gamma(float gamma); + STBIDEF void stbi_ldr_to_hdr_scale(float scale); +#endif // STBI_NO_LINEAR + +// stbi_is_hdr is always defined, but always returns false if STBI_NO_HDR +STBIDEF int stbi_is_hdr_from_callbacks(stbi_io_callbacks const *clbk, void *user); +STBIDEF int stbi_is_hdr_from_memory(stbi_uc const *buffer, int len); +#ifndef STBI_NO_STDIO +STBIDEF int stbi_is_hdr (char const *filename); +STBIDEF int stbi_is_hdr_from_file(FILE *f); +#endif // STBI_NO_STDIO + + +// get a VERY brief reason for failure +// on most compilers (and ALL modern mainstream compilers) this is threadsafe +STBIDEF const char *stbi_failure_reason (void); + +// free the loaded image -- this is just free() +STBIDEF void stbi_image_free (void *retval_from_stbi_load); + +// get image dimensions & components without fully decoding +STBIDEF int stbi_info_from_memory(stbi_uc const *buffer, int len, int *x, int *y, int *comp); +STBIDEF int stbi_info_from_callbacks(stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *comp); +STBIDEF int stbi_is_16_bit_from_memory(stbi_uc const *buffer, int len); +STBIDEF int stbi_is_16_bit_from_callbacks(stbi_io_callbacks const *clbk, void *user); + +#ifndef STBI_NO_STDIO +STBIDEF int stbi_info (char const *filename, int *x, int *y, int *comp); +STBIDEF int stbi_info_from_file (FILE *f, int *x, int *y, int *comp); +STBIDEF int stbi_is_16_bit (char const *filename); +STBIDEF int stbi_is_16_bit_from_file(FILE *f); +#endif + + + +// for image formats that explicitly notate that they have premultiplied alpha, +// we just return the colors as stored in the file. set this flag to force +// unpremultiplication. results are undefined if the unpremultiply overflow. +STBIDEF void stbi_set_unpremultiply_on_load(int flag_true_if_should_unpremultiply); + +// indicate whether we should process iphone images back to canonical format, +// or just pass them through "as-is" +STBIDEF void stbi_convert_iphone_png_to_rgb(int flag_true_if_should_convert); + +// flip the image vertically, so the first pixel in the output array is the bottom left +STBIDEF void stbi_set_flip_vertically_on_load(int flag_true_if_should_flip); + +// as above, but only applies to images loaded on the thread that calls the function +// this function is only available if your compiler supports thread-local variables; +// calling it will fail to link if your compiler doesn't +STBIDEF void stbi_set_unpremultiply_on_load_thread(int flag_true_if_should_unpremultiply); +STBIDEF void stbi_convert_iphone_png_to_rgb_thread(int flag_true_if_should_convert); +STBIDEF void stbi_set_flip_vertically_on_load_thread(int flag_true_if_should_flip); + +// ZLIB client - used by PNG, available for other purposes + +STBIDEF char *stbi_zlib_decode_malloc_guesssize(const char *buffer, int len, int initial_size, int *outlen); +STBIDEF char *stbi_zlib_decode_malloc_guesssize_headerflag(const char *buffer, int len, int initial_size, int *outlen, int parse_header); +STBIDEF char *stbi_zlib_decode_malloc(const char *buffer, int len, int *outlen); +STBIDEF int stbi_zlib_decode_buffer(char *obuffer, int olen, const char *ibuffer, int ilen); + +STBIDEF char *stbi_zlib_decode_noheader_malloc(const char *buffer, int len, int *outlen); +STBIDEF int stbi_zlib_decode_noheader_buffer(char *obuffer, int olen, const char *ibuffer, int ilen); + + +#ifdef __cplusplus +} +#endif + +// +// +//// end header file ///////////////////////////////////////////////////// +#endif // STBI_INCLUDE_STB_IMAGE_H + +#ifdef STB_IMAGE_IMPLEMENTATION + +#if defined(STBI_ONLY_JPEG) || defined(STBI_ONLY_PNG) || defined(STBI_ONLY_BMP) \ + || defined(STBI_ONLY_TGA) || defined(STBI_ONLY_GIF) || defined(STBI_ONLY_PSD) \ + || defined(STBI_ONLY_HDR) || defined(STBI_ONLY_PIC) || defined(STBI_ONLY_PNM) \ + || defined(STBI_ONLY_ZLIB) + #ifndef STBI_ONLY_JPEG + #define STBI_NO_JPEG + #endif + #ifndef STBI_ONLY_PNG + #define STBI_NO_PNG + #endif + #ifndef STBI_ONLY_BMP + #define STBI_NO_BMP + #endif + #ifndef STBI_ONLY_PSD + #define STBI_NO_PSD + #endif + #ifndef STBI_ONLY_TGA + #define STBI_NO_TGA + #endif + #ifndef STBI_ONLY_GIF + #define STBI_NO_GIF + #endif + #ifndef STBI_ONLY_HDR + #define STBI_NO_HDR + #endif + #ifndef STBI_ONLY_PIC + #define STBI_NO_PIC + #endif + #ifndef STBI_ONLY_PNM + #define STBI_NO_PNM + #endif +#endif + +#if defined(STBI_NO_PNG) && !defined(STBI_SUPPORT_ZLIB) && !defined(STBI_NO_ZLIB) +#define STBI_NO_ZLIB +#endif + + +#include +#include // ptrdiff_t on osx +#include +#include +#include + +#if !defined(STBI_NO_LINEAR) || !defined(STBI_NO_HDR) +#include // ldexp, pow +#endif + +#ifndef STBI_NO_STDIO +#include +#endif + +#ifndef STBI_ASSERT +#include +#define STBI_ASSERT(x) assert(x) +#endif + +#ifdef __cplusplus +#define STBI_EXTERN extern "C" +#else +#define STBI_EXTERN extern +#endif + + +#ifndef _MSC_VER + #ifdef __cplusplus + #define stbi_inline inline + #else + #define stbi_inline + #endif +#else + #define stbi_inline __forceinline +#endif + +#ifndef STBI_NO_THREAD_LOCALS + #if defined(__cplusplus) && __cplusplus >= 201103L + #define STBI_THREAD_LOCAL thread_local + #elif defined(__GNUC__) && __GNUC__ < 5 + #define STBI_THREAD_LOCAL __thread + #elif defined(_MSC_VER) + #define STBI_THREAD_LOCAL __declspec(thread) + #elif defined (__STDC_VERSION__) && __STDC_VERSION__ >= 201112L && !defined(__STDC_NO_THREADS__) + #define STBI_THREAD_LOCAL _Thread_local + #endif + + #ifndef STBI_THREAD_LOCAL + #if defined(__GNUC__) + #define STBI_THREAD_LOCAL __thread + #endif + #endif +#endif + +#ifdef _MSC_VER +typedef unsigned short stbi__uint16; +typedef signed short stbi__int16; +typedef unsigned int stbi__uint32; +typedef signed int stbi__int32; +#else +#include +typedef uint16_t stbi__uint16; +typedef int16_t stbi__int16; +typedef uint32_t stbi__uint32; +typedef int32_t stbi__int32; +#endif + +// should produce compiler error if size is wrong +typedef unsigned char validate_uint32[sizeof(stbi__uint32)==4 ? 1 : -1]; + +#ifdef _MSC_VER +#define STBI_NOTUSED(v) (void)(v) +#else +#define STBI_NOTUSED(v) (void)sizeof(v) +#endif + +#ifdef _MSC_VER +#define STBI_HAS_LROTL +#endif + +#ifdef STBI_HAS_LROTL + #define stbi_lrot(x,y) _lrotl(x,y) +#else + #define stbi_lrot(x,y) (((x) << (y)) | ((x) >> (-(y) & 31))) +#endif + +#if defined(STBI_MALLOC) && defined(STBI_FREE) && (defined(STBI_REALLOC) || defined(STBI_REALLOC_SIZED)) +// ok +#elif !defined(STBI_MALLOC) && !defined(STBI_FREE) && !defined(STBI_REALLOC) && !defined(STBI_REALLOC_SIZED) +// ok +#else +#error "Must define all or none of STBI_MALLOC, STBI_FREE, and STBI_REALLOC (or STBI_REALLOC_SIZED)." +#endif + +#ifndef STBI_MALLOC +#define STBI_MALLOC(sz) malloc(sz) +#define STBI_REALLOC(p,newsz) realloc(p,newsz) +#define STBI_FREE(p) if (p) free(p) +#endif + +#ifndef STBI_REALLOC_SIZED +#define STBI_REALLOC_SIZED(p,oldsz,newsz) STBI_REALLOC(p,newsz) +#endif + +// x86/x64 detection +#if defined(__x86_64__) || defined(_M_X64) +#define STBI__X64_TARGET +#elif defined(__i386) || defined(_M_IX86) +#define STBI__X86_TARGET +#endif + +#if defined(__GNUC__) && defined(STBI__X86_TARGET) && !defined(__SSE2__) && !defined(STBI_NO_SIMD) +// gcc doesn't support sse2 intrinsics unless you compile with -msse2, +// which in turn means it gets to use SSE2 everywhere. This is unfortunate, +// but previous attempts to provide the SSE2 functions with runtime +// detection caused numerous issues. The way architecture extensions are +// exposed in GCC/Clang is, sadly, not really suited for one-file libs. +// New behavior: if compiled with -msse2, we use SSE2 without any +// detection; if not, we don't use it at all. +#define STBI_NO_SIMD +#endif + +#if defined(__MINGW32__) && defined(STBI__X86_TARGET) && !defined(STBI_MINGW_ENABLE_SSE2) && !defined(STBI_NO_SIMD) +// Note that __MINGW32__ doesn't actually mean 32-bit, so we have to avoid STBI__X64_TARGET +// +// 32-bit MinGW wants ESP to be 16-byte aligned, but this is not in the +// Windows ABI and VC++ as well as Windows DLLs don't maintain that invariant. +// As a result, enabling SSE2 on 32-bit MinGW is dangerous when not +// simultaneously enabling "-mstackrealign". +// +// See https://github.com/nothings/stb/issues/81 for more information. +// +// So default to no SSE2 on 32-bit MinGW. If you've read this far and added +// -mstackrealign to your build settings, feel free to #define STBI_MINGW_ENABLE_SSE2. +#define STBI_NO_SIMD +#endif + +#if !defined(STBI_NO_SIMD) && (defined(STBI__X86_TARGET) || defined(STBI__X64_TARGET)) +#define STBI_SSE2 +#include + +#ifdef _MSC_VER + +#if _MSC_VER >= 1400 // not VC6 +#include // __cpuid +static int stbi__cpuid3(void) +{ + int info[4]; + __cpuid(info,1); + return info[3]; +} +#else +static int stbi__cpuid3(void) +{ + int res; + __asm { + mov eax,1 + cpuid + mov res,edx + } + return res; +} +#endif + +#define STBI_SIMD_ALIGN(type, name) __declspec(align(16)) type name + +#if !defined(STBI_NO_JPEG) && defined(STBI_SSE2) +static int stbi__sse2_available(void) +{ + int info3 = stbi__cpuid3(); + return ((info3 >> 26) & 1) != 0; +} +#endif + +#else // assume GCC-style if not VC++ +#define STBI_SIMD_ALIGN(type, name) type name __attribute__((aligned(16))) + +#if !defined(STBI_NO_JPEG) && defined(STBI_SSE2) +static int stbi__sse2_available(void) +{ + // If we're even attempting to compile this on GCC/Clang, that means + // -msse2 is on, which means the compiler is allowed to use SSE2 + // instructions at will, and so are we. + return 1; +} +#endif + +#endif +#endif + +// ARM NEON +#if defined(STBI_NO_SIMD) && defined(STBI_NEON) +#undef STBI_NEON +#endif + +#ifdef STBI_NEON +#include +#ifdef _MSC_VER +#define STBI_SIMD_ALIGN(type, name) __declspec(align(16)) type name +#else +#define STBI_SIMD_ALIGN(type, name) type name __attribute__((aligned(16))) +#endif +#endif + +#ifndef STBI_SIMD_ALIGN +#define STBI_SIMD_ALIGN(type, name) type name +#endif + +#ifndef STBI_MAX_DIMENSIONS +#define STBI_MAX_DIMENSIONS (1 << 24) +#endif + +/////////////////////////////////////////////// +// +// stbi__context struct and start_xxx functions + +// stbi__context structure is our basic context used by all images, so it +// contains all the IO context, plus some basic image information +typedef struct +{ + stbi__uint32 img_x, img_y; + int img_n, img_out_n; + + stbi_io_callbacks io; + void *io_user_data; + + int read_from_callbacks; + int buflen; + stbi_uc buffer_start[128]; + int callback_already_read; + + stbi_uc *img_buffer, *img_buffer_end; + stbi_uc *img_buffer_original, *img_buffer_original_end; +} stbi__context; + + +static void stbi__refill_buffer(stbi__context *s); + +// initialize a memory-decode context +static void stbi__start_mem(stbi__context *s, stbi_uc const *buffer, int len) +{ + s->io.read = NULL; + s->read_from_callbacks = 0; + s->callback_already_read = 0; + s->img_buffer = s->img_buffer_original = (stbi_uc *) buffer; + s->img_buffer_end = s->img_buffer_original_end = (stbi_uc *) buffer+len; +} + +// initialize a callback-based context +static void stbi__start_callbacks(stbi__context *s, stbi_io_callbacks *c, void *user) +{ + s->io = *c; + s->io_user_data = user; + s->buflen = sizeof(s->buffer_start); + s->read_from_callbacks = 1; + s->callback_already_read = 0; + s->img_buffer = s->img_buffer_original = s->buffer_start; + stbi__refill_buffer(s); + s->img_buffer_original_end = s->img_buffer_end; +} + +#ifndef STBI_NO_STDIO + +static int stbi__stdio_read(void *user, char *data, int size) +{ + return (int) fread(data,1,size,(FILE*) user); +} + +static void stbi__stdio_skip(void *user, int n) +{ + int ch; + fseek((FILE*) user, n, SEEK_CUR); + ch = fgetc((FILE*) user); /* have to read a byte to reset feof()'s flag */ + if (ch != EOF) { + ungetc(ch, (FILE *) user); /* push byte back onto stream if valid. */ + } +} + +static int stbi__stdio_eof(void *user) +{ + return feof((FILE*) user) || ferror((FILE *) user); +} + +static stbi_io_callbacks stbi__stdio_callbacks = +{ + stbi__stdio_read, + stbi__stdio_skip, + stbi__stdio_eof, +}; + +static void stbi__start_file(stbi__context *s, FILE *f) +{ + stbi__start_callbacks(s, &stbi__stdio_callbacks, (void *) f); +} + +//static void stop_file(stbi__context *s) { } + +#endif // !STBI_NO_STDIO + +static void stbi__rewind(stbi__context *s) +{ + // conceptually rewind SHOULD rewind to the beginning of the stream, + // but we just rewind to the beginning of the initial buffer, because + // we only use it after doing 'test', which only ever looks at at most 92 bytes + s->img_buffer = s->img_buffer_original; + s->img_buffer_end = s->img_buffer_original_end; +} + +enum +{ + STBI_ORDER_RGB, + STBI_ORDER_BGR +}; + +typedef struct +{ + int bits_per_channel; + int num_channels; + int channel_order; +} stbi__result_info; + +#ifndef STBI_NO_JPEG +static int stbi__jpeg_test(stbi__context *s); +static void *stbi__jpeg_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__jpeg_info(stbi__context *s, int *x, int *y, int *comp); +#endif + +#ifndef STBI_NO_PNG +static int stbi__png_test(stbi__context *s); +static void *stbi__png_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__png_info(stbi__context *s, int *x, int *y, int *comp); +static int stbi__png_is16(stbi__context *s); +#endif + +#ifndef STBI_NO_BMP +static int stbi__bmp_test(stbi__context *s); +static void *stbi__bmp_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__bmp_info(stbi__context *s, int *x, int *y, int *comp); +#endif + +#ifndef STBI_NO_TGA +static int stbi__tga_test(stbi__context *s); +static void *stbi__tga_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__tga_info(stbi__context *s, int *x, int *y, int *comp); +#endif + +#ifndef STBI_NO_PSD +static int stbi__psd_test(stbi__context *s); +static void *stbi__psd_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri, int bpc); +static int stbi__psd_info(stbi__context *s, int *x, int *y, int *comp); +static int stbi__psd_is16(stbi__context *s); +#endif + +#ifndef STBI_NO_HDR +static int stbi__hdr_test(stbi__context *s); +static float *stbi__hdr_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__hdr_info(stbi__context *s, int *x, int *y, int *comp); +#endif + +#ifndef STBI_NO_PIC +static int stbi__pic_test(stbi__context *s); +static void *stbi__pic_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__pic_info(stbi__context *s, int *x, int *y, int *comp); +#endif + +#ifndef STBI_NO_GIF +static int stbi__gif_test(stbi__context *s); +static void *stbi__gif_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static void *stbi__load_gif_main(stbi__context *s, int **delays, int *x, int *y, int *z, int *comp, int req_comp); +static int stbi__gif_info(stbi__context *s, int *x, int *y, int *comp); +#endif + +#ifndef STBI_NO_PNM +static int stbi__pnm_test(stbi__context *s); +static void *stbi__pnm_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__pnm_info(stbi__context *s, int *x, int *y, int *comp); +static int stbi__pnm_is16(stbi__context *s); +#endif + +static +#ifdef STBI_THREAD_LOCAL +STBI_THREAD_LOCAL +#endif +const char *stbi__g_failure_reason; + +STBIDEF const char *stbi_failure_reason(void) +{ + return stbi__g_failure_reason; +} + +#ifndef STBI_NO_FAILURE_STRINGS +static int stbi__err(const char *str) +{ + stbi__g_failure_reason = str; + return 0; +} +#endif + +static void *stbi__malloc(size_t size) +{ + return STBI_MALLOC(size); +} + +// stb_image uses ints pervasively, including for offset calculations. +// therefore the largest decoded image size we can support with the +// current code, even on 64-bit targets, is INT_MAX. this is not a +// significant limitation for the intended use case. +// +// we do, however, need to make sure our size calculations don't +// overflow. hence a few helper functions for size calculations that +// multiply integers together, making sure that they're non-negative +// and no overflow occurs. + +// return 1 if the sum is valid, 0 on overflow. +// negative terms are considered invalid. +static int stbi__addsizes_valid(int a, int b) +{ + if (b < 0) return 0; + // now 0 <= b <= INT_MAX, hence also + // 0 <= INT_MAX - b <= INTMAX. + // And "a + b <= INT_MAX" (which might overflow) is the + // same as a <= INT_MAX - b (no overflow) + return a <= INT_MAX - b; +} + +// returns 1 if the product is valid, 0 on overflow. +// negative factors are considered invalid. +static int stbi__mul2sizes_valid(int a, int b) +{ + if (a < 0 || b < 0) return 0; + if (b == 0) return 1; // mul-by-0 is always safe + // portable way to check for no overflows in a*b + return a <= INT_MAX/b; +} + +#if !defined(STBI_NO_JPEG) || !defined(STBI_NO_PNG) || !defined(STBI_NO_TGA) || !defined(STBI_NO_HDR) +// returns 1 if "a*b + add" has no negative terms/factors and doesn't overflow +static int stbi__mad2sizes_valid(int a, int b, int add) +{ + return stbi__mul2sizes_valid(a, b) && stbi__addsizes_valid(a*b, add); +} +#endif + +// returns 1 if "a*b*c + add" has no negative terms/factors and doesn't overflow +static int stbi__mad3sizes_valid(int a, int b, int c, int add) +{ + return stbi__mul2sizes_valid(a, b) && stbi__mul2sizes_valid(a*b, c) && + stbi__addsizes_valid(a*b*c, add); +} + +// returns 1 if "a*b*c*d + add" has no negative terms/factors and doesn't overflow +#if !defined(STBI_NO_LINEAR) || !defined(STBI_NO_HDR) || !defined(STBI_NO_PNM) +static int stbi__mad4sizes_valid(int a, int b, int c, int d, int add) +{ + return stbi__mul2sizes_valid(a, b) && stbi__mul2sizes_valid(a*b, c) && + stbi__mul2sizes_valid(a*b*c, d) && stbi__addsizes_valid(a*b*c*d, add); +} +#endif + +#if !defined(STBI_NO_JPEG) || !defined(STBI_NO_PNG) || !defined(STBI_NO_TGA) || !defined(STBI_NO_HDR) +// mallocs with size overflow checking +static void *stbi__malloc_mad2(int a, int b, int add) +{ + if (!stbi__mad2sizes_valid(a, b, add)) return NULL; + return stbi__malloc(a*b + add); +} +#endif + +static void *stbi__malloc_mad3(int a, int b, int c, int add) +{ + if (!stbi__mad3sizes_valid(a, b, c, add)) return NULL; + return stbi__malloc(a*b*c + add); +} + +#if !defined(STBI_NO_LINEAR) || !defined(STBI_NO_HDR) || !defined(STBI_NO_PNM) +static void *stbi__malloc_mad4(int a, int b, int c, int d, int add) +{ + if (!stbi__mad4sizes_valid(a, b, c, d, add)) return NULL; + return stbi__malloc(a*b*c*d + add); +} +#endif + +// stbi__err - error +// stbi__errpf - error returning pointer to float +// stbi__errpuc - error returning pointer to unsigned char + +#ifdef STBI_NO_FAILURE_STRINGS + #define stbi__err(x,y) 0 +#elif defined(STBI_FAILURE_USERMSG) + #define stbi__err(x,y) stbi__err(y) +#else + #define stbi__err(x,y) stbi__err(x) +#endif + +#define stbi__errpf(x,y) ((float *)(size_t) (stbi__err(x,y)?NULL:NULL)) +#define stbi__errpuc(x,y) ((unsigned char *)(size_t) (stbi__err(x,y)?NULL:NULL)) + +STBIDEF void stbi_image_free(void *retval_from_stbi_load) +{ + STBI_FREE(retval_from_stbi_load); +} + +#ifndef STBI_NO_LINEAR +static float *stbi__ldr_to_hdr(stbi_uc *data, int x, int y, int comp); +#endif + +#ifndef STBI_NO_HDR +static stbi_uc *stbi__hdr_to_ldr(float *data, int x, int y, int comp); +#endif + +static int stbi__vertically_flip_on_load_global = 0; + +STBIDEF void stbi_set_flip_vertically_on_load(int flag_true_if_should_flip) +{ + stbi__vertically_flip_on_load_global = flag_true_if_should_flip; +} + +#ifndef STBI_THREAD_LOCAL +#define stbi__vertically_flip_on_load stbi__vertically_flip_on_load_global +#else +static STBI_THREAD_LOCAL int stbi__vertically_flip_on_load_local, stbi__vertically_flip_on_load_set; + +STBIDEF void stbi_set_flip_vertically_on_load_thread(int flag_true_if_should_flip) +{ + stbi__vertically_flip_on_load_local = flag_true_if_should_flip; + stbi__vertically_flip_on_load_set = 1; +} + +#define stbi__vertically_flip_on_load (stbi__vertically_flip_on_load_set \ + ? stbi__vertically_flip_on_load_local \ + : stbi__vertically_flip_on_load_global) +#endif // STBI_THREAD_LOCAL + +static void *stbi__load_main(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri, int bpc) +{ + memset(ri, 0, sizeof(*ri)); // make sure it's initialized if we add new fields + ri->bits_per_channel = 8; // default is 8 so most paths don't have to be changed + ri->channel_order = STBI_ORDER_RGB; // all current input & output are this, but this is here so we can add BGR order + ri->num_channels = 0; + + // test the formats with a very explicit header first (at least a FOURCC + // or distinctive magic number first) + #ifndef STBI_NO_PNG + if (stbi__png_test(s)) return stbi__png_load(s,x,y,comp,req_comp, ri); + #endif + #ifndef STBI_NO_BMP + if (stbi__bmp_test(s)) return stbi__bmp_load(s,x,y,comp,req_comp, ri); + #endif + #ifndef STBI_NO_GIF + if (stbi__gif_test(s)) return stbi__gif_load(s,x,y,comp,req_comp, ri); + #endif + #ifndef STBI_NO_PSD + if (stbi__psd_test(s)) return stbi__psd_load(s,x,y,comp,req_comp, ri, bpc); + #else + STBI_NOTUSED(bpc); + #endif + #ifndef STBI_NO_PIC + if (stbi__pic_test(s)) return stbi__pic_load(s,x,y,comp,req_comp, ri); + #endif + + // then the formats that can end up attempting to load with just 1 or 2 + // bytes matching expectations; these are prone to false positives, so + // try them later + #ifndef STBI_NO_JPEG + if (stbi__jpeg_test(s)) return stbi__jpeg_load(s,x,y,comp,req_comp, ri); + #endif + #ifndef STBI_NO_PNM + if (stbi__pnm_test(s)) return stbi__pnm_load(s,x,y,comp,req_comp, ri); + #endif + + #ifndef STBI_NO_HDR + if (stbi__hdr_test(s)) { + float *hdr = stbi__hdr_load(s, x,y,comp,req_comp, ri); + return stbi__hdr_to_ldr(hdr, *x, *y, req_comp ? req_comp : *comp); + } + #endif + + #ifndef STBI_NO_TGA + // test tga last because it's a crappy test! + if (stbi__tga_test(s)) + return stbi__tga_load(s,x,y,comp,req_comp, ri); + #endif + + return stbi__errpuc("unknown image type", "Image not of any known type, or corrupt"); +} + +static stbi_uc *stbi__convert_16_to_8(stbi__uint16 *orig, int w, int h, int channels) +{ + int i; + int img_len = w * h * channels; + stbi_uc *reduced; + + reduced = (stbi_uc *) stbi__malloc(img_len); + if (reduced == NULL) return stbi__errpuc("outofmem", "Out of memory"); + + for (i = 0; i < img_len; ++i) + reduced[i] = (stbi_uc)((orig[i] >> 8) & 0xFF); // top half of each byte is sufficient approx of 16->8 bit scaling + + STBI_FREE(orig); + return reduced; +} + +static stbi__uint16 *stbi__convert_8_to_16(stbi_uc *orig, int w, int h, int channels) +{ + int i; + int img_len = w * h * channels; + stbi__uint16 *enlarged; + + enlarged = (stbi__uint16 *) stbi__malloc(img_len*2); + if (enlarged == NULL) return (stbi__uint16 *) stbi__errpuc("outofmem", "Out of memory"); + + for (i = 0; i < img_len; ++i) + enlarged[i] = (stbi__uint16)((orig[i] << 8) + orig[i]); // replicate to high and low byte, maps 0->0, 255->0xffff + + STBI_FREE(orig); + return enlarged; +} + +static void stbi__vertical_flip(void *image, int w, int h, int bytes_per_pixel) +{ + int row; + size_t bytes_per_row = (size_t)w * bytes_per_pixel; + stbi_uc temp[2048]; + stbi_uc *bytes = (stbi_uc *)image; + + for (row = 0; row < (h>>1); row++) { + stbi_uc *row0 = bytes + row*bytes_per_row; + stbi_uc *row1 = bytes + (h - row - 1)*bytes_per_row; + // swap row0 with row1 + size_t bytes_left = bytes_per_row; + while (bytes_left) { + size_t bytes_copy = (bytes_left < sizeof(temp)) ? bytes_left : sizeof(temp); + memcpy(temp, row0, bytes_copy); + memcpy(row0, row1, bytes_copy); + memcpy(row1, temp, bytes_copy); + row0 += bytes_copy; + row1 += bytes_copy; + bytes_left -= bytes_copy; + } + } +} + +#ifndef STBI_NO_GIF +static void stbi__vertical_flip_slices(void *image, int w, int h, int z, int bytes_per_pixel) +{ + int slice; + int slice_size = w * h * bytes_per_pixel; + + stbi_uc *bytes = (stbi_uc *)image; + for (slice = 0; slice < z; ++slice) { + stbi__vertical_flip(bytes, w, h, bytes_per_pixel); + bytes += slice_size; + } +} +#endif + +static unsigned char *stbi__load_and_postprocess_8bit(stbi__context *s, int *x, int *y, int *comp, int req_comp) +{ + stbi__result_info ri; + void *result = stbi__load_main(s, x, y, comp, req_comp, &ri, 8); + + if (result == NULL) + return NULL; + + // it is the responsibility of the loaders to make sure we get either 8 or 16 bit. + STBI_ASSERT(ri.bits_per_channel == 8 || ri.bits_per_channel == 16); + + if (ri.bits_per_channel != 8) { + result = stbi__convert_16_to_8((stbi__uint16 *) result, *x, *y, req_comp == 0 ? *comp : req_comp); + ri.bits_per_channel = 8; + } + + // @TODO: move stbi__convert_format to here + + if (stbi__vertically_flip_on_load) { + int channels = req_comp ? req_comp : *comp; + stbi__vertical_flip(result, *x, *y, channels * sizeof(stbi_uc)); + } + + return (unsigned char *) result; +} + +static stbi__uint16 *stbi__load_and_postprocess_16bit(stbi__context *s, int *x, int *y, int *comp, int req_comp) +{ + stbi__result_info ri; + void *result = stbi__load_main(s, x, y, comp, req_comp, &ri, 16); + + if (result == NULL) + return NULL; + + // it is the responsibility of the loaders to make sure we get either 8 or 16 bit. + STBI_ASSERT(ri.bits_per_channel == 8 || ri.bits_per_channel == 16); + + if (ri.bits_per_channel != 16) { + result = stbi__convert_8_to_16((stbi_uc *) result, *x, *y, req_comp == 0 ? *comp : req_comp); + ri.bits_per_channel = 16; + } + + // @TODO: move stbi__convert_format16 to here + // @TODO: special case RGB-to-Y (and RGBA-to-YA) for 8-bit-to-16-bit case to keep more precision + + if (stbi__vertically_flip_on_load) { + int channels = req_comp ? req_comp : *comp; + stbi__vertical_flip(result, *x, *y, channels * sizeof(stbi__uint16)); + } + + return (stbi__uint16 *) result; +} + +#if !defined(STBI_NO_HDR) && !defined(STBI_NO_LINEAR) +static void stbi__float_postprocess(float *result, int *x, int *y, int *comp, int req_comp) +{ + if (stbi__vertically_flip_on_load && result != NULL) { + int channels = req_comp ? req_comp : *comp; + stbi__vertical_flip(result, *x, *y, channels * sizeof(float)); + } +} +#endif + +#ifndef STBI_NO_STDIO + +#if defined(_WIN32) && defined(STBI_WINDOWS_UTF8) +STBI_EXTERN __declspec(dllimport) int __stdcall MultiByteToWideChar(unsigned int cp, unsigned long flags, const char *str, int cbmb, wchar_t *widestr, int cchwide); +STBI_EXTERN __declspec(dllimport) int __stdcall WideCharToMultiByte(unsigned int cp, unsigned long flags, const wchar_t *widestr, int cchwide, char *str, int cbmb, const char *defchar, int *used_default); +#endif + +#if defined(_WIN32) && defined(STBI_WINDOWS_UTF8) +STBIDEF int stbi_convert_wchar_to_utf8(char *buffer, size_t bufferlen, const wchar_t* input) +{ + return WideCharToMultiByte(65001 /* UTF8 */, 0, input, -1, buffer, (int) bufferlen, NULL, NULL); +} +#endif + +static FILE *stbi__fopen(char const *filename, char const *mode) +{ + FILE *f; +#if defined(_WIN32) && defined(STBI_WINDOWS_UTF8) + wchar_t wMode[64]; + wchar_t wFilename[1024]; + if (0 == MultiByteToWideChar(65001 /* UTF8 */, 0, filename, -1, wFilename, sizeof(wFilename)/sizeof(*wFilename))) + return 0; + + if (0 == MultiByteToWideChar(65001 /* UTF8 */, 0, mode, -1, wMode, sizeof(wMode)/sizeof(*wMode))) + return 0; + +#if defined(_MSC_VER) && _MSC_VER >= 1400 + if (0 != _wfopen_s(&f, wFilename, wMode)) + f = 0; +#else + f = _wfopen(wFilename, wMode); +#endif + +#elif defined(_MSC_VER) && _MSC_VER >= 1400 + if (0 != fopen_s(&f, filename, mode)) + f=0; +#else + f = fopen(filename, mode); +#endif + return f; +} + + +STBIDEF stbi_uc *stbi_load(char const *filename, int *x, int *y, int *comp, int req_comp) +{ + FILE *f = stbi__fopen(filename, "rb"); + unsigned char *result; + if (!f) return stbi__errpuc("can't fopen", "Unable to open file"); + result = stbi_load_from_file(f,x,y,comp,req_comp); + fclose(f); + return result; +} + +STBIDEF stbi_uc *stbi_load_from_file(FILE *f, int *x, int *y, int *comp, int req_comp) +{ + unsigned char *result; + stbi__context s; + stbi__start_file(&s,f); + result = stbi__load_and_postprocess_8bit(&s,x,y,comp,req_comp); + if (result) { + // need to 'unget' all the characters in the IO buffer + fseek(f, - (int) (s.img_buffer_end - s.img_buffer), SEEK_CUR); + } + return result; +} + +STBIDEF stbi__uint16 *stbi_load_from_file_16(FILE *f, int *x, int *y, int *comp, int req_comp) +{ + stbi__uint16 *result; + stbi__context s; + stbi__start_file(&s,f); + result = stbi__load_and_postprocess_16bit(&s,x,y,comp,req_comp); + if (result) { + // need to 'unget' all the characters in the IO buffer + fseek(f, - (int) (s.img_buffer_end - s.img_buffer), SEEK_CUR); + } + return result; +} + +STBIDEF stbi_us *stbi_load_16(char const *filename, int *x, int *y, int *comp, int req_comp) +{ + FILE *f = stbi__fopen(filename, "rb"); + stbi__uint16 *result; + if (!f) return (stbi_us *) stbi__errpuc("can't fopen", "Unable to open file"); + result = stbi_load_from_file_16(f,x,y,comp,req_comp); + fclose(f); + return result; +} + + +#endif //!STBI_NO_STDIO + +STBIDEF stbi_us *stbi_load_16_from_memory(stbi_uc const *buffer, int len, int *x, int *y, int *channels_in_file, int desired_channels) +{ + stbi__context s; + stbi__start_mem(&s,buffer,len); + return stbi__load_and_postprocess_16bit(&s,x,y,channels_in_file,desired_channels); +} + +STBIDEF stbi_us *stbi_load_16_from_callbacks(stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *channels_in_file, int desired_channels) +{ + stbi__context s; + stbi__start_callbacks(&s, (stbi_io_callbacks *)clbk, user); + return stbi__load_and_postprocess_16bit(&s,x,y,channels_in_file,desired_channels); +} + +STBIDEF stbi_uc *stbi_load_from_memory(stbi_uc const *buffer, int len, int *x, int *y, int *comp, int req_comp) +{ + stbi__context s; + stbi__start_mem(&s,buffer,len); + return stbi__load_and_postprocess_8bit(&s,x,y,comp,req_comp); +} + +STBIDEF stbi_uc *stbi_load_from_callbacks(stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *comp, int req_comp) +{ + stbi__context s; + stbi__start_callbacks(&s, (stbi_io_callbacks *) clbk, user); + return stbi__load_and_postprocess_8bit(&s,x,y,comp,req_comp); +} + +#ifndef STBI_NO_GIF +STBIDEF stbi_uc *stbi_load_gif_from_memory(stbi_uc const *buffer, int len, int **delays, int *x, int *y, int *z, int *comp, int req_comp) +{ + unsigned char *result; + stbi__context s; + stbi__start_mem(&s,buffer,len); + + result = (unsigned char*) stbi__load_gif_main(&s, delays, x, y, z, comp, req_comp); + if (stbi__vertically_flip_on_load) { + stbi__vertical_flip_slices( result, *x, *y, *z, *comp ); + } + + return result; +} +#endif + +#ifndef STBI_NO_LINEAR +static float *stbi__loadf_main(stbi__context *s, int *x, int *y, int *comp, int req_comp) +{ + unsigned char *data; + #ifndef STBI_NO_HDR + if (stbi__hdr_test(s)) { + stbi__result_info ri; + float *hdr_data = stbi__hdr_load(s,x,y,comp,req_comp, &ri); + if (hdr_data) + stbi__float_postprocess(hdr_data,x,y,comp,req_comp); + return hdr_data; + } + #endif + data = stbi__load_and_postprocess_8bit(s, x, y, comp, req_comp); + if (data) + return stbi__ldr_to_hdr(data, *x, *y, req_comp ? req_comp : *comp); + return stbi__errpf("unknown image type", "Image not of any known type, or corrupt"); +} + +STBIDEF float *stbi_loadf_from_memory(stbi_uc const *buffer, int len, int *x, int *y, int *comp, int req_comp) +{ + stbi__context s; + stbi__start_mem(&s,buffer,len); + return stbi__loadf_main(&s,x,y,comp,req_comp); +} + +STBIDEF float *stbi_loadf_from_callbacks(stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *comp, int req_comp) +{ + stbi__context s; + stbi__start_callbacks(&s, (stbi_io_callbacks *) clbk, user); + return stbi__loadf_main(&s,x,y,comp,req_comp); +} + +#ifndef STBI_NO_STDIO +STBIDEF float *stbi_loadf(char const *filename, int *x, int *y, int *comp, int req_comp) +{ + float *result; + FILE *f = stbi__fopen(filename, "rb"); + if (!f) return stbi__errpf("can't fopen", "Unable to open file"); + result = stbi_loadf_from_file(f,x,y,comp,req_comp); + fclose(f); + return result; +} + +STBIDEF float *stbi_loadf_from_file(FILE *f, int *x, int *y, int *comp, int req_comp) +{ + stbi__context s; + stbi__start_file(&s,f); + return stbi__loadf_main(&s,x,y,comp,req_comp); +} +#endif // !STBI_NO_STDIO + +#endif // !STBI_NO_LINEAR + +// these is-hdr-or-not is defined independent of whether STBI_NO_LINEAR is +// defined, for API simplicity; if STBI_NO_LINEAR is defined, it always +// reports false! + +STBIDEF int stbi_is_hdr_from_memory(stbi_uc const *buffer, int len) +{ + #ifndef STBI_NO_HDR + stbi__context s; + stbi__start_mem(&s,buffer,len); + return stbi__hdr_test(&s); + #else + STBI_NOTUSED(buffer); + STBI_NOTUSED(len); + return 0; + #endif +} + +#ifndef STBI_NO_STDIO +STBIDEF int stbi_is_hdr (char const *filename) +{ + FILE *f = stbi__fopen(filename, "rb"); + int result=0; + if (f) { + result = stbi_is_hdr_from_file(f); + fclose(f); + } + return result; +} + +STBIDEF int stbi_is_hdr_from_file(FILE *f) +{ + #ifndef STBI_NO_HDR + long pos = ftell(f); + int res; + stbi__context s; + stbi__start_file(&s,f); + res = stbi__hdr_test(&s); + fseek(f, pos, SEEK_SET); + return res; + #else + STBI_NOTUSED(f); + return 0; + #endif +} +#endif // !STBI_NO_STDIO + +STBIDEF int stbi_is_hdr_from_callbacks(stbi_io_callbacks const *clbk, void *user) +{ + #ifndef STBI_NO_HDR + stbi__context s; + stbi__start_callbacks(&s, (stbi_io_callbacks *) clbk, user); + return stbi__hdr_test(&s); + #else + STBI_NOTUSED(clbk); + STBI_NOTUSED(user); + return 0; + #endif +} + +#ifndef STBI_NO_LINEAR +static float stbi__l2h_gamma=2.2f, stbi__l2h_scale=1.0f; + +STBIDEF void stbi_ldr_to_hdr_gamma(float gamma) { stbi__l2h_gamma = gamma; } +STBIDEF void stbi_ldr_to_hdr_scale(float scale) { stbi__l2h_scale = scale; } +#endif + +static float stbi__h2l_gamma_i=1.0f/2.2f, stbi__h2l_scale_i=1.0f; + +STBIDEF void stbi_hdr_to_ldr_gamma(float gamma) { stbi__h2l_gamma_i = 1/gamma; } +STBIDEF void stbi_hdr_to_ldr_scale(float scale) { stbi__h2l_scale_i = 1/scale; } + + +////////////////////////////////////////////////////////////////////////////// +// +// Common code used by all image loaders +// + +enum +{ + STBI__SCAN_load=0, + STBI__SCAN_type, + STBI__SCAN_header +}; + +static void stbi__refill_buffer(stbi__context *s) +{ + int n = (s->io.read)(s->io_user_data,(char*)s->buffer_start,s->buflen); + s->callback_already_read += (int) (s->img_buffer - s->img_buffer_original); + if (n == 0) { + // at end of file, treat same as if from memory, but need to handle case + // where s->img_buffer isn't pointing to safe memory, e.g. 0-byte file + s->read_from_callbacks = 0; + s->img_buffer = s->buffer_start; + s->img_buffer_end = s->buffer_start+1; + *s->img_buffer = 0; + } else { + s->img_buffer = s->buffer_start; + s->img_buffer_end = s->buffer_start + n; + } +} + +stbi_inline static stbi_uc stbi__get8(stbi__context *s) +{ + if (s->img_buffer < s->img_buffer_end) + return *s->img_buffer++; + if (s->read_from_callbacks) { + stbi__refill_buffer(s); + return *s->img_buffer++; + } + return 0; +} + +#if defined(STBI_NO_JPEG) && defined(STBI_NO_HDR) && defined(STBI_NO_PIC) && defined(STBI_NO_PNM) +// nothing +#else +stbi_inline static int stbi__at_eof(stbi__context *s) +{ + if (s->io.read) { + if (!(s->io.eof)(s->io_user_data)) return 0; + // if feof() is true, check if buffer = end + // special case: we've only got the special 0 character at the end + if (s->read_from_callbacks == 0) return 1; + } + + return s->img_buffer >= s->img_buffer_end; +} +#endif + +#if defined(STBI_NO_JPEG) && defined(STBI_NO_PNG) && defined(STBI_NO_BMP) && defined(STBI_NO_PSD) && defined(STBI_NO_TGA) && defined(STBI_NO_GIF) && defined(STBI_NO_PIC) +// nothing +#else +static void stbi__skip(stbi__context *s, int n) +{ + if (n == 0) return; // already there! + if (n < 0) { + s->img_buffer = s->img_buffer_end; + return; + } + if (s->io.read) { + int blen = (int) (s->img_buffer_end - s->img_buffer); + if (blen < n) { + s->img_buffer = s->img_buffer_end; + (s->io.skip)(s->io_user_data, n - blen); + return; + } + } + s->img_buffer += n; +} +#endif + +#if defined(STBI_NO_PNG) && defined(STBI_NO_TGA) && defined(STBI_NO_HDR) && defined(STBI_NO_PNM) +// nothing +#else +static int stbi__getn(stbi__context *s, stbi_uc *buffer, int n) +{ + if (s->io.read) { + int blen = (int) (s->img_buffer_end - s->img_buffer); + if (blen < n) { + int res, count; + + memcpy(buffer, s->img_buffer, blen); + + count = (s->io.read)(s->io_user_data, (char*) buffer + blen, n - blen); + res = (count == (n-blen)); + s->img_buffer = s->img_buffer_end; + return res; + } + } + + if (s->img_buffer+n <= s->img_buffer_end) { + memcpy(buffer, s->img_buffer, n); + s->img_buffer += n; + return 1; + } else + return 0; +} +#endif + +#if defined(STBI_NO_JPEG) && defined(STBI_NO_PNG) && defined(STBI_NO_PSD) && defined(STBI_NO_PIC) +// nothing +#else +static int stbi__get16be(stbi__context *s) +{ + int z = stbi__get8(s); + return (z << 8) + stbi__get8(s); +} +#endif + +#if defined(STBI_NO_PNG) && defined(STBI_NO_PSD) && defined(STBI_NO_PIC) +// nothing +#else +static stbi__uint32 stbi__get32be(stbi__context *s) +{ + stbi__uint32 z = stbi__get16be(s); + return (z << 16) + stbi__get16be(s); +} +#endif + +#if defined(STBI_NO_BMP) && defined(STBI_NO_TGA) && defined(STBI_NO_GIF) +// nothing +#else +static int stbi__get16le(stbi__context *s) +{ + int z = stbi__get8(s); + return z + (stbi__get8(s) << 8); +} +#endif + +#ifndef STBI_NO_BMP +static stbi__uint32 stbi__get32le(stbi__context *s) +{ + stbi__uint32 z = stbi__get16le(s); + z += (stbi__uint32)stbi__get16le(s) << 16; + return z; +} +#endif + +#define STBI__BYTECAST(x) ((stbi_uc) ((x) & 255)) // truncate int to byte without warnings + +#if defined(STBI_NO_JPEG) && defined(STBI_NO_PNG) && defined(STBI_NO_BMP) && defined(STBI_NO_PSD) && defined(STBI_NO_TGA) && defined(STBI_NO_GIF) && defined(STBI_NO_PIC) && defined(STBI_NO_PNM) +// nothing +#else +////////////////////////////////////////////////////////////////////////////// +// +// generic converter from built-in img_n to req_comp +// individual types do this automatically as much as possible (e.g. jpeg +// does all cases internally since it needs to colorspace convert anyway, +// and it never has alpha, so very few cases ). png can automatically +// interleave an alpha=255 channel, but falls back to this for other cases +// +// assume data buffer is malloced, so malloc a new one and free that one +// only failure mode is malloc failing + +static stbi_uc stbi__compute_y(int r, int g, int b) +{ + return (stbi_uc) (((r*77) + (g*150) + (29*b)) >> 8); +} +#endif + +#if defined(STBI_NO_PNG) && defined(STBI_NO_BMP) && defined(STBI_NO_PSD) && defined(STBI_NO_TGA) && defined(STBI_NO_GIF) && defined(STBI_NO_PIC) && defined(STBI_NO_PNM) +// nothing +#else +static unsigned char *stbi__convert_format(unsigned char *data, int img_n, int req_comp, unsigned int x, unsigned int y) +{ + int i,j; + unsigned char *good; + + if (req_comp == img_n) return data; + STBI_ASSERT(req_comp >= 1 && req_comp <= 4); + + good = (unsigned char *) stbi__malloc_mad3(req_comp, x, y, 0); + if (good == NULL) { + STBI_FREE(data); + return stbi__errpuc("outofmem", "Out of memory"); + } + + for (j=0; j < (int) y; ++j) { + unsigned char *src = data + j * x * img_n ; + unsigned char *dest = good + j * x * req_comp; + + #define STBI__COMBO(a,b) ((a)*8+(b)) + #define STBI__CASE(a,b) case STBI__COMBO(a,b): for(i=x-1; i >= 0; --i, src += a, dest += b) + // convert source image with img_n components to one with req_comp components; + // avoid switch per pixel, so use switch per scanline and massive macros + switch (STBI__COMBO(img_n, req_comp)) { + STBI__CASE(1,2) { dest[0]=src[0]; dest[1]=255; } break; + STBI__CASE(1,3) { dest[0]=dest[1]=dest[2]=src[0]; } break; + STBI__CASE(1,4) { dest[0]=dest[1]=dest[2]=src[0]; dest[3]=255; } break; + STBI__CASE(2,1) { dest[0]=src[0]; } break; + STBI__CASE(2,3) { dest[0]=dest[1]=dest[2]=src[0]; } break; + STBI__CASE(2,4) { dest[0]=dest[1]=dest[2]=src[0]; dest[3]=src[1]; } break; + STBI__CASE(3,4) { dest[0]=src[0];dest[1]=src[1];dest[2]=src[2];dest[3]=255; } break; + STBI__CASE(3,1) { dest[0]=stbi__compute_y(src[0],src[1],src[2]); } break; + STBI__CASE(3,2) { dest[0]=stbi__compute_y(src[0],src[1],src[2]); dest[1] = 255; } break; + STBI__CASE(4,1) { dest[0]=stbi__compute_y(src[0],src[1],src[2]); } break; + STBI__CASE(4,2) { dest[0]=stbi__compute_y(src[0],src[1],src[2]); dest[1] = src[3]; } break; + STBI__CASE(4,3) { dest[0]=src[0];dest[1]=src[1];dest[2]=src[2]; } break; + default: STBI_ASSERT(0); STBI_FREE(data); STBI_FREE(good); return stbi__errpuc("unsupported", "Unsupported format conversion"); + } + #undef STBI__CASE + } + + STBI_FREE(data); + return good; +} +#endif + +#if defined(STBI_NO_PNG) && defined(STBI_NO_PSD) +// nothing +#else +static stbi__uint16 stbi__compute_y_16(int r, int g, int b) +{ + return (stbi__uint16) (((r*77) + (g*150) + (29*b)) >> 8); +} +#endif + +#if defined(STBI_NO_PNG) && defined(STBI_NO_PSD) +// nothing +#else +static stbi__uint16 *stbi__convert_format16(stbi__uint16 *data, int img_n, int req_comp, unsigned int x, unsigned int y) +{ + int i,j; + stbi__uint16 *good; + + if (req_comp == img_n) return data; + STBI_ASSERT(req_comp >= 1 && req_comp <= 4); + + good = (stbi__uint16 *) stbi__malloc(req_comp * x * y * 2); + if (good == NULL) { + STBI_FREE(data); + return (stbi__uint16 *) stbi__errpuc("outofmem", "Out of memory"); + } + + for (j=0; j < (int) y; ++j) { + stbi__uint16 *src = data + j * x * img_n ; + stbi__uint16 *dest = good + j * x * req_comp; + + #define STBI__COMBO(a,b) ((a)*8+(b)) + #define STBI__CASE(a,b) case STBI__COMBO(a,b): for(i=x-1; i >= 0; --i, src += a, dest += b) + // convert source image with img_n components to one with req_comp components; + // avoid switch per pixel, so use switch per scanline and massive macros + switch (STBI__COMBO(img_n, req_comp)) { + STBI__CASE(1,2) { dest[0]=src[0]; dest[1]=0xffff; } break; + STBI__CASE(1,3) { dest[0]=dest[1]=dest[2]=src[0]; } break; + STBI__CASE(1,4) { dest[0]=dest[1]=dest[2]=src[0]; dest[3]=0xffff; } break; + STBI__CASE(2,1) { dest[0]=src[0]; } break; + STBI__CASE(2,3) { dest[0]=dest[1]=dest[2]=src[0]; } break; + STBI__CASE(2,4) { dest[0]=dest[1]=dest[2]=src[0]; dest[3]=src[1]; } break; + STBI__CASE(3,4) { dest[0]=src[0];dest[1]=src[1];dest[2]=src[2];dest[3]=0xffff; } break; + STBI__CASE(3,1) { dest[0]=stbi__compute_y_16(src[0],src[1],src[2]); } break; + STBI__CASE(3,2) { dest[0]=stbi__compute_y_16(src[0],src[1],src[2]); dest[1] = 0xffff; } break; + STBI__CASE(4,1) { dest[0]=stbi__compute_y_16(src[0],src[1],src[2]); } break; + STBI__CASE(4,2) { dest[0]=stbi__compute_y_16(src[0],src[1],src[2]); dest[1] = src[3]; } break; + STBI__CASE(4,3) { dest[0]=src[0];dest[1]=src[1];dest[2]=src[2]; } break; + default: STBI_ASSERT(0); STBI_FREE(data); STBI_FREE(good); return (stbi__uint16*) stbi__errpuc("unsupported", "Unsupported format conversion"); + } + #undef STBI__CASE + } + + STBI_FREE(data); + return good; +} +#endif + +#ifndef STBI_NO_LINEAR +static float *stbi__ldr_to_hdr(stbi_uc *data, int x, int y, int comp) +{ + int i,k,n; + float *output; + if (!data) return NULL; + output = (float *) stbi__malloc_mad4(x, y, comp, sizeof(float), 0); + if (output == NULL) { STBI_FREE(data); return stbi__errpf("outofmem", "Out of memory"); } + // compute number of non-alpha components + if (comp & 1) n = comp; else n = comp-1; + for (i=0; i < x*y; ++i) { + for (k=0; k < n; ++k) { + output[i*comp + k] = (float) (pow(data[i*comp+k]/255.0f, stbi__l2h_gamma) * stbi__l2h_scale); + } + } + if (n < comp) { + for (i=0; i < x*y; ++i) { + output[i*comp + n] = data[i*comp + n]/255.0f; + } + } + STBI_FREE(data); + return output; +} +#endif + +#ifndef STBI_NO_HDR +#define stbi__float2int(x) ((int) (x)) +static stbi_uc *stbi__hdr_to_ldr(float *data, int x, int y, int comp) +{ + int i,k,n; + stbi_uc *output; + if (!data) return NULL; + output = (stbi_uc *) stbi__malloc_mad3(x, y, comp, 0); + if (output == NULL) { STBI_FREE(data); return stbi__errpuc("outofmem", "Out of memory"); } + // compute number of non-alpha components + if (comp & 1) n = comp; else n = comp-1; + for (i=0; i < x*y; ++i) { + for (k=0; k < n; ++k) { + float z = (float) pow(data[i*comp+k]*stbi__h2l_scale_i, stbi__h2l_gamma_i) * 255 + 0.5f; + if (z < 0) z = 0; + if (z > 255) z = 255; + output[i*comp + k] = (stbi_uc) stbi__float2int(z); + } + if (k < comp) { + float z = data[i*comp+k] * 255 + 0.5f; + if (z < 0) z = 0; + if (z > 255) z = 255; + output[i*comp + k] = (stbi_uc) stbi__float2int(z); + } + } + STBI_FREE(data); + return output; +} +#endif + +////////////////////////////////////////////////////////////////////////////// +// +// "baseline" JPEG/JFIF decoder +// +// simple implementation +// - doesn't support delayed output of y-dimension +// - simple interface (only one output format: 8-bit interleaved RGB) +// - doesn't try to recover corrupt jpegs +// - doesn't allow partial loading, loading multiple at once +// - still fast on x86 (copying globals into locals doesn't help x86) +// - allocates lots of intermediate memory (full size of all components) +// - non-interleaved case requires this anyway +// - allows good upsampling (see next) +// high-quality +// - upsampled channels are bilinearly interpolated, even across blocks +// - quality integer IDCT derived from IJG's 'slow' +// performance +// - fast huffman; reasonable integer IDCT +// - some SIMD kernels for common paths on targets with SSE2/NEON +// - uses a lot of intermediate memory, could cache poorly + +#ifndef STBI_NO_JPEG + +// huffman decoding acceleration +#define FAST_BITS 9 // larger handles more cases; smaller stomps less cache + +typedef struct +{ + stbi_uc fast[1 << FAST_BITS]; + // weirdly, repacking this into AoS is a 10% speed loss, instead of a win + stbi__uint16 code[256]; + stbi_uc values[256]; + stbi_uc size[257]; + unsigned int maxcode[18]; + int delta[17]; // old 'firstsymbol' - old 'firstcode' +} stbi__huffman; + +typedef struct +{ + stbi__context *s; + stbi__huffman huff_dc[4]; + stbi__huffman huff_ac[4]; + stbi__uint16 dequant[4][64]; + stbi__int16 fast_ac[4][1 << FAST_BITS]; + +// sizes for components, interleaved MCUs + int img_h_max, img_v_max; + int img_mcu_x, img_mcu_y; + int img_mcu_w, img_mcu_h; + +// definition of jpeg image component + struct + { + int id; + int h,v; + int tq; + int hd,ha; + int dc_pred; + + int x,y,w2,h2; + stbi_uc *data; + void *raw_data, *raw_coeff; + stbi_uc *linebuf; + short *coeff; // progressive only + int coeff_w, coeff_h; // number of 8x8 coefficient blocks + } img_comp[4]; + + stbi__uint32 code_buffer; // jpeg entropy-coded buffer + int code_bits; // number of valid bits + unsigned char marker; // marker seen while filling entropy buffer + int nomore; // flag if we saw a marker so must stop + + int progressive; + int spec_start; + int spec_end; + int succ_high; + int succ_low; + int eob_run; + int jfif; + int app14_color_transform; // Adobe APP14 tag + int rgb; + + int scan_n, order[4]; + int restart_interval, todo; + +// kernels + void (*idct_block_kernel)(stbi_uc *out, int out_stride, short data[64]); + void (*YCbCr_to_RGB_kernel)(stbi_uc *out, const stbi_uc *y, const stbi_uc *pcb, const stbi_uc *pcr, int count, int step); + stbi_uc *(*resample_row_hv_2_kernel)(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs); +} stbi__jpeg; + +static int stbi__build_huffman(stbi__huffman *h, int *count) +{ + int i,j,k=0; + unsigned int code; + // build size list for each symbol (from JPEG spec) + for (i=0; i < 16; ++i) + for (j=0; j < count[i]; ++j) + h->size[k++] = (stbi_uc) (i+1); + h->size[k] = 0; + + // compute actual symbols (from jpeg spec) + code = 0; + k = 0; + for(j=1; j <= 16; ++j) { + // compute delta to add to code to compute symbol id + h->delta[j] = k - code; + if (h->size[k] == j) { + while (h->size[k] == j) + h->code[k++] = (stbi__uint16) (code++); + if (code-1 >= (1u << j)) return stbi__err("bad code lengths","Corrupt JPEG"); + } + // compute largest code + 1 for this size, preshifted as needed later + h->maxcode[j] = code << (16-j); + code <<= 1; + } + h->maxcode[j] = 0xffffffff; + + // build non-spec acceleration table; 255 is flag for not-accelerated + memset(h->fast, 255, 1 << FAST_BITS); + for (i=0; i < k; ++i) { + int s = h->size[i]; + if (s <= FAST_BITS) { + int c = h->code[i] << (FAST_BITS-s); + int m = 1 << (FAST_BITS-s); + for (j=0; j < m; ++j) { + h->fast[c+j] = (stbi_uc) i; + } + } + } + return 1; +} + +// build a table that decodes both magnitude and value of small ACs in +// one go. +static void stbi__build_fast_ac(stbi__int16 *fast_ac, stbi__huffman *h) +{ + int i; + for (i=0; i < (1 << FAST_BITS); ++i) { + stbi_uc fast = h->fast[i]; + fast_ac[i] = 0; + if (fast < 255) { + int rs = h->values[fast]; + int run = (rs >> 4) & 15; + int magbits = rs & 15; + int len = h->size[fast]; + + if (magbits && len + magbits <= FAST_BITS) { + // magnitude code followed by receive_extend code + int k = ((i << len) & ((1 << FAST_BITS) - 1)) >> (FAST_BITS - magbits); + int m = 1 << (magbits - 1); + if (k < m) k += (~0U << magbits) + 1; + // if the result is small enough, we can fit it in fast_ac table + if (k >= -128 && k <= 127) + fast_ac[i] = (stbi__int16) ((k * 256) + (run * 16) + (len + magbits)); + } + } + } +} + +static void stbi__grow_buffer_unsafe(stbi__jpeg *j) +{ + do { + unsigned int b = j->nomore ? 0 : stbi__get8(j->s); + if (b == 0xff) { + int c = stbi__get8(j->s); + while (c == 0xff) c = stbi__get8(j->s); // consume fill bytes + if (c != 0) { + j->marker = (unsigned char) c; + j->nomore = 1; + return; + } + } + j->code_buffer |= b << (24 - j->code_bits); + j->code_bits += 8; + } while (j->code_bits <= 24); +} + +// (1 << n) - 1 +static const stbi__uint32 stbi__bmask[17]={0,1,3,7,15,31,63,127,255,511,1023,2047,4095,8191,16383,32767,65535}; + +// decode a jpeg huffman value from the bitstream +stbi_inline static int stbi__jpeg_huff_decode(stbi__jpeg *j, stbi__huffman *h) +{ + unsigned int temp; + int c,k; + + if (j->code_bits < 16) stbi__grow_buffer_unsafe(j); + + // look at the top FAST_BITS and determine what symbol ID it is, + // if the code is <= FAST_BITS + c = (j->code_buffer >> (32 - FAST_BITS)) & ((1 << FAST_BITS)-1); + k = h->fast[c]; + if (k < 255) { + int s = h->size[k]; + if (s > j->code_bits) + return -1; + j->code_buffer <<= s; + j->code_bits -= s; + return h->values[k]; + } + + // naive test is to shift the code_buffer down so k bits are + // valid, then test against maxcode. To speed this up, we've + // preshifted maxcode left so that it has (16-k) 0s at the + // end; in other words, regardless of the number of bits, it + // wants to be compared against something shifted to have 16; + // that way we don't need to shift inside the loop. + temp = j->code_buffer >> 16; + for (k=FAST_BITS+1 ; ; ++k) + if (temp < h->maxcode[k]) + break; + if (k == 17) { + // error! code not found + j->code_bits -= 16; + return -1; + } + + if (k > j->code_bits) + return -1; + + // convert the huffman code to the symbol id + c = ((j->code_buffer >> (32 - k)) & stbi__bmask[k]) + h->delta[k]; + STBI_ASSERT((((j->code_buffer) >> (32 - h->size[c])) & stbi__bmask[h->size[c]]) == h->code[c]); + + // convert the id to a symbol + j->code_bits -= k; + j->code_buffer <<= k; + return h->values[c]; +} + +// bias[n] = (-1<code_bits < n) stbi__grow_buffer_unsafe(j); + + sgn = j->code_buffer >> 31; // sign bit always in MSB; 0 if MSB clear (positive), 1 if MSB set (negative) + k = stbi_lrot(j->code_buffer, n); + j->code_buffer = k & ~stbi__bmask[n]; + k &= stbi__bmask[n]; + j->code_bits -= n; + return k + (stbi__jbias[n] & (sgn - 1)); +} + +// get some unsigned bits +stbi_inline static int stbi__jpeg_get_bits(stbi__jpeg *j, int n) +{ + unsigned int k; + if (j->code_bits < n) stbi__grow_buffer_unsafe(j); + k = stbi_lrot(j->code_buffer, n); + j->code_buffer = k & ~stbi__bmask[n]; + k &= stbi__bmask[n]; + j->code_bits -= n; + return k; +} + +stbi_inline static int stbi__jpeg_get_bit(stbi__jpeg *j) +{ + unsigned int k; + if (j->code_bits < 1) stbi__grow_buffer_unsafe(j); + k = j->code_buffer; + j->code_buffer <<= 1; + --j->code_bits; + return k & 0x80000000; +} + +// given a value that's at position X in the zigzag stream, +// where does it appear in the 8x8 matrix coded as row-major? +static const stbi_uc stbi__jpeg_dezigzag[64+15] = +{ + 0, 1, 8, 16, 9, 2, 3, 10, + 17, 24, 32, 25, 18, 11, 4, 5, + 12, 19, 26, 33, 40, 48, 41, 34, + 27, 20, 13, 6, 7, 14, 21, 28, + 35, 42, 49, 56, 57, 50, 43, 36, + 29, 22, 15, 23, 30, 37, 44, 51, + 58, 59, 52, 45, 38, 31, 39, 46, + 53, 60, 61, 54, 47, 55, 62, 63, + // let corrupt input sample past end + 63, 63, 63, 63, 63, 63, 63, 63, + 63, 63, 63, 63, 63, 63, 63 +}; + +// decode one 64-entry block-- +static int stbi__jpeg_decode_block(stbi__jpeg *j, short data[64], stbi__huffman *hdc, stbi__huffman *hac, stbi__int16 *fac, int b, stbi__uint16 *dequant) +{ + int diff,dc,k; + int t; + + if (j->code_bits < 16) stbi__grow_buffer_unsafe(j); + t = stbi__jpeg_huff_decode(j, hdc); + if (t < 0 || t > 15) return stbi__err("bad huffman code","Corrupt JPEG"); + + // 0 all the ac values now so we can do it 32-bits at a time + memset(data,0,64*sizeof(data[0])); + + diff = t ? stbi__extend_receive(j, t) : 0; + dc = j->img_comp[b].dc_pred + diff; + j->img_comp[b].dc_pred = dc; + data[0] = (short) (dc * dequant[0]); + + // decode AC components, see JPEG spec + k = 1; + do { + unsigned int zig; + int c,r,s; + if (j->code_bits < 16) stbi__grow_buffer_unsafe(j); + c = (j->code_buffer >> (32 - FAST_BITS)) & ((1 << FAST_BITS)-1); + r = fac[c]; + if (r) { // fast-AC path + k += (r >> 4) & 15; // run + s = r & 15; // combined length + j->code_buffer <<= s; + j->code_bits -= s; + // decode into unzigzag'd location + zig = stbi__jpeg_dezigzag[k++]; + data[zig] = (short) ((r >> 8) * dequant[zig]); + } else { + int rs = stbi__jpeg_huff_decode(j, hac); + if (rs < 0) return stbi__err("bad huffman code","Corrupt JPEG"); + s = rs & 15; + r = rs >> 4; + if (s == 0) { + if (rs != 0xf0) break; // end block + k += 16; + } else { + k += r; + // decode into unzigzag'd location + zig = stbi__jpeg_dezigzag[k++]; + data[zig] = (short) (stbi__extend_receive(j,s) * dequant[zig]); + } + } + } while (k < 64); + return 1; +} + +static int stbi__jpeg_decode_block_prog_dc(stbi__jpeg *j, short data[64], stbi__huffman *hdc, int b) +{ + int diff,dc; + int t; + if (j->spec_end != 0) return stbi__err("can't merge dc and ac", "Corrupt JPEG"); + + if (j->code_bits < 16) stbi__grow_buffer_unsafe(j); + + if (j->succ_high == 0) { + // first scan for DC coefficient, must be first + memset(data,0,64*sizeof(data[0])); // 0 all the ac values now + t = stbi__jpeg_huff_decode(j, hdc); + if (t < 0 || t > 15) return stbi__err("can't merge dc and ac", "Corrupt JPEG"); + diff = t ? stbi__extend_receive(j, t) : 0; + + dc = j->img_comp[b].dc_pred + diff; + j->img_comp[b].dc_pred = dc; + data[0] = (short) (dc * (1 << j->succ_low)); + } else { + // refinement scan for DC coefficient + if (stbi__jpeg_get_bit(j)) + data[0] += (short) (1 << j->succ_low); + } + return 1; +} + +// @OPTIMIZE: store non-zigzagged during the decode passes, +// and only de-zigzag when dequantizing +static int stbi__jpeg_decode_block_prog_ac(stbi__jpeg *j, short data[64], stbi__huffman *hac, stbi__int16 *fac) +{ + int k; + if (j->spec_start == 0) return stbi__err("can't merge dc and ac", "Corrupt JPEG"); + + if (j->succ_high == 0) { + int shift = j->succ_low; + + if (j->eob_run) { + --j->eob_run; + return 1; + } + + k = j->spec_start; + do { + unsigned int zig; + int c,r,s; + if (j->code_bits < 16) stbi__grow_buffer_unsafe(j); + c = (j->code_buffer >> (32 - FAST_BITS)) & ((1 << FAST_BITS)-1); + r = fac[c]; + if (r) { // fast-AC path + k += (r >> 4) & 15; // run + s = r & 15; // combined length + j->code_buffer <<= s; + j->code_bits -= s; + zig = stbi__jpeg_dezigzag[k++]; + data[zig] = (short) ((r >> 8) * (1 << shift)); + } else { + int rs = stbi__jpeg_huff_decode(j, hac); + if (rs < 0) return stbi__err("bad huffman code","Corrupt JPEG"); + s = rs & 15; + r = rs >> 4; + if (s == 0) { + if (r < 15) { + j->eob_run = (1 << r); + if (r) + j->eob_run += stbi__jpeg_get_bits(j, r); + --j->eob_run; + break; + } + k += 16; + } else { + k += r; + zig = stbi__jpeg_dezigzag[k++]; + data[zig] = (short) (stbi__extend_receive(j,s) * (1 << shift)); + } + } + } while (k <= j->spec_end); + } else { + // refinement scan for these AC coefficients + + short bit = (short) (1 << j->succ_low); + + if (j->eob_run) { + --j->eob_run; + for (k = j->spec_start; k <= j->spec_end; ++k) { + short *p = &data[stbi__jpeg_dezigzag[k]]; + if (*p != 0) + if (stbi__jpeg_get_bit(j)) + if ((*p & bit)==0) { + if (*p > 0) + *p += bit; + else + *p -= bit; + } + } + } else { + k = j->spec_start; + do { + int r,s; + int rs = stbi__jpeg_huff_decode(j, hac); // @OPTIMIZE see if we can use the fast path here, advance-by-r is so slow, eh + if (rs < 0) return stbi__err("bad huffman code","Corrupt JPEG"); + s = rs & 15; + r = rs >> 4; + if (s == 0) { + if (r < 15) { + j->eob_run = (1 << r) - 1; + if (r) + j->eob_run += stbi__jpeg_get_bits(j, r); + r = 64; // force end of block + } else { + // r=15 s=0 should write 16 0s, so we just do + // a run of 15 0s and then write s (which is 0), + // so we don't have to do anything special here + } + } else { + if (s != 1) return stbi__err("bad huffman code", "Corrupt JPEG"); + // sign bit + if (stbi__jpeg_get_bit(j)) + s = bit; + else + s = -bit; + } + + // advance by r + while (k <= j->spec_end) { + short *p = &data[stbi__jpeg_dezigzag[k++]]; + if (*p != 0) { + if (stbi__jpeg_get_bit(j)) + if ((*p & bit)==0) { + if (*p > 0) + *p += bit; + else + *p -= bit; + } + } else { + if (r == 0) { + *p = (short) s; + break; + } + --r; + } + } + } while (k <= j->spec_end); + } + } + return 1; +} + +// take a -128..127 value and stbi__clamp it and convert to 0..255 +stbi_inline static stbi_uc stbi__clamp(int x) +{ + // trick to use a single test to catch both cases + if ((unsigned int) x > 255) { + if (x < 0) return 0; + if (x > 255) return 255; + } + return (stbi_uc) x; +} + +#define stbi__f2f(x) ((int) (((x) * 4096 + 0.5))) +#define stbi__fsh(x) ((x) * 4096) + +// derived from jidctint -- DCT_ISLOW +#define STBI__IDCT_1D(s0,s1,s2,s3,s4,s5,s6,s7) \ + int t0,t1,t2,t3,p1,p2,p3,p4,p5,x0,x1,x2,x3; \ + p2 = s2; \ + p3 = s6; \ + p1 = (p2+p3) * stbi__f2f(0.5411961f); \ + t2 = p1 + p3*stbi__f2f(-1.847759065f); \ + t3 = p1 + p2*stbi__f2f( 0.765366865f); \ + p2 = s0; \ + p3 = s4; \ + t0 = stbi__fsh(p2+p3); \ + t1 = stbi__fsh(p2-p3); \ + x0 = t0+t3; \ + x3 = t0-t3; \ + x1 = t1+t2; \ + x2 = t1-t2; \ + t0 = s7; \ + t1 = s5; \ + t2 = s3; \ + t3 = s1; \ + p3 = t0+t2; \ + p4 = t1+t3; \ + p1 = t0+t3; \ + p2 = t1+t2; \ + p5 = (p3+p4)*stbi__f2f( 1.175875602f); \ + t0 = t0*stbi__f2f( 0.298631336f); \ + t1 = t1*stbi__f2f( 2.053119869f); \ + t2 = t2*stbi__f2f( 3.072711026f); \ + t3 = t3*stbi__f2f( 1.501321110f); \ + p1 = p5 + p1*stbi__f2f(-0.899976223f); \ + p2 = p5 + p2*stbi__f2f(-2.562915447f); \ + p3 = p3*stbi__f2f(-1.961570560f); \ + p4 = p4*stbi__f2f(-0.390180644f); \ + t3 += p1+p4; \ + t2 += p2+p3; \ + t1 += p2+p4; \ + t0 += p1+p3; + +static void stbi__idct_block(stbi_uc *out, int out_stride, short data[64]) +{ + int i,val[64],*v=val; + stbi_uc *o; + short *d = data; + + // columns + for (i=0; i < 8; ++i,++d, ++v) { + // if all zeroes, shortcut -- this avoids dequantizing 0s and IDCTing + if (d[ 8]==0 && d[16]==0 && d[24]==0 && d[32]==0 + && d[40]==0 && d[48]==0 && d[56]==0) { + // no shortcut 0 seconds + // (1|2|3|4|5|6|7)==0 0 seconds + // all separate -0.047 seconds + // 1 && 2|3 && 4|5 && 6|7: -0.047 seconds + int dcterm = d[0]*4; + v[0] = v[8] = v[16] = v[24] = v[32] = v[40] = v[48] = v[56] = dcterm; + } else { + STBI__IDCT_1D(d[ 0],d[ 8],d[16],d[24],d[32],d[40],d[48],d[56]) + // constants scaled things up by 1<<12; let's bring them back + // down, but keep 2 extra bits of precision + x0 += 512; x1 += 512; x2 += 512; x3 += 512; + v[ 0] = (x0+t3) >> 10; + v[56] = (x0-t3) >> 10; + v[ 8] = (x1+t2) >> 10; + v[48] = (x1-t2) >> 10; + v[16] = (x2+t1) >> 10; + v[40] = (x2-t1) >> 10; + v[24] = (x3+t0) >> 10; + v[32] = (x3-t0) >> 10; + } + } + + for (i=0, v=val, o=out; i < 8; ++i,v+=8,o+=out_stride) { + // no fast case since the first 1D IDCT spread components out + STBI__IDCT_1D(v[0],v[1],v[2],v[3],v[4],v[5],v[6],v[7]) + // constants scaled things up by 1<<12, plus we had 1<<2 from first + // loop, plus horizontal and vertical each scale by sqrt(8) so together + // we've got an extra 1<<3, so 1<<17 total we need to remove. + // so we want to round that, which means adding 0.5 * 1<<17, + // aka 65536. Also, we'll end up with -128 to 127 that we want + // to encode as 0..255 by adding 128, so we'll add that before the shift + x0 += 65536 + (128<<17); + x1 += 65536 + (128<<17); + x2 += 65536 + (128<<17); + x3 += 65536 + (128<<17); + // tried computing the shifts into temps, or'ing the temps to see + // if any were out of range, but that was slower + o[0] = stbi__clamp((x0+t3) >> 17); + o[7] = stbi__clamp((x0-t3) >> 17); + o[1] = stbi__clamp((x1+t2) >> 17); + o[6] = stbi__clamp((x1-t2) >> 17); + o[2] = stbi__clamp((x2+t1) >> 17); + o[5] = stbi__clamp((x2-t1) >> 17); + o[3] = stbi__clamp((x3+t0) >> 17); + o[4] = stbi__clamp((x3-t0) >> 17); + } +} + +#ifdef STBI_SSE2 +// sse2 integer IDCT. not the fastest possible implementation but it +// produces bit-identical results to the generic C version so it's +// fully "transparent". +static void stbi__idct_simd(stbi_uc *out, int out_stride, short data[64]) +{ + // This is constructed to match our regular (generic) integer IDCT exactly. + __m128i row0, row1, row2, row3, row4, row5, row6, row7; + __m128i tmp; + + // dot product constant: even elems=x, odd elems=y + #define dct_const(x,y) _mm_setr_epi16((x),(y),(x),(y),(x),(y),(x),(y)) + + // out(0) = c0[even]*x + c0[odd]*y (c0, x, y 16-bit, out 32-bit) + // out(1) = c1[even]*x + c1[odd]*y + #define dct_rot(out0,out1, x,y,c0,c1) \ + __m128i c0##lo = _mm_unpacklo_epi16((x),(y)); \ + __m128i c0##hi = _mm_unpackhi_epi16((x),(y)); \ + __m128i out0##_l = _mm_madd_epi16(c0##lo, c0); \ + __m128i out0##_h = _mm_madd_epi16(c0##hi, c0); \ + __m128i out1##_l = _mm_madd_epi16(c0##lo, c1); \ + __m128i out1##_h = _mm_madd_epi16(c0##hi, c1) + + // out = in << 12 (in 16-bit, out 32-bit) + #define dct_widen(out, in) \ + __m128i out##_l = _mm_srai_epi32(_mm_unpacklo_epi16(_mm_setzero_si128(), (in)), 4); \ + __m128i out##_h = _mm_srai_epi32(_mm_unpackhi_epi16(_mm_setzero_si128(), (in)), 4) + + // wide add + #define dct_wadd(out, a, b) \ + __m128i out##_l = _mm_add_epi32(a##_l, b##_l); \ + __m128i out##_h = _mm_add_epi32(a##_h, b##_h) + + // wide sub + #define dct_wsub(out, a, b) \ + __m128i out##_l = _mm_sub_epi32(a##_l, b##_l); \ + __m128i out##_h = _mm_sub_epi32(a##_h, b##_h) + + // butterfly a/b, add bias, then shift by "s" and pack + #define dct_bfly32o(out0, out1, a,b,bias,s) \ + { \ + __m128i abiased_l = _mm_add_epi32(a##_l, bias); \ + __m128i abiased_h = _mm_add_epi32(a##_h, bias); \ + dct_wadd(sum, abiased, b); \ + dct_wsub(dif, abiased, b); \ + out0 = _mm_packs_epi32(_mm_srai_epi32(sum_l, s), _mm_srai_epi32(sum_h, s)); \ + out1 = _mm_packs_epi32(_mm_srai_epi32(dif_l, s), _mm_srai_epi32(dif_h, s)); \ + } + + // 8-bit interleave step (for transposes) + #define dct_interleave8(a, b) \ + tmp = a; \ + a = _mm_unpacklo_epi8(a, b); \ + b = _mm_unpackhi_epi8(tmp, b) + + // 16-bit interleave step (for transposes) + #define dct_interleave16(a, b) \ + tmp = a; \ + a = _mm_unpacklo_epi16(a, b); \ + b = _mm_unpackhi_epi16(tmp, b) + + #define dct_pass(bias,shift) \ + { \ + /* even part */ \ + dct_rot(t2e,t3e, row2,row6, rot0_0,rot0_1); \ + __m128i sum04 = _mm_add_epi16(row0, row4); \ + __m128i dif04 = _mm_sub_epi16(row0, row4); \ + dct_widen(t0e, sum04); \ + dct_widen(t1e, dif04); \ + dct_wadd(x0, t0e, t3e); \ + dct_wsub(x3, t0e, t3e); \ + dct_wadd(x1, t1e, t2e); \ + dct_wsub(x2, t1e, t2e); \ + /* odd part */ \ + dct_rot(y0o,y2o, row7,row3, rot2_0,rot2_1); \ + dct_rot(y1o,y3o, row5,row1, rot3_0,rot3_1); \ + __m128i sum17 = _mm_add_epi16(row1, row7); \ + __m128i sum35 = _mm_add_epi16(row3, row5); \ + dct_rot(y4o,y5o, sum17,sum35, rot1_0,rot1_1); \ + dct_wadd(x4, y0o, y4o); \ + dct_wadd(x5, y1o, y5o); \ + dct_wadd(x6, y2o, y5o); \ + dct_wadd(x7, y3o, y4o); \ + dct_bfly32o(row0,row7, x0,x7,bias,shift); \ + dct_bfly32o(row1,row6, x1,x6,bias,shift); \ + dct_bfly32o(row2,row5, x2,x5,bias,shift); \ + dct_bfly32o(row3,row4, x3,x4,bias,shift); \ + } + + __m128i rot0_0 = dct_const(stbi__f2f(0.5411961f), stbi__f2f(0.5411961f) + stbi__f2f(-1.847759065f)); + __m128i rot0_1 = dct_const(stbi__f2f(0.5411961f) + stbi__f2f( 0.765366865f), stbi__f2f(0.5411961f)); + __m128i rot1_0 = dct_const(stbi__f2f(1.175875602f) + stbi__f2f(-0.899976223f), stbi__f2f(1.175875602f)); + __m128i rot1_1 = dct_const(stbi__f2f(1.175875602f), stbi__f2f(1.175875602f) + stbi__f2f(-2.562915447f)); + __m128i rot2_0 = dct_const(stbi__f2f(-1.961570560f) + stbi__f2f( 0.298631336f), stbi__f2f(-1.961570560f)); + __m128i rot2_1 = dct_const(stbi__f2f(-1.961570560f), stbi__f2f(-1.961570560f) + stbi__f2f( 3.072711026f)); + __m128i rot3_0 = dct_const(stbi__f2f(-0.390180644f) + stbi__f2f( 2.053119869f), stbi__f2f(-0.390180644f)); + __m128i rot3_1 = dct_const(stbi__f2f(-0.390180644f), stbi__f2f(-0.390180644f) + stbi__f2f( 1.501321110f)); + + // rounding biases in column/row passes, see stbi__idct_block for explanation. + __m128i bias_0 = _mm_set1_epi32(512); + __m128i bias_1 = _mm_set1_epi32(65536 + (128<<17)); + + // load + row0 = _mm_load_si128((const __m128i *) (data + 0*8)); + row1 = _mm_load_si128((const __m128i *) (data + 1*8)); + row2 = _mm_load_si128((const __m128i *) (data + 2*8)); + row3 = _mm_load_si128((const __m128i *) (data + 3*8)); + row4 = _mm_load_si128((const __m128i *) (data + 4*8)); + row5 = _mm_load_si128((const __m128i *) (data + 5*8)); + row6 = _mm_load_si128((const __m128i *) (data + 6*8)); + row7 = _mm_load_si128((const __m128i *) (data + 7*8)); + + // column pass + dct_pass(bias_0, 10); + + { + // 16bit 8x8 transpose pass 1 + dct_interleave16(row0, row4); + dct_interleave16(row1, row5); + dct_interleave16(row2, row6); + dct_interleave16(row3, row7); + + // transpose pass 2 + dct_interleave16(row0, row2); + dct_interleave16(row1, row3); + dct_interleave16(row4, row6); + dct_interleave16(row5, row7); + + // transpose pass 3 + dct_interleave16(row0, row1); + dct_interleave16(row2, row3); + dct_interleave16(row4, row5); + dct_interleave16(row6, row7); + } + + // row pass + dct_pass(bias_1, 17); + + { + // pack + __m128i p0 = _mm_packus_epi16(row0, row1); // a0a1a2a3...a7b0b1b2b3...b7 + __m128i p1 = _mm_packus_epi16(row2, row3); + __m128i p2 = _mm_packus_epi16(row4, row5); + __m128i p3 = _mm_packus_epi16(row6, row7); + + // 8bit 8x8 transpose pass 1 + dct_interleave8(p0, p2); // a0e0a1e1... + dct_interleave8(p1, p3); // c0g0c1g1... + + // transpose pass 2 + dct_interleave8(p0, p1); // a0c0e0g0... + dct_interleave8(p2, p3); // b0d0f0h0... + + // transpose pass 3 + dct_interleave8(p0, p2); // a0b0c0d0... + dct_interleave8(p1, p3); // a4b4c4d4... + + // store + _mm_storel_epi64((__m128i *) out, p0); out += out_stride; + _mm_storel_epi64((__m128i *) out, _mm_shuffle_epi32(p0, 0x4e)); out += out_stride; + _mm_storel_epi64((__m128i *) out, p2); out += out_stride; + _mm_storel_epi64((__m128i *) out, _mm_shuffle_epi32(p2, 0x4e)); out += out_stride; + _mm_storel_epi64((__m128i *) out, p1); out += out_stride; + _mm_storel_epi64((__m128i *) out, _mm_shuffle_epi32(p1, 0x4e)); out += out_stride; + _mm_storel_epi64((__m128i *) out, p3); out += out_stride; + _mm_storel_epi64((__m128i *) out, _mm_shuffle_epi32(p3, 0x4e)); + } + +#undef dct_const +#undef dct_rot +#undef dct_widen +#undef dct_wadd +#undef dct_wsub +#undef dct_bfly32o +#undef dct_interleave8 +#undef dct_interleave16 +#undef dct_pass +} + +#endif // STBI_SSE2 + +#ifdef STBI_NEON + +// NEON integer IDCT. should produce bit-identical +// results to the generic C version. +static void stbi__idct_simd(stbi_uc *out, int out_stride, short data[64]) +{ + int16x8_t row0, row1, row2, row3, row4, row5, row6, row7; + + int16x4_t rot0_0 = vdup_n_s16(stbi__f2f(0.5411961f)); + int16x4_t rot0_1 = vdup_n_s16(stbi__f2f(-1.847759065f)); + int16x4_t rot0_2 = vdup_n_s16(stbi__f2f( 0.765366865f)); + int16x4_t rot1_0 = vdup_n_s16(stbi__f2f( 1.175875602f)); + int16x4_t rot1_1 = vdup_n_s16(stbi__f2f(-0.899976223f)); + int16x4_t rot1_2 = vdup_n_s16(stbi__f2f(-2.562915447f)); + int16x4_t rot2_0 = vdup_n_s16(stbi__f2f(-1.961570560f)); + int16x4_t rot2_1 = vdup_n_s16(stbi__f2f(-0.390180644f)); + int16x4_t rot3_0 = vdup_n_s16(stbi__f2f( 0.298631336f)); + int16x4_t rot3_1 = vdup_n_s16(stbi__f2f( 2.053119869f)); + int16x4_t rot3_2 = vdup_n_s16(stbi__f2f( 3.072711026f)); + int16x4_t rot3_3 = vdup_n_s16(stbi__f2f( 1.501321110f)); + +#define dct_long_mul(out, inq, coeff) \ + int32x4_t out##_l = vmull_s16(vget_low_s16(inq), coeff); \ + int32x4_t out##_h = vmull_s16(vget_high_s16(inq), coeff) + +#define dct_long_mac(out, acc, inq, coeff) \ + int32x4_t out##_l = vmlal_s16(acc##_l, vget_low_s16(inq), coeff); \ + int32x4_t out##_h = vmlal_s16(acc##_h, vget_high_s16(inq), coeff) + +#define dct_widen(out, inq) \ + int32x4_t out##_l = vshll_n_s16(vget_low_s16(inq), 12); \ + int32x4_t out##_h = vshll_n_s16(vget_high_s16(inq), 12) + +// wide add +#define dct_wadd(out, a, b) \ + int32x4_t out##_l = vaddq_s32(a##_l, b##_l); \ + int32x4_t out##_h = vaddq_s32(a##_h, b##_h) + +// wide sub +#define dct_wsub(out, a, b) \ + int32x4_t out##_l = vsubq_s32(a##_l, b##_l); \ + int32x4_t out##_h = vsubq_s32(a##_h, b##_h) + +// butterfly a/b, then shift using "shiftop" by "s" and pack +#define dct_bfly32o(out0,out1, a,b,shiftop,s) \ + { \ + dct_wadd(sum, a, b); \ + dct_wsub(dif, a, b); \ + out0 = vcombine_s16(shiftop(sum_l, s), shiftop(sum_h, s)); \ + out1 = vcombine_s16(shiftop(dif_l, s), shiftop(dif_h, s)); \ + } + +#define dct_pass(shiftop, shift) \ + { \ + /* even part */ \ + int16x8_t sum26 = vaddq_s16(row2, row6); \ + dct_long_mul(p1e, sum26, rot0_0); \ + dct_long_mac(t2e, p1e, row6, rot0_1); \ + dct_long_mac(t3e, p1e, row2, rot0_2); \ + int16x8_t sum04 = vaddq_s16(row0, row4); \ + int16x8_t dif04 = vsubq_s16(row0, row4); \ + dct_widen(t0e, sum04); \ + dct_widen(t1e, dif04); \ + dct_wadd(x0, t0e, t3e); \ + dct_wsub(x3, t0e, t3e); \ + dct_wadd(x1, t1e, t2e); \ + dct_wsub(x2, t1e, t2e); \ + /* odd part */ \ + int16x8_t sum15 = vaddq_s16(row1, row5); \ + int16x8_t sum17 = vaddq_s16(row1, row7); \ + int16x8_t sum35 = vaddq_s16(row3, row5); \ + int16x8_t sum37 = vaddq_s16(row3, row7); \ + int16x8_t sumodd = vaddq_s16(sum17, sum35); \ + dct_long_mul(p5o, sumodd, rot1_0); \ + dct_long_mac(p1o, p5o, sum17, rot1_1); \ + dct_long_mac(p2o, p5o, sum35, rot1_2); \ + dct_long_mul(p3o, sum37, rot2_0); \ + dct_long_mul(p4o, sum15, rot2_1); \ + dct_wadd(sump13o, p1o, p3o); \ + dct_wadd(sump24o, p2o, p4o); \ + dct_wadd(sump23o, p2o, p3o); \ + dct_wadd(sump14o, p1o, p4o); \ + dct_long_mac(x4, sump13o, row7, rot3_0); \ + dct_long_mac(x5, sump24o, row5, rot3_1); \ + dct_long_mac(x6, sump23o, row3, rot3_2); \ + dct_long_mac(x7, sump14o, row1, rot3_3); \ + dct_bfly32o(row0,row7, x0,x7,shiftop,shift); \ + dct_bfly32o(row1,row6, x1,x6,shiftop,shift); \ + dct_bfly32o(row2,row5, x2,x5,shiftop,shift); \ + dct_bfly32o(row3,row4, x3,x4,shiftop,shift); \ + } + + // load + row0 = vld1q_s16(data + 0*8); + row1 = vld1q_s16(data + 1*8); + row2 = vld1q_s16(data + 2*8); + row3 = vld1q_s16(data + 3*8); + row4 = vld1q_s16(data + 4*8); + row5 = vld1q_s16(data + 5*8); + row6 = vld1q_s16(data + 6*8); + row7 = vld1q_s16(data + 7*8); + + // add DC bias + row0 = vaddq_s16(row0, vsetq_lane_s16(1024, vdupq_n_s16(0), 0)); + + // column pass + dct_pass(vrshrn_n_s32, 10); + + // 16bit 8x8 transpose + { +// these three map to a single VTRN.16, VTRN.32, and VSWP, respectively. +// whether compilers actually get this is another story, sadly. +#define dct_trn16(x, y) { int16x8x2_t t = vtrnq_s16(x, y); x = t.val[0]; y = t.val[1]; } +#define dct_trn32(x, y) { int32x4x2_t t = vtrnq_s32(vreinterpretq_s32_s16(x), vreinterpretq_s32_s16(y)); x = vreinterpretq_s16_s32(t.val[0]); y = vreinterpretq_s16_s32(t.val[1]); } +#define dct_trn64(x, y) { int16x8_t x0 = x; int16x8_t y0 = y; x = vcombine_s16(vget_low_s16(x0), vget_low_s16(y0)); y = vcombine_s16(vget_high_s16(x0), vget_high_s16(y0)); } + + // pass 1 + dct_trn16(row0, row1); // a0b0a2b2a4b4a6b6 + dct_trn16(row2, row3); + dct_trn16(row4, row5); + dct_trn16(row6, row7); + + // pass 2 + dct_trn32(row0, row2); // a0b0c0d0a4b4c4d4 + dct_trn32(row1, row3); + dct_trn32(row4, row6); + dct_trn32(row5, row7); + + // pass 3 + dct_trn64(row0, row4); // a0b0c0d0e0f0g0h0 + dct_trn64(row1, row5); + dct_trn64(row2, row6); + dct_trn64(row3, row7); + +#undef dct_trn16 +#undef dct_trn32 +#undef dct_trn64 + } + + // row pass + // vrshrn_n_s32 only supports shifts up to 16, we need + // 17. so do a non-rounding shift of 16 first then follow + // up with a rounding shift by 1. + dct_pass(vshrn_n_s32, 16); + + { + // pack and round + uint8x8_t p0 = vqrshrun_n_s16(row0, 1); + uint8x8_t p1 = vqrshrun_n_s16(row1, 1); + uint8x8_t p2 = vqrshrun_n_s16(row2, 1); + uint8x8_t p3 = vqrshrun_n_s16(row3, 1); + uint8x8_t p4 = vqrshrun_n_s16(row4, 1); + uint8x8_t p5 = vqrshrun_n_s16(row5, 1); + uint8x8_t p6 = vqrshrun_n_s16(row6, 1); + uint8x8_t p7 = vqrshrun_n_s16(row7, 1); + + // again, these can translate into one instruction, but often don't. +#define dct_trn8_8(x, y) { uint8x8x2_t t = vtrn_u8(x, y); x = t.val[0]; y = t.val[1]; } +#define dct_trn8_16(x, y) { uint16x4x2_t t = vtrn_u16(vreinterpret_u16_u8(x), vreinterpret_u16_u8(y)); x = vreinterpret_u8_u16(t.val[0]); y = vreinterpret_u8_u16(t.val[1]); } +#define dct_trn8_32(x, y) { uint32x2x2_t t = vtrn_u32(vreinterpret_u32_u8(x), vreinterpret_u32_u8(y)); x = vreinterpret_u8_u32(t.val[0]); y = vreinterpret_u8_u32(t.val[1]); } + + // sadly can't use interleaved stores here since we only write + // 8 bytes to each scan line! + + // 8x8 8-bit transpose pass 1 + dct_trn8_8(p0, p1); + dct_trn8_8(p2, p3); + dct_trn8_8(p4, p5); + dct_trn8_8(p6, p7); + + // pass 2 + dct_trn8_16(p0, p2); + dct_trn8_16(p1, p3); + dct_trn8_16(p4, p6); + dct_trn8_16(p5, p7); + + // pass 3 + dct_trn8_32(p0, p4); + dct_trn8_32(p1, p5); + dct_trn8_32(p2, p6); + dct_trn8_32(p3, p7); + + // store + vst1_u8(out, p0); out += out_stride; + vst1_u8(out, p1); out += out_stride; + vst1_u8(out, p2); out += out_stride; + vst1_u8(out, p3); out += out_stride; + vst1_u8(out, p4); out += out_stride; + vst1_u8(out, p5); out += out_stride; + vst1_u8(out, p6); out += out_stride; + vst1_u8(out, p7); + +#undef dct_trn8_8 +#undef dct_trn8_16 +#undef dct_trn8_32 + } + +#undef dct_long_mul +#undef dct_long_mac +#undef dct_widen +#undef dct_wadd +#undef dct_wsub +#undef dct_bfly32o +#undef dct_pass +} + +#endif // STBI_NEON + +#define STBI__MARKER_none 0xff +// if there's a pending marker from the entropy stream, return that +// otherwise, fetch from the stream and get a marker. if there's no +// marker, return 0xff, which is never a valid marker value +static stbi_uc stbi__get_marker(stbi__jpeg *j) +{ + stbi_uc x; + if (j->marker != STBI__MARKER_none) { x = j->marker; j->marker = STBI__MARKER_none; return x; } + x = stbi__get8(j->s); + if (x != 0xff) return STBI__MARKER_none; + while (x == 0xff) + x = stbi__get8(j->s); // consume repeated 0xff fill bytes + return x; +} + +// in each scan, we'll have scan_n components, and the order +// of the components is specified by order[] +#define STBI__RESTART(x) ((x) >= 0xd0 && (x) <= 0xd7) + +// after a restart interval, stbi__jpeg_reset the entropy decoder and +// the dc prediction +static void stbi__jpeg_reset(stbi__jpeg *j) +{ + j->code_bits = 0; + j->code_buffer = 0; + j->nomore = 0; + j->img_comp[0].dc_pred = j->img_comp[1].dc_pred = j->img_comp[2].dc_pred = j->img_comp[3].dc_pred = 0; + j->marker = STBI__MARKER_none; + j->todo = j->restart_interval ? j->restart_interval : 0x7fffffff; + j->eob_run = 0; + // no more than 1<<31 MCUs if no restart_interal? that's plenty safe, + // since we don't even allow 1<<30 pixels +} + +static int stbi__parse_entropy_coded_data(stbi__jpeg *z) +{ + stbi__jpeg_reset(z); + if (!z->progressive) { + if (z->scan_n == 1) { + int i,j; + STBI_SIMD_ALIGN(short, data[64]); + int n = z->order[0]; + // non-interleaved data, we just need to process one block at a time, + // in trivial scanline order + // number of blocks to do just depends on how many actual "pixels" this + // component has, independent of interleaved MCU blocking and such + int w = (z->img_comp[n].x+7) >> 3; + int h = (z->img_comp[n].y+7) >> 3; + for (j=0; j < h; ++j) { + for (i=0; i < w; ++i) { + int ha = z->img_comp[n].ha; + if (!stbi__jpeg_decode_block(z, data, z->huff_dc+z->img_comp[n].hd, z->huff_ac+ha, z->fast_ac[ha], n, z->dequant[z->img_comp[n].tq])) return 0; + z->idct_block_kernel(z->img_comp[n].data+z->img_comp[n].w2*j*8+i*8, z->img_comp[n].w2, data); + // every data block is an MCU, so countdown the restart interval + if (--z->todo <= 0) { + if (z->code_bits < 24) stbi__grow_buffer_unsafe(z); + // if it's NOT a restart, then just bail, so we get corrupt data + // rather than no data + if (!STBI__RESTART(z->marker)) return 1; + stbi__jpeg_reset(z); + } + } + } + return 1; + } else { // interleaved + int i,j,k,x,y; + STBI_SIMD_ALIGN(short, data[64]); + for (j=0; j < z->img_mcu_y; ++j) { + for (i=0; i < z->img_mcu_x; ++i) { + // scan an interleaved mcu... process scan_n components in order + for (k=0; k < z->scan_n; ++k) { + int n = z->order[k]; + // scan out an mcu's worth of this component; that's just determined + // by the basic H and V specified for the component + for (y=0; y < z->img_comp[n].v; ++y) { + for (x=0; x < z->img_comp[n].h; ++x) { + int x2 = (i*z->img_comp[n].h + x)*8; + int y2 = (j*z->img_comp[n].v + y)*8; + int ha = z->img_comp[n].ha; + if (!stbi__jpeg_decode_block(z, data, z->huff_dc+z->img_comp[n].hd, z->huff_ac+ha, z->fast_ac[ha], n, z->dequant[z->img_comp[n].tq])) return 0; + z->idct_block_kernel(z->img_comp[n].data+z->img_comp[n].w2*y2+x2, z->img_comp[n].w2, data); + } + } + } + // after all interleaved components, that's an interleaved MCU, + // so now count down the restart interval + if (--z->todo <= 0) { + if (z->code_bits < 24) stbi__grow_buffer_unsafe(z); + if (!STBI__RESTART(z->marker)) return 1; + stbi__jpeg_reset(z); + } + } + } + return 1; + } + } else { + if (z->scan_n == 1) { + int i,j; + int n = z->order[0]; + // non-interleaved data, we just need to process one block at a time, + // in trivial scanline order + // number of blocks to do just depends on how many actual "pixels" this + // component has, independent of interleaved MCU blocking and such + int w = (z->img_comp[n].x+7) >> 3; + int h = (z->img_comp[n].y+7) >> 3; + for (j=0; j < h; ++j) { + for (i=0; i < w; ++i) { + short *data = z->img_comp[n].coeff + 64 * (i + j * z->img_comp[n].coeff_w); + if (z->spec_start == 0) { + if (!stbi__jpeg_decode_block_prog_dc(z, data, &z->huff_dc[z->img_comp[n].hd], n)) + return 0; + } else { + int ha = z->img_comp[n].ha; + if (!stbi__jpeg_decode_block_prog_ac(z, data, &z->huff_ac[ha], z->fast_ac[ha])) + return 0; + } + // every data block is an MCU, so countdown the restart interval + if (--z->todo <= 0) { + if (z->code_bits < 24) stbi__grow_buffer_unsafe(z); + if (!STBI__RESTART(z->marker)) return 1; + stbi__jpeg_reset(z); + } + } + } + return 1; + } else { // interleaved + int i,j,k,x,y; + for (j=0; j < z->img_mcu_y; ++j) { + for (i=0; i < z->img_mcu_x; ++i) { + // scan an interleaved mcu... process scan_n components in order + for (k=0; k < z->scan_n; ++k) { + int n = z->order[k]; + // scan out an mcu's worth of this component; that's just determined + // by the basic H and V specified for the component + for (y=0; y < z->img_comp[n].v; ++y) { + for (x=0; x < z->img_comp[n].h; ++x) { + int x2 = (i*z->img_comp[n].h + x); + int y2 = (j*z->img_comp[n].v + y); + short *data = z->img_comp[n].coeff + 64 * (x2 + y2 * z->img_comp[n].coeff_w); + if (!stbi__jpeg_decode_block_prog_dc(z, data, &z->huff_dc[z->img_comp[n].hd], n)) + return 0; + } + } + } + // after all interleaved components, that's an interleaved MCU, + // so now count down the restart interval + if (--z->todo <= 0) { + if (z->code_bits < 24) stbi__grow_buffer_unsafe(z); + if (!STBI__RESTART(z->marker)) return 1; + stbi__jpeg_reset(z); + } + } + } + return 1; + } + } +} + +static void stbi__jpeg_dequantize(short *data, stbi__uint16 *dequant) +{ + int i; + for (i=0; i < 64; ++i) + data[i] *= dequant[i]; +} + +static void stbi__jpeg_finish(stbi__jpeg *z) +{ + if (z->progressive) { + // dequantize and idct the data + int i,j,n; + for (n=0; n < z->s->img_n; ++n) { + int w = (z->img_comp[n].x+7) >> 3; + int h = (z->img_comp[n].y+7) >> 3; + for (j=0; j < h; ++j) { + for (i=0; i < w; ++i) { + short *data = z->img_comp[n].coeff + 64 * (i + j * z->img_comp[n].coeff_w); + stbi__jpeg_dequantize(data, z->dequant[z->img_comp[n].tq]); + z->idct_block_kernel(z->img_comp[n].data+z->img_comp[n].w2*j*8+i*8, z->img_comp[n].w2, data); + } + } + } + } +} + +static int stbi__process_marker(stbi__jpeg *z, int m) +{ + int L; + switch (m) { + case STBI__MARKER_none: // no marker found + return stbi__err("expected marker","Corrupt JPEG"); + + case 0xDD: // DRI - specify restart interval + if (stbi__get16be(z->s) != 4) return stbi__err("bad DRI len","Corrupt JPEG"); + z->restart_interval = stbi__get16be(z->s); + return 1; + + case 0xDB: // DQT - define quantization table + L = stbi__get16be(z->s)-2; + while (L > 0) { + int q = stbi__get8(z->s); + int p = q >> 4, sixteen = (p != 0); + int t = q & 15,i; + if (p != 0 && p != 1) return stbi__err("bad DQT type","Corrupt JPEG"); + if (t > 3) return stbi__err("bad DQT table","Corrupt JPEG"); + + for (i=0; i < 64; ++i) + z->dequant[t][stbi__jpeg_dezigzag[i]] = (stbi__uint16)(sixteen ? stbi__get16be(z->s) : stbi__get8(z->s)); + L -= (sixteen ? 129 : 65); + } + return L==0; + + case 0xC4: // DHT - define huffman table + L = stbi__get16be(z->s)-2; + while (L > 0) { + stbi_uc *v; + int sizes[16],i,n=0; + int q = stbi__get8(z->s); + int tc = q >> 4; + int th = q & 15; + if (tc > 1 || th > 3) return stbi__err("bad DHT header","Corrupt JPEG"); + for (i=0; i < 16; ++i) { + sizes[i] = stbi__get8(z->s); + n += sizes[i]; + } + L -= 17; + if (tc == 0) { + if (!stbi__build_huffman(z->huff_dc+th, sizes)) return 0; + v = z->huff_dc[th].values; + } else { + if (!stbi__build_huffman(z->huff_ac+th, sizes)) return 0; + v = z->huff_ac[th].values; + } + for (i=0; i < n; ++i) + v[i] = stbi__get8(z->s); + if (tc != 0) + stbi__build_fast_ac(z->fast_ac[th], z->huff_ac + th); + L -= n; + } + return L==0; + } + + // check for comment block or APP blocks + if ((m >= 0xE0 && m <= 0xEF) || m == 0xFE) { + L = stbi__get16be(z->s); + if (L < 2) { + if (m == 0xFE) + return stbi__err("bad COM len","Corrupt JPEG"); + else + return stbi__err("bad APP len","Corrupt JPEG"); + } + L -= 2; + + if (m == 0xE0 && L >= 5) { // JFIF APP0 segment + static const unsigned char tag[5] = {'J','F','I','F','\0'}; + int ok = 1; + int i; + for (i=0; i < 5; ++i) + if (stbi__get8(z->s) != tag[i]) + ok = 0; + L -= 5; + if (ok) + z->jfif = 1; + } else if (m == 0xEE && L >= 12) { // Adobe APP14 segment + static const unsigned char tag[6] = {'A','d','o','b','e','\0'}; + int ok = 1; + int i; + for (i=0; i < 6; ++i) + if (stbi__get8(z->s) != tag[i]) + ok = 0; + L -= 6; + if (ok) { + stbi__get8(z->s); // version + stbi__get16be(z->s); // flags0 + stbi__get16be(z->s); // flags1 + z->app14_color_transform = stbi__get8(z->s); // color transform + L -= 6; + } + } + + stbi__skip(z->s, L); + return 1; + } + + return stbi__err("unknown marker","Corrupt JPEG"); +} + +// after we see SOS +static int stbi__process_scan_header(stbi__jpeg *z) +{ + int i; + int Ls = stbi__get16be(z->s); + z->scan_n = stbi__get8(z->s); + if (z->scan_n < 1 || z->scan_n > 4 || z->scan_n > (int) z->s->img_n) return stbi__err("bad SOS component count","Corrupt JPEG"); + if (Ls != 6+2*z->scan_n) return stbi__err("bad SOS len","Corrupt JPEG"); + for (i=0; i < z->scan_n; ++i) { + int id = stbi__get8(z->s), which; + int q = stbi__get8(z->s); + for (which = 0; which < z->s->img_n; ++which) + if (z->img_comp[which].id == id) + break; + if (which == z->s->img_n) return 0; // no match + z->img_comp[which].hd = q >> 4; if (z->img_comp[which].hd > 3) return stbi__err("bad DC huff","Corrupt JPEG"); + z->img_comp[which].ha = q & 15; if (z->img_comp[which].ha > 3) return stbi__err("bad AC huff","Corrupt JPEG"); + z->order[i] = which; + } + + { + int aa; + z->spec_start = stbi__get8(z->s); + z->spec_end = stbi__get8(z->s); // should be 63, but might be 0 + aa = stbi__get8(z->s); + z->succ_high = (aa >> 4); + z->succ_low = (aa & 15); + if (z->progressive) { + if (z->spec_start > 63 || z->spec_end > 63 || z->spec_start > z->spec_end || z->succ_high > 13 || z->succ_low > 13) + return stbi__err("bad SOS", "Corrupt JPEG"); + } else { + if (z->spec_start != 0) return stbi__err("bad SOS","Corrupt JPEG"); + if (z->succ_high != 0 || z->succ_low != 0) return stbi__err("bad SOS","Corrupt JPEG"); + z->spec_end = 63; + } + } + + return 1; +} + +static int stbi__free_jpeg_components(stbi__jpeg *z, int ncomp, int why) +{ + int i; + for (i=0; i < ncomp; ++i) { + if (z->img_comp[i].raw_data) { + STBI_FREE(z->img_comp[i].raw_data); + z->img_comp[i].raw_data = NULL; + z->img_comp[i].data = NULL; + } + if (z->img_comp[i].raw_coeff) { + STBI_FREE(z->img_comp[i].raw_coeff); + z->img_comp[i].raw_coeff = 0; + z->img_comp[i].coeff = 0; + } + if (z->img_comp[i].linebuf) { + STBI_FREE(z->img_comp[i].linebuf); + z->img_comp[i].linebuf = NULL; + } + } + return why; +} + +static int stbi__process_frame_header(stbi__jpeg *z, int scan) +{ + stbi__context *s = z->s; + int Lf,p,i,q, h_max=1,v_max=1,c; + Lf = stbi__get16be(s); if (Lf < 11) return stbi__err("bad SOF len","Corrupt JPEG"); // JPEG + p = stbi__get8(s); if (p != 8) return stbi__err("only 8-bit","JPEG format not supported: 8-bit only"); // JPEG baseline + s->img_y = stbi__get16be(s); if (s->img_y == 0) return stbi__err("no header height", "JPEG format not supported: delayed height"); // Legal, but we don't handle it--but neither does IJG + s->img_x = stbi__get16be(s); if (s->img_x == 0) return stbi__err("0 width","Corrupt JPEG"); // JPEG requires + if (s->img_y > STBI_MAX_DIMENSIONS) return stbi__err("too large","Very large image (corrupt?)"); + if (s->img_x > STBI_MAX_DIMENSIONS) return stbi__err("too large","Very large image (corrupt?)"); + c = stbi__get8(s); + if (c != 3 && c != 1 && c != 4) return stbi__err("bad component count","Corrupt JPEG"); + s->img_n = c; + for (i=0; i < c; ++i) { + z->img_comp[i].data = NULL; + z->img_comp[i].linebuf = NULL; + } + + if (Lf != 8+3*s->img_n) return stbi__err("bad SOF len","Corrupt JPEG"); + + z->rgb = 0; + for (i=0; i < s->img_n; ++i) { + static const unsigned char rgb[3] = { 'R', 'G', 'B' }; + z->img_comp[i].id = stbi__get8(s); + if (s->img_n == 3 && z->img_comp[i].id == rgb[i]) + ++z->rgb; + q = stbi__get8(s); + z->img_comp[i].h = (q >> 4); if (!z->img_comp[i].h || z->img_comp[i].h > 4) return stbi__err("bad H","Corrupt JPEG"); + z->img_comp[i].v = q & 15; if (!z->img_comp[i].v || z->img_comp[i].v > 4) return stbi__err("bad V","Corrupt JPEG"); + z->img_comp[i].tq = stbi__get8(s); if (z->img_comp[i].tq > 3) return stbi__err("bad TQ","Corrupt JPEG"); + } + + if (scan != STBI__SCAN_load) return 1; + + if (!stbi__mad3sizes_valid(s->img_x, s->img_y, s->img_n, 0)) return stbi__err("too large", "Image too large to decode"); + + for (i=0; i < s->img_n; ++i) { + if (z->img_comp[i].h > h_max) h_max = z->img_comp[i].h; + if (z->img_comp[i].v > v_max) v_max = z->img_comp[i].v; + } + + // check that plane subsampling factors are integer ratios; our resamplers can't deal with fractional ratios + // and I've never seen a non-corrupted JPEG file actually use them + for (i=0; i < s->img_n; ++i) { + if (h_max % z->img_comp[i].h != 0) return stbi__err("bad H","Corrupt JPEG"); + if (v_max % z->img_comp[i].v != 0) return stbi__err("bad V","Corrupt JPEG"); + } + + // compute interleaved mcu info + z->img_h_max = h_max; + z->img_v_max = v_max; + z->img_mcu_w = h_max * 8; + z->img_mcu_h = v_max * 8; + // these sizes can't be more than 17 bits + z->img_mcu_x = (s->img_x + z->img_mcu_w-1) / z->img_mcu_w; + z->img_mcu_y = (s->img_y + z->img_mcu_h-1) / z->img_mcu_h; + + for (i=0; i < s->img_n; ++i) { + // number of effective pixels (e.g. for non-interleaved MCU) + z->img_comp[i].x = (s->img_x * z->img_comp[i].h + h_max-1) / h_max; + z->img_comp[i].y = (s->img_y * z->img_comp[i].v + v_max-1) / v_max; + // to simplify generation, we'll allocate enough memory to decode + // the bogus oversized data from using interleaved MCUs and their + // big blocks (e.g. a 16x16 iMCU on an image of width 33); we won't + // discard the extra data until colorspace conversion + // + // img_mcu_x, img_mcu_y: <=17 bits; comp[i].h and .v are <=4 (checked earlier) + // so these muls can't overflow with 32-bit ints (which we require) + z->img_comp[i].w2 = z->img_mcu_x * z->img_comp[i].h * 8; + z->img_comp[i].h2 = z->img_mcu_y * z->img_comp[i].v * 8; + z->img_comp[i].coeff = 0; + z->img_comp[i].raw_coeff = 0; + z->img_comp[i].linebuf = NULL; + z->img_comp[i].raw_data = stbi__malloc_mad2(z->img_comp[i].w2, z->img_comp[i].h2, 15); + if (z->img_comp[i].raw_data == NULL) + return stbi__free_jpeg_components(z, i+1, stbi__err("outofmem", "Out of memory")); + // align blocks for idct using mmx/sse + z->img_comp[i].data = (stbi_uc*) (((size_t) z->img_comp[i].raw_data + 15) & ~15); + if (z->progressive) { + // w2, h2 are multiples of 8 (see above) + z->img_comp[i].coeff_w = z->img_comp[i].w2 / 8; + z->img_comp[i].coeff_h = z->img_comp[i].h2 / 8; + z->img_comp[i].raw_coeff = stbi__malloc_mad3(z->img_comp[i].w2, z->img_comp[i].h2, sizeof(short), 15); + if (z->img_comp[i].raw_coeff == NULL) + return stbi__free_jpeg_components(z, i+1, stbi__err("outofmem", "Out of memory")); + z->img_comp[i].coeff = (short*) (((size_t) z->img_comp[i].raw_coeff + 15) & ~15); + } + } + + return 1; +} + +// use comparisons since in some cases we handle more than one case (e.g. SOF) +#define stbi__DNL(x) ((x) == 0xdc) +#define stbi__SOI(x) ((x) == 0xd8) +#define stbi__EOI(x) ((x) == 0xd9) +#define stbi__SOF(x) ((x) == 0xc0 || (x) == 0xc1 || (x) == 0xc2) +#define stbi__SOS(x) ((x) == 0xda) + +#define stbi__SOF_progressive(x) ((x) == 0xc2) + +static int stbi__decode_jpeg_header(stbi__jpeg *z, int scan) +{ + int m; + z->jfif = 0; + z->app14_color_transform = -1; // valid values are 0,1,2 + z->marker = STBI__MARKER_none; // initialize cached marker to empty + m = stbi__get_marker(z); + if (!stbi__SOI(m)) return stbi__err("no SOI","Corrupt JPEG"); + if (scan == STBI__SCAN_type) return 1; + m = stbi__get_marker(z); + while (!stbi__SOF(m)) { + if (!stbi__process_marker(z,m)) return 0; + m = stbi__get_marker(z); + while (m == STBI__MARKER_none) { + // some files have extra padding after their blocks, so ok, we'll scan + if (stbi__at_eof(z->s)) return stbi__err("no SOF", "Corrupt JPEG"); + m = stbi__get_marker(z); + } + } + z->progressive = stbi__SOF_progressive(m); + if (!stbi__process_frame_header(z, scan)) return 0; + return 1; +} + +// decode image to YCbCr format +static int stbi__decode_jpeg_image(stbi__jpeg *j) +{ + int m; + for (m = 0; m < 4; m++) { + j->img_comp[m].raw_data = NULL; + j->img_comp[m].raw_coeff = NULL; + } + j->restart_interval = 0; + if (!stbi__decode_jpeg_header(j, STBI__SCAN_load)) return 0; + m = stbi__get_marker(j); + while (!stbi__EOI(m)) { + if (stbi__SOS(m)) { + if (!stbi__process_scan_header(j)) return 0; + if (!stbi__parse_entropy_coded_data(j)) return 0; + if (j->marker == STBI__MARKER_none ) { + // handle 0s at the end of image data from IP Kamera 9060 + while (!stbi__at_eof(j->s)) { + int x = stbi__get8(j->s); + if (x == 255) { + j->marker = stbi__get8(j->s); + break; + } + } + // if we reach eof without hitting a marker, stbi__get_marker() below will fail and we'll eventually return 0 + } + } else if (stbi__DNL(m)) { + int Ld = stbi__get16be(j->s); + stbi__uint32 NL = stbi__get16be(j->s); + if (Ld != 4) return stbi__err("bad DNL len", "Corrupt JPEG"); + if (NL != j->s->img_y) return stbi__err("bad DNL height", "Corrupt JPEG"); + } else { + if (!stbi__process_marker(j, m)) return 0; + } + m = stbi__get_marker(j); + } + if (j->progressive) + stbi__jpeg_finish(j); + return 1; +} + +// static jfif-centered resampling (across block boundaries) + +typedef stbi_uc *(*resample_row_func)(stbi_uc *out, stbi_uc *in0, stbi_uc *in1, + int w, int hs); + +#define stbi__div4(x) ((stbi_uc) ((x) >> 2)) + +static stbi_uc *resample_row_1(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs) +{ + STBI_NOTUSED(out); + STBI_NOTUSED(in_far); + STBI_NOTUSED(w); + STBI_NOTUSED(hs); + return in_near; +} + +static stbi_uc* stbi__resample_row_v_2(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs) +{ + // need to generate two samples vertically for every one in input + int i; + STBI_NOTUSED(hs); + for (i=0; i < w; ++i) + out[i] = stbi__div4(3*in_near[i] + in_far[i] + 2); + return out; +} + +static stbi_uc* stbi__resample_row_h_2(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs) +{ + // need to generate two samples horizontally for every one in input + int i; + stbi_uc *input = in_near; + + if (w == 1) { + // if only one sample, can't do any interpolation + out[0] = out[1] = input[0]; + return out; + } + + out[0] = input[0]; + out[1] = stbi__div4(input[0]*3 + input[1] + 2); + for (i=1; i < w-1; ++i) { + int n = 3*input[i]+2; + out[i*2+0] = stbi__div4(n+input[i-1]); + out[i*2+1] = stbi__div4(n+input[i+1]); + } + out[i*2+0] = stbi__div4(input[w-2]*3 + input[w-1] + 2); + out[i*2+1] = input[w-1]; + + STBI_NOTUSED(in_far); + STBI_NOTUSED(hs); + + return out; +} + +#define stbi__div16(x) ((stbi_uc) ((x) >> 4)) + +static stbi_uc *stbi__resample_row_hv_2(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs) +{ + // need to generate 2x2 samples for every one in input + int i,t0,t1; + if (w == 1) { + out[0] = out[1] = stbi__div4(3*in_near[0] + in_far[0] + 2); + return out; + } + + t1 = 3*in_near[0] + in_far[0]; + out[0] = stbi__div4(t1+2); + for (i=1; i < w; ++i) { + t0 = t1; + t1 = 3*in_near[i]+in_far[i]; + out[i*2-1] = stbi__div16(3*t0 + t1 + 8); + out[i*2 ] = stbi__div16(3*t1 + t0 + 8); + } + out[w*2-1] = stbi__div4(t1+2); + + STBI_NOTUSED(hs); + + return out; +} + +#if defined(STBI_SSE2) || defined(STBI_NEON) +static stbi_uc *stbi__resample_row_hv_2_simd(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs) +{ + // need to generate 2x2 samples for every one in input + int i=0,t0,t1; + + if (w == 1) { + out[0] = out[1] = stbi__div4(3*in_near[0] + in_far[0] + 2); + return out; + } + + t1 = 3*in_near[0] + in_far[0]; + // process groups of 8 pixels for as long as we can. + // note we can't handle the last pixel in a row in this loop + // because we need to handle the filter boundary conditions. + for (; i < ((w-1) & ~7); i += 8) { +#if defined(STBI_SSE2) + // load and perform the vertical filtering pass + // this uses 3*x + y = 4*x + (y - x) + __m128i zero = _mm_setzero_si128(); + __m128i farb = _mm_loadl_epi64((__m128i *) (in_far + i)); + __m128i nearb = _mm_loadl_epi64((__m128i *) (in_near + i)); + __m128i farw = _mm_unpacklo_epi8(farb, zero); + __m128i nearw = _mm_unpacklo_epi8(nearb, zero); + __m128i diff = _mm_sub_epi16(farw, nearw); + __m128i nears = _mm_slli_epi16(nearw, 2); + __m128i curr = _mm_add_epi16(nears, diff); // current row + + // horizontal filter works the same based on shifted vers of current + // row. "prev" is current row shifted right by 1 pixel; we need to + // insert the previous pixel value (from t1). + // "next" is current row shifted left by 1 pixel, with first pixel + // of next block of 8 pixels added in. + __m128i prv0 = _mm_slli_si128(curr, 2); + __m128i nxt0 = _mm_srli_si128(curr, 2); + __m128i prev = _mm_insert_epi16(prv0, t1, 0); + __m128i next = _mm_insert_epi16(nxt0, 3*in_near[i+8] + in_far[i+8], 7); + + // horizontal filter, polyphase implementation since it's convenient: + // even pixels = 3*cur + prev = cur*4 + (prev - cur) + // odd pixels = 3*cur + next = cur*4 + (next - cur) + // note the shared term. + __m128i bias = _mm_set1_epi16(8); + __m128i curs = _mm_slli_epi16(curr, 2); + __m128i prvd = _mm_sub_epi16(prev, curr); + __m128i nxtd = _mm_sub_epi16(next, curr); + __m128i curb = _mm_add_epi16(curs, bias); + __m128i even = _mm_add_epi16(prvd, curb); + __m128i odd = _mm_add_epi16(nxtd, curb); + + // interleave even and odd pixels, then undo scaling. + __m128i int0 = _mm_unpacklo_epi16(even, odd); + __m128i int1 = _mm_unpackhi_epi16(even, odd); + __m128i de0 = _mm_srli_epi16(int0, 4); + __m128i de1 = _mm_srli_epi16(int1, 4); + + // pack and write output + __m128i outv = _mm_packus_epi16(de0, de1); + _mm_storeu_si128((__m128i *) (out + i*2), outv); +#elif defined(STBI_NEON) + // load and perform the vertical filtering pass + // this uses 3*x + y = 4*x + (y - x) + uint8x8_t farb = vld1_u8(in_far + i); + uint8x8_t nearb = vld1_u8(in_near + i); + int16x8_t diff = vreinterpretq_s16_u16(vsubl_u8(farb, nearb)); + int16x8_t nears = vreinterpretq_s16_u16(vshll_n_u8(nearb, 2)); + int16x8_t curr = vaddq_s16(nears, diff); // current row + + // horizontal filter works the same based on shifted vers of current + // row. "prev" is current row shifted right by 1 pixel; we need to + // insert the previous pixel value (from t1). + // "next" is current row shifted left by 1 pixel, with first pixel + // of next block of 8 pixels added in. + int16x8_t prv0 = vextq_s16(curr, curr, 7); + int16x8_t nxt0 = vextq_s16(curr, curr, 1); + int16x8_t prev = vsetq_lane_s16(t1, prv0, 0); + int16x8_t next = vsetq_lane_s16(3*in_near[i+8] + in_far[i+8], nxt0, 7); + + // horizontal filter, polyphase implementation since it's convenient: + // even pixels = 3*cur + prev = cur*4 + (prev - cur) + // odd pixels = 3*cur + next = cur*4 + (next - cur) + // note the shared term. + int16x8_t curs = vshlq_n_s16(curr, 2); + int16x8_t prvd = vsubq_s16(prev, curr); + int16x8_t nxtd = vsubq_s16(next, curr); + int16x8_t even = vaddq_s16(curs, prvd); + int16x8_t odd = vaddq_s16(curs, nxtd); + + // undo scaling and round, then store with even/odd phases interleaved + uint8x8x2_t o; + o.val[0] = vqrshrun_n_s16(even, 4); + o.val[1] = vqrshrun_n_s16(odd, 4); + vst2_u8(out + i*2, o); +#endif + + // "previous" value for next iter + t1 = 3*in_near[i+7] + in_far[i+7]; + } + + t0 = t1; + t1 = 3*in_near[i] + in_far[i]; + out[i*2] = stbi__div16(3*t1 + t0 + 8); + + for (++i; i < w; ++i) { + t0 = t1; + t1 = 3*in_near[i]+in_far[i]; + out[i*2-1] = stbi__div16(3*t0 + t1 + 8); + out[i*2 ] = stbi__div16(3*t1 + t0 + 8); + } + out[w*2-1] = stbi__div4(t1+2); + + STBI_NOTUSED(hs); + + return out; +} +#endif + +static stbi_uc *stbi__resample_row_generic(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs) +{ + // resample with nearest-neighbor + int i,j; + STBI_NOTUSED(in_far); + for (i=0; i < w; ++i) + for (j=0; j < hs; ++j) + out[i*hs+j] = in_near[i]; + return out; +} + +// this is a reduced-precision calculation of YCbCr-to-RGB introduced +// to make sure the code produces the same results in both SIMD and scalar +#define stbi__float2fixed(x) (((int) ((x) * 4096.0f + 0.5f)) << 8) +static void stbi__YCbCr_to_RGB_row(stbi_uc *out, const stbi_uc *y, const stbi_uc *pcb, const stbi_uc *pcr, int count, int step) +{ + int i; + for (i=0; i < count; ++i) { + int y_fixed = (y[i] << 20) + (1<<19); // rounding + int r,g,b; + int cr = pcr[i] - 128; + int cb = pcb[i] - 128; + r = y_fixed + cr* stbi__float2fixed(1.40200f); + g = y_fixed + (cr*-stbi__float2fixed(0.71414f)) + ((cb*-stbi__float2fixed(0.34414f)) & 0xffff0000); + b = y_fixed + cb* stbi__float2fixed(1.77200f); + r >>= 20; + g >>= 20; + b >>= 20; + if ((unsigned) r > 255) { if (r < 0) r = 0; else r = 255; } + if ((unsigned) g > 255) { if (g < 0) g = 0; else g = 255; } + if ((unsigned) b > 255) { if (b < 0) b = 0; else b = 255; } + out[0] = (stbi_uc)r; + out[1] = (stbi_uc)g; + out[2] = (stbi_uc)b; + out[3] = 255; + out += step; + } +} + +#if defined(STBI_SSE2) || defined(STBI_NEON) +static void stbi__YCbCr_to_RGB_simd(stbi_uc *out, stbi_uc const *y, stbi_uc const *pcb, stbi_uc const *pcr, int count, int step) +{ + int i = 0; + +#ifdef STBI_SSE2 + // step == 3 is pretty ugly on the final interleave, and i'm not convinced + // it's useful in practice (you wouldn't use it for textures, for example). + // so just accelerate step == 4 case. + if (step == 4) { + // this is a fairly straightforward implementation and not super-optimized. + __m128i signflip = _mm_set1_epi8(-0x80); + __m128i cr_const0 = _mm_set1_epi16( (short) ( 1.40200f*4096.0f+0.5f)); + __m128i cr_const1 = _mm_set1_epi16( - (short) ( 0.71414f*4096.0f+0.5f)); + __m128i cb_const0 = _mm_set1_epi16( - (short) ( 0.34414f*4096.0f+0.5f)); + __m128i cb_const1 = _mm_set1_epi16( (short) ( 1.77200f*4096.0f+0.5f)); + __m128i y_bias = _mm_set1_epi8((char) (unsigned char) 128); + __m128i xw = _mm_set1_epi16(255); // alpha channel + + for (; i+7 < count; i += 8) { + // load + __m128i y_bytes = _mm_loadl_epi64((__m128i *) (y+i)); + __m128i cr_bytes = _mm_loadl_epi64((__m128i *) (pcr+i)); + __m128i cb_bytes = _mm_loadl_epi64((__m128i *) (pcb+i)); + __m128i cr_biased = _mm_xor_si128(cr_bytes, signflip); // -128 + __m128i cb_biased = _mm_xor_si128(cb_bytes, signflip); // -128 + + // unpack to short (and left-shift cr, cb by 8) + __m128i yw = _mm_unpacklo_epi8(y_bias, y_bytes); + __m128i crw = _mm_unpacklo_epi8(_mm_setzero_si128(), cr_biased); + __m128i cbw = _mm_unpacklo_epi8(_mm_setzero_si128(), cb_biased); + + // color transform + __m128i yws = _mm_srli_epi16(yw, 4); + __m128i cr0 = _mm_mulhi_epi16(cr_const0, crw); + __m128i cb0 = _mm_mulhi_epi16(cb_const0, cbw); + __m128i cb1 = _mm_mulhi_epi16(cbw, cb_const1); + __m128i cr1 = _mm_mulhi_epi16(crw, cr_const1); + __m128i rws = _mm_add_epi16(cr0, yws); + __m128i gwt = _mm_add_epi16(cb0, yws); + __m128i bws = _mm_add_epi16(yws, cb1); + __m128i gws = _mm_add_epi16(gwt, cr1); + + // descale + __m128i rw = _mm_srai_epi16(rws, 4); + __m128i bw = _mm_srai_epi16(bws, 4); + __m128i gw = _mm_srai_epi16(gws, 4); + + // back to byte, set up for transpose + __m128i brb = _mm_packus_epi16(rw, bw); + __m128i gxb = _mm_packus_epi16(gw, xw); + + // transpose to interleave channels + __m128i t0 = _mm_unpacklo_epi8(brb, gxb); + __m128i t1 = _mm_unpackhi_epi8(brb, gxb); + __m128i o0 = _mm_unpacklo_epi16(t0, t1); + __m128i o1 = _mm_unpackhi_epi16(t0, t1); + + // store + _mm_storeu_si128((__m128i *) (out + 0), o0); + _mm_storeu_si128((__m128i *) (out + 16), o1); + out += 32; + } + } +#endif + +#ifdef STBI_NEON + // in this version, step=3 support would be easy to add. but is there demand? + if (step == 4) { + // this is a fairly straightforward implementation and not super-optimized. + uint8x8_t signflip = vdup_n_u8(0x80); + int16x8_t cr_const0 = vdupq_n_s16( (short) ( 1.40200f*4096.0f+0.5f)); + int16x8_t cr_const1 = vdupq_n_s16( - (short) ( 0.71414f*4096.0f+0.5f)); + int16x8_t cb_const0 = vdupq_n_s16( - (short) ( 0.34414f*4096.0f+0.5f)); + int16x8_t cb_const1 = vdupq_n_s16( (short) ( 1.77200f*4096.0f+0.5f)); + + for (; i+7 < count; i += 8) { + // load + uint8x8_t y_bytes = vld1_u8(y + i); + uint8x8_t cr_bytes = vld1_u8(pcr + i); + uint8x8_t cb_bytes = vld1_u8(pcb + i); + int8x8_t cr_biased = vreinterpret_s8_u8(vsub_u8(cr_bytes, signflip)); + int8x8_t cb_biased = vreinterpret_s8_u8(vsub_u8(cb_bytes, signflip)); + + // expand to s16 + int16x8_t yws = vreinterpretq_s16_u16(vshll_n_u8(y_bytes, 4)); + int16x8_t crw = vshll_n_s8(cr_biased, 7); + int16x8_t cbw = vshll_n_s8(cb_biased, 7); + + // color transform + int16x8_t cr0 = vqdmulhq_s16(crw, cr_const0); + int16x8_t cb0 = vqdmulhq_s16(cbw, cb_const0); + int16x8_t cr1 = vqdmulhq_s16(crw, cr_const1); + int16x8_t cb1 = vqdmulhq_s16(cbw, cb_const1); + int16x8_t rws = vaddq_s16(yws, cr0); + int16x8_t gws = vaddq_s16(vaddq_s16(yws, cb0), cr1); + int16x8_t bws = vaddq_s16(yws, cb1); + + // undo scaling, round, convert to byte + uint8x8x4_t o; + o.val[0] = vqrshrun_n_s16(rws, 4); + o.val[1] = vqrshrun_n_s16(gws, 4); + o.val[2] = vqrshrun_n_s16(bws, 4); + o.val[3] = vdup_n_u8(255); + + // store, interleaving r/g/b/a + vst4_u8(out, o); + out += 8*4; + } + } +#endif + + for (; i < count; ++i) { + int y_fixed = (y[i] << 20) + (1<<19); // rounding + int r,g,b; + int cr = pcr[i] - 128; + int cb = pcb[i] - 128; + r = y_fixed + cr* stbi__float2fixed(1.40200f); + g = y_fixed + cr*-stbi__float2fixed(0.71414f) + ((cb*-stbi__float2fixed(0.34414f)) & 0xffff0000); + b = y_fixed + cb* stbi__float2fixed(1.77200f); + r >>= 20; + g >>= 20; + b >>= 20; + if ((unsigned) r > 255) { if (r < 0) r = 0; else r = 255; } + if ((unsigned) g > 255) { if (g < 0) g = 0; else g = 255; } + if ((unsigned) b > 255) { if (b < 0) b = 0; else b = 255; } + out[0] = (stbi_uc)r; + out[1] = (stbi_uc)g; + out[2] = (stbi_uc)b; + out[3] = 255; + out += step; + } +} +#endif + +// set up the kernels +static void stbi__setup_jpeg(stbi__jpeg *j) +{ + j->idct_block_kernel = stbi__idct_block; + j->YCbCr_to_RGB_kernel = stbi__YCbCr_to_RGB_row; + j->resample_row_hv_2_kernel = stbi__resample_row_hv_2; + +#ifdef STBI_SSE2 + if (stbi__sse2_available()) { + j->idct_block_kernel = stbi__idct_simd; + j->YCbCr_to_RGB_kernel = stbi__YCbCr_to_RGB_simd; + j->resample_row_hv_2_kernel = stbi__resample_row_hv_2_simd; + } +#endif + +#ifdef STBI_NEON + j->idct_block_kernel = stbi__idct_simd; + j->YCbCr_to_RGB_kernel = stbi__YCbCr_to_RGB_simd; + j->resample_row_hv_2_kernel = stbi__resample_row_hv_2_simd; +#endif +} + +// clean up the temporary component buffers +static void stbi__cleanup_jpeg(stbi__jpeg *j) +{ + stbi__free_jpeg_components(j, j->s->img_n, 0); +} + +typedef struct +{ + resample_row_func resample; + stbi_uc *line0,*line1; + int hs,vs; // expansion factor in each axis + int w_lores; // horizontal pixels pre-expansion + int ystep; // how far through vertical expansion we are + int ypos; // which pre-expansion row we're on +} stbi__resample; + +// fast 0..255 * 0..255 => 0..255 rounded multiplication +static stbi_uc stbi__blinn_8x8(stbi_uc x, stbi_uc y) +{ + unsigned int t = x*y + 128; + return (stbi_uc) ((t + (t >>8)) >> 8); +} + +static stbi_uc *load_jpeg_image(stbi__jpeg *z, int *out_x, int *out_y, int *comp, int req_comp) +{ + int n, decode_n, is_rgb; + z->s->img_n = 0; // make stbi__cleanup_jpeg safe + + // validate req_comp + if (req_comp < 0 || req_comp > 4) return stbi__errpuc("bad req_comp", "Internal error"); + + // load a jpeg image from whichever source, but leave in YCbCr format + if (!stbi__decode_jpeg_image(z)) { stbi__cleanup_jpeg(z); return NULL; } + + // determine actual number of components to generate + n = req_comp ? req_comp : z->s->img_n >= 3 ? 3 : 1; + + is_rgb = z->s->img_n == 3 && (z->rgb == 3 || (z->app14_color_transform == 0 && !z->jfif)); + + if (z->s->img_n == 3 && n < 3 && !is_rgb) + decode_n = 1; + else + decode_n = z->s->img_n; + + // nothing to do if no components requested; check this now to avoid + // accessing uninitialized coutput[0] later + if (decode_n <= 0) { stbi__cleanup_jpeg(z); return NULL; } + + // resample and color-convert + { + int k; + unsigned int i,j; + stbi_uc *output; + stbi_uc *coutput[4] = { NULL, NULL, NULL, NULL }; + + stbi__resample res_comp[4]; + + for (k=0; k < decode_n; ++k) { + stbi__resample *r = &res_comp[k]; + + // allocate line buffer big enough for upsampling off the edges + // with upsample factor of 4 + z->img_comp[k].linebuf = (stbi_uc *) stbi__malloc(z->s->img_x + 3); + if (!z->img_comp[k].linebuf) { stbi__cleanup_jpeg(z); return stbi__errpuc("outofmem", "Out of memory"); } + + r->hs = z->img_h_max / z->img_comp[k].h; + r->vs = z->img_v_max / z->img_comp[k].v; + r->ystep = r->vs >> 1; + r->w_lores = (z->s->img_x + r->hs-1) / r->hs; + r->ypos = 0; + r->line0 = r->line1 = z->img_comp[k].data; + + if (r->hs == 1 && r->vs == 1) r->resample = resample_row_1; + else if (r->hs == 1 && r->vs == 2) r->resample = stbi__resample_row_v_2; + else if (r->hs == 2 && r->vs == 1) r->resample = stbi__resample_row_h_2; + else if (r->hs == 2 && r->vs == 2) r->resample = z->resample_row_hv_2_kernel; + else r->resample = stbi__resample_row_generic; + } + + // can't error after this so, this is safe + output = (stbi_uc *) stbi__malloc_mad3(n, z->s->img_x, z->s->img_y, 1); + if (!output) { stbi__cleanup_jpeg(z); return stbi__errpuc("outofmem", "Out of memory"); } + + // now go ahead and resample + for (j=0; j < z->s->img_y; ++j) { + stbi_uc *out = output + n * z->s->img_x * j; + for (k=0; k < decode_n; ++k) { + stbi__resample *r = &res_comp[k]; + int y_bot = r->ystep >= (r->vs >> 1); + coutput[k] = r->resample(z->img_comp[k].linebuf, + y_bot ? r->line1 : r->line0, + y_bot ? r->line0 : r->line1, + r->w_lores, r->hs); + if (++r->ystep >= r->vs) { + r->ystep = 0; + r->line0 = r->line1; + if (++r->ypos < z->img_comp[k].y) + r->line1 += z->img_comp[k].w2; + } + } + if (n >= 3) { + stbi_uc *y = coutput[0]; + if (z->s->img_n == 3) { + if (is_rgb) { + for (i=0; i < z->s->img_x; ++i) { + out[0] = y[i]; + out[1] = coutput[1][i]; + out[2] = coutput[2][i]; + out[3] = 255; + out += n; + } + } else { + z->YCbCr_to_RGB_kernel(out, y, coutput[1], coutput[2], z->s->img_x, n); + } + } else if (z->s->img_n == 4) { + if (z->app14_color_transform == 0) { // CMYK + for (i=0; i < z->s->img_x; ++i) { + stbi_uc m = coutput[3][i]; + out[0] = stbi__blinn_8x8(coutput[0][i], m); + out[1] = stbi__blinn_8x8(coutput[1][i], m); + out[2] = stbi__blinn_8x8(coutput[2][i], m); + out[3] = 255; + out += n; + } + } else if (z->app14_color_transform == 2) { // YCCK + z->YCbCr_to_RGB_kernel(out, y, coutput[1], coutput[2], z->s->img_x, n); + for (i=0; i < z->s->img_x; ++i) { + stbi_uc m = coutput[3][i]; + out[0] = stbi__blinn_8x8(255 - out[0], m); + out[1] = stbi__blinn_8x8(255 - out[1], m); + out[2] = stbi__blinn_8x8(255 - out[2], m); + out += n; + } + } else { // YCbCr + alpha? Ignore the fourth channel for now + z->YCbCr_to_RGB_kernel(out, y, coutput[1], coutput[2], z->s->img_x, n); + } + } else + for (i=0; i < z->s->img_x; ++i) { + out[0] = out[1] = out[2] = y[i]; + out[3] = 255; // not used if n==3 + out += n; + } + } else { + if (is_rgb) { + if (n == 1) + for (i=0; i < z->s->img_x; ++i) + *out++ = stbi__compute_y(coutput[0][i], coutput[1][i], coutput[2][i]); + else { + for (i=0; i < z->s->img_x; ++i, out += 2) { + out[0] = stbi__compute_y(coutput[0][i], coutput[1][i], coutput[2][i]); + out[1] = 255; + } + } + } else if (z->s->img_n == 4 && z->app14_color_transform == 0) { + for (i=0; i < z->s->img_x; ++i) { + stbi_uc m = coutput[3][i]; + stbi_uc r = stbi__blinn_8x8(coutput[0][i], m); + stbi_uc g = stbi__blinn_8x8(coutput[1][i], m); + stbi_uc b = stbi__blinn_8x8(coutput[2][i], m); + out[0] = stbi__compute_y(r, g, b); + out[1] = 255; + out += n; + } + } else if (z->s->img_n == 4 && z->app14_color_transform == 2) { + for (i=0; i < z->s->img_x; ++i) { + out[0] = stbi__blinn_8x8(255 - coutput[0][i], coutput[3][i]); + out[1] = 255; + out += n; + } + } else { + stbi_uc *y = coutput[0]; + if (n == 1) + for (i=0; i < z->s->img_x; ++i) out[i] = y[i]; + else + for (i=0; i < z->s->img_x; ++i) { *out++ = y[i]; *out++ = 255; } + } + } + } + stbi__cleanup_jpeg(z); + *out_x = z->s->img_x; + *out_y = z->s->img_y; + if (comp) *comp = z->s->img_n >= 3 ? 3 : 1; // report original components, not output + return output; + } +} + +static void *stbi__jpeg_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + unsigned char* result; + stbi__jpeg* j = (stbi__jpeg*) stbi__malloc(sizeof(stbi__jpeg)); + if (!j) return stbi__errpuc("outofmem", "Out of memory"); + STBI_NOTUSED(ri); + j->s = s; + stbi__setup_jpeg(j); + result = load_jpeg_image(j, x,y,comp,req_comp); + STBI_FREE(j); + return result; +} + +static int stbi__jpeg_test(stbi__context *s) +{ + int r; + stbi__jpeg* j = (stbi__jpeg*)stbi__malloc(sizeof(stbi__jpeg)); + if (!j) return stbi__err("outofmem", "Out of memory"); + j->s = s; + stbi__setup_jpeg(j); + r = stbi__decode_jpeg_header(j, STBI__SCAN_type); + stbi__rewind(s); + STBI_FREE(j); + return r; +} + +static int stbi__jpeg_info_raw(stbi__jpeg *j, int *x, int *y, int *comp) +{ + if (!stbi__decode_jpeg_header(j, STBI__SCAN_header)) { + stbi__rewind( j->s ); + return 0; + } + if (x) *x = j->s->img_x; + if (y) *y = j->s->img_y; + if (comp) *comp = j->s->img_n >= 3 ? 3 : 1; + return 1; +} + +static int stbi__jpeg_info(stbi__context *s, int *x, int *y, int *comp) +{ + int result; + stbi__jpeg* j = (stbi__jpeg*) (stbi__malloc(sizeof(stbi__jpeg))); + if (!j) return stbi__err("outofmem", "Out of memory"); + j->s = s; + result = stbi__jpeg_info_raw(j, x, y, comp); + STBI_FREE(j); + return result; +} +#endif + +// public domain zlib decode v0.2 Sean Barrett 2006-11-18 +// simple implementation +// - all input must be provided in an upfront buffer +// - all output is written to a single output buffer (can malloc/realloc) +// performance +// - fast huffman + +#ifndef STBI_NO_ZLIB + +// fast-way is faster to check than jpeg huffman, but slow way is slower +#define STBI__ZFAST_BITS 9 // accelerate all cases in default tables +#define STBI__ZFAST_MASK ((1 << STBI__ZFAST_BITS) - 1) +#define STBI__ZNSYMS 288 // number of symbols in literal/length alphabet + +// zlib-style huffman encoding +// (jpegs packs from left, zlib from right, so can't share code) +typedef struct +{ + stbi__uint16 fast[1 << STBI__ZFAST_BITS]; + stbi__uint16 firstcode[16]; + int maxcode[17]; + stbi__uint16 firstsymbol[16]; + stbi_uc size[STBI__ZNSYMS]; + stbi__uint16 value[STBI__ZNSYMS]; +} stbi__zhuffman; + +stbi_inline static int stbi__bitreverse16(int n) +{ + n = ((n & 0xAAAA) >> 1) | ((n & 0x5555) << 1); + n = ((n & 0xCCCC) >> 2) | ((n & 0x3333) << 2); + n = ((n & 0xF0F0) >> 4) | ((n & 0x0F0F) << 4); + n = ((n & 0xFF00) >> 8) | ((n & 0x00FF) << 8); + return n; +} + +stbi_inline static int stbi__bit_reverse(int v, int bits) +{ + STBI_ASSERT(bits <= 16); + // to bit reverse n bits, reverse 16 and shift + // e.g. 11 bits, bit reverse and shift away 5 + return stbi__bitreverse16(v) >> (16-bits); +} + +static int stbi__zbuild_huffman(stbi__zhuffman *z, const stbi_uc *sizelist, int num) +{ + int i,k=0; + int code, next_code[16], sizes[17]; + + // DEFLATE spec for generating codes + memset(sizes, 0, sizeof(sizes)); + memset(z->fast, 0, sizeof(z->fast)); + for (i=0; i < num; ++i) + ++sizes[sizelist[i]]; + sizes[0] = 0; + for (i=1; i < 16; ++i) + if (sizes[i] > (1 << i)) + return stbi__err("bad sizes", "Corrupt PNG"); + code = 0; + for (i=1; i < 16; ++i) { + next_code[i] = code; + z->firstcode[i] = (stbi__uint16) code; + z->firstsymbol[i] = (stbi__uint16) k; + code = (code + sizes[i]); + if (sizes[i]) + if (code-1 >= (1 << i)) return stbi__err("bad codelengths","Corrupt PNG"); + z->maxcode[i] = code << (16-i); // preshift for inner loop + code <<= 1; + k += sizes[i]; + } + z->maxcode[16] = 0x10000; // sentinel + for (i=0; i < num; ++i) { + int s = sizelist[i]; + if (s) { + int c = next_code[s] - z->firstcode[s] + z->firstsymbol[s]; + stbi__uint16 fastv = (stbi__uint16) ((s << 9) | i); + z->size [c] = (stbi_uc ) s; + z->value[c] = (stbi__uint16) i; + if (s <= STBI__ZFAST_BITS) { + int j = stbi__bit_reverse(next_code[s],s); + while (j < (1 << STBI__ZFAST_BITS)) { + z->fast[j] = fastv; + j += (1 << s); + } + } + ++next_code[s]; + } + } + return 1; +} + +// zlib-from-memory implementation for PNG reading +// because PNG allows splitting the zlib stream arbitrarily, +// and it's annoying structurally to have PNG call ZLIB call PNG, +// we require PNG read all the IDATs and combine them into a single +// memory buffer + +typedef struct +{ + stbi_uc *zbuffer, *zbuffer_end; + int num_bits; + stbi__uint32 code_buffer; + + char *zout; + char *zout_start; + char *zout_end; + int z_expandable; + + stbi__zhuffman z_length, z_distance; +} stbi__zbuf; + +stbi_inline static int stbi__zeof(stbi__zbuf *z) +{ + return (z->zbuffer >= z->zbuffer_end); +} + +stbi_inline static stbi_uc stbi__zget8(stbi__zbuf *z) +{ + return stbi__zeof(z) ? 0 : *z->zbuffer++; +} + +static void stbi__fill_bits(stbi__zbuf *z) +{ + do { + if (z->code_buffer >= (1U << z->num_bits)) { + z->zbuffer = z->zbuffer_end; /* treat this as EOF so we fail. */ + return; + } + z->code_buffer |= (unsigned int) stbi__zget8(z) << z->num_bits; + z->num_bits += 8; + } while (z->num_bits <= 24); +} + +stbi_inline static unsigned int stbi__zreceive(stbi__zbuf *z, int n) +{ + unsigned int k; + if (z->num_bits < n) stbi__fill_bits(z); + k = z->code_buffer & ((1 << n) - 1); + z->code_buffer >>= n; + z->num_bits -= n; + return k; +} + +static int stbi__zhuffman_decode_slowpath(stbi__zbuf *a, stbi__zhuffman *z) +{ + int b,s,k; + // not resolved by fast table, so compute it the slow way + // use jpeg approach, which requires MSbits at top + k = stbi__bit_reverse(a->code_buffer, 16); + for (s=STBI__ZFAST_BITS+1; ; ++s) + if (k < z->maxcode[s]) + break; + if (s >= 16) return -1; // invalid code! + // code size is s, so: + b = (k >> (16-s)) - z->firstcode[s] + z->firstsymbol[s]; + if (b >= STBI__ZNSYMS) return -1; // some data was corrupt somewhere! + if (z->size[b] != s) return -1; // was originally an assert, but report failure instead. + a->code_buffer >>= s; + a->num_bits -= s; + return z->value[b]; +} + +stbi_inline static int stbi__zhuffman_decode(stbi__zbuf *a, stbi__zhuffman *z) +{ + int b,s; + if (a->num_bits < 16) { + if (stbi__zeof(a)) { + return -1; /* report error for unexpected end of data. */ + } + stbi__fill_bits(a); + } + b = z->fast[a->code_buffer & STBI__ZFAST_MASK]; + if (b) { + s = b >> 9; + a->code_buffer >>= s; + a->num_bits -= s; + return b & 511; + } + return stbi__zhuffman_decode_slowpath(a, z); +} + +static int stbi__zexpand(stbi__zbuf *z, char *zout, int n) // need to make room for n bytes +{ + char *q; + unsigned int cur, limit, old_limit; + z->zout = zout; + if (!z->z_expandable) return stbi__err("output buffer limit","Corrupt PNG"); + cur = (unsigned int) (z->zout - z->zout_start); + limit = old_limit = (unsigned) (z->zout_end - z->zout_start); + if (UINT_MAX - cur < (unsigned) n) return stbi__err("outofmem", "Out of memory"); + while (cur + n > limit) { + if(limit > UINT_MAX / 2) return stbi__err("outofmem", "Out of memory"); + limit *= 2; + } + q = (char *) STBI_REALLOC_SIZED(z->zout_start, old_limit, limit); + STBI_NOTUSED(old_limit); + if (q == NULL) return stbi__err("outofmem", "Out of memory"); + z->zout_start = q; + z->zout = q + cur; + z->zout_end = q + limit; + return 1; +} + +static const int stbi__zlength_base[31] = { + 3,4,5,6,7,8,9,10,11,13, + 15,17,19,23,27,31,35,43,51,59, + 67,83,99,115,131,163,195,227,258,0,0 }; + +static const int stbi__zlength_extra[31]= +{ 0,0,0,0,0,0,0,0,1,1,1,1,2,2,2,2,3,3,3,3,4,4,4,4,5,5,5,5,0,0,0 }; + +static const int stbi__zdist_base[32] = { 1,2,3,4,5,7,9,13,17,25,33,49,65,97,129,193, +257,385,513,769,1025,1537,2049,3073,4097,6145,8193,12289,16385,24577,0,0}; + +static const int stbi__zdist_extra[32] = +{ 0,0,0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,12,12,13,13}; + +static int stbi__parse_huffman_block(stbi__zbuf *a) +{ + char *zout = a->zout; + for(;;) { + int z = stbi__zhuffman_decode(a, &a->z_length); + if (z < 256) { + if (z < 0) return stbi__err("bad huffman code","Corrupt PNG"); // error in huffman codes + if (zout >= a->zout_end) { + if (!stbi__zexpand(a, zout, 1)) return 0; + zout = a->zout; + } + *zout++ = (char) z; + } else { + stbi_uc *p; + int len,dist; + if (z == 256) { + a->zout = zout; + return 1; + } + z -= 257; + len = stbi__zlength_base[z]; + if (stbi__zlength_extra[z]) len += stbi__zreceive(a, stbi__zlength_extra[z]); + z = stbi__zhuffman_decode(a, &a->z_distance); + if (z < 0) return stbi__err("bad huffman code","Corrupt PNG"); + dist = stbi__zdist_base[z]; + if (stbi__zdist_extra[z]) dist += stbi__zreceive(a, stbi__zdist_extra[z]); + if (zout - a->zout_start < dist) return stbi__err("bad dist","Corrupt PNG"); + if (zout + len > a->zout_end) { + if (!stbi__zexpand(a, zout, len)) return 0; + zout = a->zout; + } + p = (stbi_uc *) (zout - dist); + if (dist == 1) { // run of one byte; common in images. + stbi_uc v = *p; + if (len) { do *zout++ = v; while (--len); } + } else { + if (len) { do *zout++ = *p++; while (--len); } + } + } + } +} + +static int stbi__compute_huffman_codes(stbi__zbuf *a) +{ + static const stbi_uc length_dezigzag[19] = { 16,17,18,0,8,7,9,6,10,5,11,4,12,3,13,2,14,1,15 }; + stbi__zhuffman z_codelength; + stbi_uc lencodes[286+32+137];//padding for maximum single op + stbi_uc codelength_sizes[19]; + int i,n; + + int hlit = stbi__zreceive(a,5) + 257; + int hdist = stbi__zreceive(a,5) + 1; + int hclen = stbi__zreceive(a,4) + 4; + int ntot = hlit + hdist; + + memset(codelength_sizes, 0, sizeof(codelength_sizes)); + for (i=0; i < hclen; ++i) { + int s = stbi__zreceive(a,3); + codelength_sizes[length_dezigzag[i]] = (stbi_uc) s; + } + if (!stbi__zbuild_huffman(&z_codelength, codelength_sizes, 19)) return 0; + + n = 0; + while (n < ntot) { + int c = stbi__zhuffman_decode(a, &z_codelength); + if (c < 0 || c >= 19) return stbi__err("bad codelengths", "Corrupt PNG"); + if (c < 16) + lencodes[n++] = (stbi_uc) c; + else { + stbi_uc fill = 0; + if (c == 16) { + c = stbi__zreceive(a,2)+3; + if (n == 0) return stbi__err("bad codelengths", "Corrupt PNG"); + fill = lencodes[n-1]; + } else if (c == 17) { + c = stbi__zreceive(a,3)+3; + } else if (c == 18) { + c = stbi__zreceive(a,7)+11; + } else { + return stbi__err("bad codelengths", "Corrupt PNG"); + } + if (ntot - n < c) return stbi__err("bad codelengths", "Corrupt PNG"); + memset(lencodes+n, fill, c); + n += c; + } + } + if (n != ntot) return stbi__err("bad codelengths","Corrupt PNG"); + if (!stbi__zbuild_huffman(&a->z_length, lencodes, hlit)) return 0; + if (!stbi__zbuild_huffman(&a->z_distance, lencodes+hlit, hdist)) return 0; + return 1; +} + +static int stbi__parse_uncompressed_block(stbi__zbuf *a) +{ + stbi_uc header[4]; + int len,nlen,k; + if (a->num_bits & 7) + stbi__zreceive(a, a->num_bits & 7); // discard + // drain the bit-packed data into header + k = 0; + while (a->num_bits > 0) { + header[k++] = (stbi_uc) (a->code_buffer & 255); // suppress MSVC run-time check + a->code_buffer >>= 8; + a->num_bits -= 8; + } + if (a->num_bits < 0) return stbi__err("zlib corrupt","Corrupt PNG"); + // now fill header the normal way + while (k < 4) + header[k++] = stbi__zget8(a); + len = header[1] * 256 + header[0]; + nlen = header[3] * 256 + header[2]; + if (nlen != (len ^ 0xffff)) return stbi__err("zlib corrupt","Corrupt PNG"); + if (a->zbuffer + len > a->zbuffer_end) return stbi__err("read past buffer","Corrupt PNG"); + if (a->zout + len > a->zout_end) + if (!stbi__zexpand(a, a->zout, len)) return 0; + memcpy(a->zout, a->zbuffer, len); + a->zbuffer += len; + a->zout += len; + return 1; +} + +static int stbi__parse_zlib_header(stbi__zbuf *a) +{ + int cmf = stbi__zget8(a); + int cm = cmf & 15; + /* int cinfo = cmf >> 4; */ + int flg = stbi__zget8(a); + if (stbi__zeof(a)) return stbi__err("bad zlib header","Corrupt PNG"); // zlib spec + if ((cmf*256+flg) % 31 != 0) return stbi__err("bad zlib header","Corrupt PNG"); // zlib spec + if (flg & 32) return stbi__err("no preset dict","Corrupt PNG"); // preset dictionary not allowed in png + if (cm != 8) return stbi__err("bad compression","Corrupt PNG"); // DEFLATE required for png + // window = 1 << (8 + cinfo)... but who cares, we fully buffer output + return 1; +} + +static const stbi_uc stbi__zdefault_length[STBI__ZNSYMS] = +{ + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, + 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, + 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, + 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, + 7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7, 7,7,7,7,7,7,7,7,8,8,8,8,8,8,8,8 +}; +static const stbi_uc stbi__zdefault_distance[32] = +{ + 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 +}; +/* +Init algorithm: +{ + int i; // use <= to match clearly with spec + for (i=0; i <= 143; ++i) stbi__zdefault_length[i] = 8; + for ( ; i <= 255; ++i) stbi__zdefault_length[i] = 9; + for ( ; i <= 279; ++i) stbi__zdefault_length[i] = 7; + for ( ; i <= 287; ++i) stbi__zdefault_length[i] = 8; + + for (i=0; i <= 31; ++i) stbi__zdefault_distance[i] = 5; +} +*/ + +static int stbi__parse_zlib(stbi__zbuf *a, int parse_header) +{ + int final, type; + if (parse_header) + if (!stbi__parse_zlib_header(a)) return 0; + a->num_bits = 0; + a->code_buffer = 0; + do { + final = stbi__zreceive(a,1); + type = stbi__zreceive(a,2); + if (type == 0) { + if (!stbi__parse_uncompressed_block(a)) return 0; + } else if (type == 3) { + return 0; + } else { + if (type == 1) { + // use fixed code lengths + if (!stbi__zbuild_huffman(&a->z_length , stbi__zdefault_length , STBI__ZNSYMS)) return 0; + if (!stbi__zbuild_huffman(&a->z_distance, stbi__zdefault_distance, 32)) return 0; + } else { + if (!stbi__compute_huffman_codes(a)) return 0; + } + if (!stbi__parse_huffman_block(a)) return 0; + } + } while (!final); + return 1; +} + +static int stbi__do_zlib(stbi__zbuf *a, char *obuf, int olen, int exp, int parse_header) +{ + a->zout_start = obuf; + a->zout = obuf; + a->zout_end = obuf + olen; + a->z_expandable = exp; + + return stbi__parse_zlib(a, parse_header); +} + +STBIDEF char *stbi_zlib_decode_malloc_guesssize(const char *buffer, int len, int initial_size, int *outlen) +{ + stbi__zbuf a; + char *p = (char *) stbi__malloc(initial_size); + if (p == NULL) return NULL; + a.zbuffer = (stbi_uc *) buffer; + a.zbuffer_end = (stbi_uc *) buffer + len; + if (stbi__do_zlib(&a, p, initial_size, 1, 1)) { + if (outlen) *outlen = (int) (a.zout - a.zout_start); + return a.zout_start; + } else { + STBI_FREE(a.zout_start); + return NULL; + } +} + +STBIDEF char *stbi_zlib_decode_malloc(char const *buffer, int len, int *outlen) +{ + return stbi_zlib_decode_malloc_guesssize(buffer, len, 16384, outlen); +} + +STBIDEF char *stbi_zlib_decode_malloc_guesssize_headerflag(const char *buffer, int len, int initial_size, int *outlen, int parse_header) +{ + stbi__zbuf a; + char *p = (char *) stbi__malloc(initial_size); + if (p == NULL) return NULL; + a.zbuffer = (stbi_uc *) buffer; + a.zbuffer_end = (stbi_uc *) buffer + len; + if (stbi__do_zlib(&a, p, initial_size, 1, parse_header)) { + if (outlen) *outlen = (int) (a.zout - a.zout_start); + return a.zout_start; + } else { + STBI_FREE(a.zout_start); + return NULL; + } +} + +STBIDEF int stbi_zlib_decode_buffer(char *obuffer, int olen, char const *ibuffer, int ilen) +{ + stbi__zbuf a; + a.zbuffer = (stbi_uc *) ibuffer; + a.zbuffer_end = (stbi_uc *) ibuffer + ilen; + if (stbi__do_zlib(&a, obuffer, olen, 0, 1)) + return (int) (a.zout - a.zout_start); + else + return -1; +} + +STBIDEF char *stbi_zlib_decode_noheader_malloc(char const *buffer, int len, int *outlen) +{ + stbi__zbuf a; + char *p = (char *) stbi__malloc(16384); + if (p == NULL) return NULL; + a.zbuffer = (stbi_uc *) buffer; + a.zbuffer_end = (stbi_uc *) buffer+len; + if (stbi__do_zlib(&a, p, 16384, 1, 0)) { + if (outlen) *outlen = (int) (a.zout - a.zout_start); + return a.zout_start; + } else { + STBI_FREE(a.zout_start); + return NULL; + } +} + +STBIDEF int stbi_zlib_decode_noheader_buffer(char *obuffer, int olen, const char *ibuffer, int ilen) +{ + stbi__zbuf a; + a.zbuffer = (stbi_uc *) ibuffer; + a.zbuffer_end = (stbi_uc *) ibuffer + ilen; + if (stbi__do_zlib(&a, obuffer, olen, 0, 0)) + return (int) (a.zout - a.zout_start); + else + return -1; +} +#endif + +// public domain "baseline" PNG decoder v0.10 Sean Barrett 2006-11-18 +// simple implementation +// - only 8-bit samples +// - no CRC checking +// - allocates lots of intermediate memory +// - avoids problem of streaming data between subsystems +// - avoids explicit window management +// performance +// - uses stb_zlib, a PD zlib implementation with fast huffman decoding + +#ifndef STBI_NO_PNG +typedef struct +{ + stbi__uint32 length; + stbi__uint32 type; +} stbi__pngchunk; + +static stbi__pngchunk stbi__get_chunk_header(stbi__context *s) +{ + stbi__pngchunk c; + c.length = stbi__get32be(s); + c.type = stbi__get32be(s); + return c; +} + +static int stbi__check_png_header(stbi__context *s) +{ + static const stbi_uc png_sig[8] = { 137,80,78,71,13,10,26,10 }; + int i; + for (i=0; i < 8; ++i) + if (stbi__get8(s) != png_sig[i]) return stbi__err("bad png sig","Not a PNG"); + return 1; +} + +typedef struct +{ + stbi__context *s; + stbi_uc *idata, *expanded, *out; + int depth; +} stbi__png; + + +enum { + STBI__F_none=0, + STBI__F_sub=1, + STBI__F_up=2, + STBI__F_avg=3, + STBI__F_paeth=4, + // synthetic filters used for first scanline to avoid needing a dummy row of 0s + STBI__F_avg_first, + STBI__F_paeth_first +}; + +static stbi_uc first_row_filter[5] = +{ + STBI__F_none, + STBI__F_sub, + STBI__F_none, + STBI__F_avg_first, + STBI__F_paeth_first +}; + +static int stbi__paeth(int a, int b, int c) +{ + int p = a + b - c; + int pa = abs(p-a); + int pb = abs(p-b); + int pc = abs(p-c); + if (pa <= pb && pa <= pc) return a; + if (pb <= pc) return b; + return c; +} + +static const stbi_uc stbi__depth_scale_table[9] = { 0, 0xff, 0x55, 0, 0x11, 0,0,0, 0x01 }; + +// create the png data from post-deflated data +static int stbi__create_png_image_raw(stbi__png *a, stbi_uc *raw, stbi__uint32 raw_len, int out_n, stbi__uint32 x, stbi__uint32 y, int depth, int color) +{ + int bytes = (depth == 16? 2 : 1); + stbi__context *s = a->s; + stbi__uint32 i,j,stride = x*out_n*bytes; + stbi__uint32 img_len, img_width_bytes; + int k; + int img_n = s->img_n; // copy it into a local for later + + int output_bytes = out_n*bytes; + int filter_bytes = img_n*bytes; + int width = x; + + STBI_ASSERT(out_n == s->img_n || out_n == s->img_n+1); + a->out = (stbi_uc *) stbi__malloc_mad3(x, y, output_bytes, 0); // extra bytes to write off the end into + if (!a->out) return stbi__err("outofmem", "Out of memory"); + + if (!stbi__mad3sizes_valid(img_n, x, depth, 7)) return stbi__err("too large", "Corrupt PNG"); + img_width_bytes = (((img_n * x * depth) + 7) >> 3); + img_len = (img_width_bytes + 1) * y; + + // we used to check for exact match between raw_len and img_len on non-interlaced PNGs, + // but issue #276 reported a PNG in the wild that had extra data at the end (all zeros), + // so just check for raw_len < img_len always. + if (raw_len < img_len) return stbi__err("not enough pixels","Corrupt PNG"); + + for (j=0; j < y; ++j) { + stbi_uc *cur = a->out + stride*j; + stbi_uc *prior; + int filter = *raw++; + + if (filter > 4) + return stbi__err("invalid filter","Corrupt PNG"); + + if (depth < 8) { + if (img_width_bytes > x) return stbi__err("invalid width","Corrupt PNG"); + cur += x*out_n - img_width_bytes; // store output to the rightmost img_len bytes, so we can decode in place + filter_bytes = 1; + width = img_width_bytes; + } + prior = cur - stride; // bugfix: need to compute this after 'cur +=' computation above + + // if first row, use special filter that doesn't sample previous row + if (j == 0) filter = first_row_filter[filter]; + + // handle first byte explicitly + for (k=0; k < filter_bytes; ++k) { + switch (filter) { + case STBI__F_none : cur[k] = raw[k]; break; + case STBI__F_sub : cur[k] = raw[k]; break; + case STBI__F_up : cur[k] = STBI__BYTECAST(raw[k] + prior[k]); break; + case STBI__F_avg : cur[k] = STBI__BYTECAST(raw[k] + (prior[k]>>1)); break; + case STBI__F_paeth : cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(0,prior[k],0)); break; + case STBI__F_avg_first : cur[k] = raw[k]; break; + case STBI__F_paeth_first: cur[k] = raw[k]; break; + } + } + + if (depth == 8) { + if (img_n != out_n) + cur[img_n] = 255; // first pixel + raw += img_n; + cur += out_n; + prior += out_n; + } else if (depth == 16) { + if (img_n != out_n) { + cur[filter_bytes] = 255; // first pixel top byte + cur[filter_bytes+1] = 255; // first pixel bottom byte + } + raw += filter_bytes; + cur += output_bytes; + prior += output_bytes; + } else { + raw += 1; + cur += 1; + prior += 1; + } + + // this is a little gross, so that we don't switch per-pixel or per-component + if (depth < 8 || img_n == out_n) { + int nk = (width - 1)*filter_bytes; + #define STBI__CASE(f) \ + case f: \ + for (k=0; k < nk; ++k) + switch (filter) { + // "none" filter turns into a memcpy here; make that explicit. + case STBI__F_none: memcpy(cur, raw, nk); break; + STBI__CASE(STBI__F_sub) { cur[k] = STBI__BYTECAST(raw[k] + cur[k-filter_bytes]); } break; + STBI__CASE(STBI__F_up) { cur[k] = STBI__BYTECAST(raw[k] + prior[k]); } break; + STBI__CASE(STBI__F_avg) { cur[k] = STBI__BYTECAST(raw[k] + ((prior[k] + cur[k-filter_bytes])>>1)); } break; + STBI__CASE(STBI__F_paeth) { cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k-filter_bytes],prior[k],prior[k-filter_bytes])); } break; + STBI__CASE(STBI__F_avg_first) { cur[k] = STBI__BYTECAST(raw[k] + (cur[k-filter_bytes] >> 1)); } break; + STBI__CASE(STBI__F_paeth_first) { cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k-filter_bytes],0,0)); } break; + } + #undef STBI__CASE + raw += nk; + } else { + STBI_ASSERT(img_n+1 == out_n); + #define STBI__CASE(f) \ + case f: \ + for (i=x-1; i >= 1; --i, cur[filter_bytes]=255,raw+=filter_bytes,cur+=output_bytes,prior+=output_bytes) \ + for (k=0; k < filter_bytes; ++k) + switch (filter) { + STBI__CASE(STBI__F_none) { cur[k] = raw[k]; } break; + STBI__CASE(STBI__F_sub) { cur[k] = STBI__BYTECAST(raw[k] + cur[k- output_bytes]); } break; + STBI__CASE(STBI__F_up) { cur[k] = STBI__BYTECAST(raw[k] + prior[k]); } break; + STBI__CASE(STBI__F_avg) { cur[k] = STBI__BYTECAST(raw[k] + ((prior[k] + cur[k- output_bytes])>>1)); } break; + STBI__CASE(STBI__F_paeth) { cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k- output_bytes],prior[k],prior[k- output_bytes])); } break; + STBI__CASE(STBI__F_avg_first) { cur[k] = STBI__BYTECAST(raw[k] + (cur[k- output_bytes] >> 1)); } break; + STBI__CASE(STBI__F_paeth_first) { cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k- output_bytes],0,0)); } break; + } + #undef STBI__CASE + + // the loop above sets the high byte of the pixels' alpha, but for + // 16 bit png files we also need the low byte set. we'll do that here. + if (depth == 16) { + cur = a->out + stride*j; // start at the beginning of the row again + for (i=0; i < x; ++i,cur+=output_bytes) { + cur[filter_bytes+1] = 255; + } + } + } + } + + // we make a separate pass to expand bits to pixels; for performance, + // this could run two scanlines behind the above code, so it won't + // intefere with filtering but will still be in the cache. + if (depth < 8) { + for (j=0; j < y; ++j) { + stbi_uc *cur = a->out + stride*j; + stbi_uc *in = a->out + stride*j + x*out_n - img_width_bytes; + // unpack 1/2/4-bit into a 8-bit buffer. allows us to keep the common 8-bit path optimal at minimal cost for 1/2/4-bit + // png guarante byte alignment, if width is not multiple of 8/4/2 we'll decode dummy trailing data that will be skipped in the later loop + stbi_uc scale = (color == 0) ? stbi__depth_scale_table[depth] : 1; // scale grayscale values to 0..255 range + + // note that the final byte might overshoot and write more data than desired. + // we can allocate enough data that this never writes out of memory, but it + // could also overwrite the next scanline. can it overwrite non-empty data + // on the next scanline? yes, consider 1-pixel-wide scanlines with 1-bit-per-pixel. + // so we need to explicitly clamp the final ones + + if (depth == 4) { + for (k=x*img_n; k >= 2; k-=2, ++in) { + *cur++ = scale * ((*in >> 4) ); + *cur++ = scale * ((*in ) & 0x0f); + } + if (k > 0) *cur++ = scale * ((*in >> 4) ); + } else if (depth == 2) { + for (k=x*img_n; k >= 4; k-=4, ++in) { + *cur++ = scale * ((*in >> 6) ); + *cur++ = scale * ((*in >> 4) & 0x03); + *cur++ = scale * ((*in >> 2) & 0x03); + *cur++ = scale * ((*in ) & 0x03); + } + if (k > 0) *cur++ = scale * ((*in >> 6) ); + if (k > 1) *cur++ = scale * ((*in >> 4) & 0x03); + if (k > 2) *cur++ = scale * ((*in >> 2) & 0x03); + } else if (depth == 1) { + for (k=x*img_n; k >= 8; k-=8, ++in) { + *cur++ = scale * ((*in >> 7) ); + *cur++ = scale * ((*in >> 6) & 0x01); + *cur++ = scale * ((*in >> 5) & 0x01); + *cur++ = scale * ((*in >> 4) & 0x01); + *cur++ = scale * ((*in >> 3) & 0x01); + *cur++ = scale * ((*in >> 2) & 0x01); + *cur++ = scale * ((*in >> 1) & 0x01); + *cur++ = scale * ((*in ) & 0x01); + } + if (k > 0) *cur++ = scale * ((*in >> 7) ); + if (k > 1) *cur++ = scale * ((*in >> 6) & 0x01); + if (k > 2) *cur++ = scale * ((*in >> 5) & 0x01); + if (k > 3) *cur++ = scale * ((*in >> 4) & 0x01); + if (k > 4) *cur++ = scale * ((*in >> 3) & 0x01); + if (k > 5) *cur++ = scale * ((*in >> 2) & 0x01); + if (k > 6) *cur++ = scale * ((*in >> 1) & 0x01); + } + if (img_n != out_n) { + int q; + // insert alpha = 255 + cur = a->out + stride*j; + if (img_n == 1) { + for (q=x-1; q >= 0; --q) { + cur[q*2+1] = 255; + cur[q*2+0] = cur[q]; + } + } else { + STBI_ASSERT(img_n == 3); + for (q=x-1; q >= 0; --q) { + cur[q*4+3] = 255; + cur[q*4+2] = cur[q*3+2]; + cur[q*4+1] = cur[q*3+1]; + cur[q*4+0] = cur[q*3+0]; + } + } + } + } + } else if (depth == 16) { + // force the image data from big-endian to platform-native. + // this is done in a separate pass due to the decoding relying + // on the data being untouched, but could probably be done + // per-line during decode if care is taken. + stbi_uc *cur = a->out; + stbi__uint16 *cur16 = (stbi__uint16*)cur; + + for(i=0; i < x*y*out_n; ++i,cur16++,cur+=2) { + *cur16 = (cur[0] << 8) | cur[1]; + } + } + + return 1; +} + +static int stbi__create_png_image(stbi__png *a, stbi_uc *image_data, stbi__uint32 image_data_len, int out_n, int depth, int color, int interlaced) +{ + int bytes = (depth == 16 ? 2 : 1); + int out_bytes = out_n * bytes; + stbi_uc *final; + int p; + if (!interlaced) + return stbi__create_png_image_raw(a, image_data, image_data_len, out_n, a->s->img_x, a->s->img_y, depth, color); + + // de-interlacing + final = (stbi_uc *) stbi__malloc_mad3(a->s->img_x, a->s->img_y, out_bytes, 0); + if (!final) return stbi__err("outofmem", "Out of memory"); + for (p=0; p < 7; ++p) { + int xorig[] = { 0,4,0,2,0,1,0 }; + int yorig[] = { 0,0,4,0,2,0,1 }; + int xspc[] = { 8,8,4,4,2,2,1 }; + int yspc[] = { 8,8,8,4,4,2,2 }; + int i,j,x,y; + // pass1_x[4] = 0, pass1_x[5] = 1, pass1_x[12] = 1 + x = (a->s->img_x - xorig[p] + xspc[p]-1) / xspc[p]; + y = (a->s->img_y - yorig[p] + yspc[p]-1) / yspc[p]; + if (x && y) { + stbi__uint32 img_len = ((((a->s->img_n * x * depth) + 7) >> 3) + 1) * y; + if (!stbi__create_png_image_raw(a, image_data, image_data_len, out_n, x, y, depth, color)) { + STBI_FREE(final); + return 0; + } + for (j=0; j < y; ++j) { + for (i=0; i < x; ++i) { + int out_y = j*yspc[p]+yorig[p]; + int out_x = i*xspc[p]+xorig[p]; + memcpy(final + out_y*a->s->img_x*out_bytes + out_x*out_bytes, + a->out + (j*x+i)*out_bytes, out_bytes); + } + } + STBI_FREE(a->out); + image_data += img_len; + image_data_len -= img_len; + } + } + a->out = final; + + return 1; +} + +static int stbi__compute_transparency(stbi__png *z, stbi_uc tc[3], int out_n) +{ + stbi__context *s = z->s; + stbi__uint32 i, pixel_count = s->img_x * s->img_y; + stbi_uc *p = z->out; + + // compute color-based transparency, assuming we've + // already got 255 as the alpha value in the output + STBI_ASSERT(out_n == 2 || out_n == 4); + + if (out_n == 2) { + for (i=0; i < pixel_count; ++i) { + p[1] = (p[0] == tc[0] ? 0 : 255); + p += 2; + } + } else { + for (i=0; i < pixel_count; ++i) { + if (p[0] == tc[0] && p[1] == tc[1] && p[2] == tc[2]) + p[3] = 0; + p += 4; + } + } + return 1; +} + +static int stbi__compute_transparency16(stbi__png *z, stbi__uint16 tc[3], int out_n) +{ + stbi__context *s = z->s; + stbi__uint32 i, pixel_count = s->img_x * s->img_y; + stbi__uint16 *p = (stbi__uint16*) z->out; + + // compute color-based transparency, assuming we've + // already got 65535 as the alpha value in the output + STBI_ASSERT(out_n == 2 || out_n == 4); + + if (out_n == 2) { + for (i = 0; i < pixel_count; ++i) { + p[1] = (p[0] == tc[0] ? 0 : 65535); + p += 2; + } + } else { + for (i = 0; i < pixel_count; ++i) { + if (p[0] == tc[0] && p[1] == tc[1] && p[2] == tc[2]) + p[3] = 0; + p += 4; + } + } + return 1; +} + +static int stbi__expand_png_palette(stbi__png *a, stbi_uc *palette, int len, int pal_img_n) +{ + stbi__uint32 i, pixel_count = a->s->img_x * a->s->img_y; + stbi_uc *p, *temp_out, *orig = a->out; + + p = (stbi_uc *) stbi__malloc_mad2(pixel_count, pal_img_n, 0); + if (p == NULL) return stbi__err("outofmem", "Out of memory"); + + // between here and free(out) below, exitting would leak + temp_out = p; + + if (pal_img_n == 3) { + for (i=0; i < pixel_count; ++i) { + int n = orig[i]*4; + p[0] = palette[n ]; + p[1] = palette[n+1]; + p[2] = palette[n+2]; + p += 3; + } + } else { + for (i=0; i < pixel_count; ++i) { + int n = orig[i]*4; + p[0] = palette[n ]; + p[1] = palette[n+1]; + p[2] = palette[n+2]; + p[3] = palette[n+3]; + p += 4; + } + } + STBI_FREE(a->out); + a->out = temp_out; + + STBI_NOTUSED(len); + + return 1; +} + +static int stbi__unpremultiply_on_load_global = 0; +static int stbi__de_iphone_flag_global = 0; + +STBIDEF void stbi_set_unpremultiply_on_load(int flag_true_if_should_unpremultiply) +{ + stbi__unpremultiply_on_load_global = flag_true_if_should_unpremultiply; +} + +STBIDEF void stbi_convert_iphone_png_to_rgb(int flag_true_if_should_convert) +{ + stbi__de_iphone_flag_global = flag_true_if_should_convert; +} + +#ifndef STBI_THREAD_LOCAL +#define stbi__unpremultiply_on_load stbi__unpremultiply_on_load_global +#define stbi__de_iphone_flag stbi__de_iphone_flag_global +#else +static STBI_THREAD_LOCAL int stbi__unpremultiply_on_load_local, stbi__unpremultiply_on_load_set; +static STBI_THREAD_LOCAL int stbi__de_iphone_flag_local, stbi__de_iphone_flag_set; + +STBIDEF void stbi__unpremultiply_on_load_thread(int flag_true_if_should_unpremultiply) +{ + stbi__unpremultiply_on_load_local = flag_true_if_should_unpremultiply; + stbi__unpremultiply_on_load_set = 1; +} + +STBIDEF void stbi_convert_iphone_png_to_rgb_thread(int flag_true_if_should_convert) +{ + stbi__de_iphone_flag_local = flag_true_if_should_convert; + stbi__de_iphone_flag_set = 1; +} + +#define stbi__unpremultiply_on_load (stbi__unpremultiply_on_load_set \ + ? stbi__unpremultiply_on_load_local \ + : stbi__unpremultiply_on_load_global) +#define stbi__de_iphone_flag (stbi__de_iphone_flag_set \ + ? stbi__de_iphone_flag_local \ + : stbi__de_iphone_flag_global) +#endif // STBI_THREAD_LOCAL + +static void stbi__de_iphone(stbi__png *z) +{ + stbi__context *s = z->s; + stbi__uint32 i, pixel_count = s->img_x * s->img_y; + stbi_uc *p = z->out; + + if (s->img_out_n == 3) { // convert bgr to rgb + for (i=0; i < pixel_count; ++i) { + stbi_uc t = p[0]; + p[0] = p[2]; + p[2] = t; + p += 3; + } + } else { + STBI_ASSERT(s->img_out_n == 4); + if (stbi__unpremultiply_on_load) { + // convert bgr to rgb and unpremultiply + for (i=0; i < pixel_count; ++i) { + stbi_uc a = p[3]; + stbi_uc t = p[0]; + if (a) { + stbi_uc half = a / 2; + p[0] = (p[2] * 255 + half) / a; + p[1] = (p[1] * 255 + half) / a; + p[2] = ( t * 255 + half) / a; + } else { + p[0] = p[2]; + p[2] = t; + } + p += 4; + } + } else { + // convert bgr to rgb + for (i=0; i < pixel_count; ++i) { + stbi_uc t = p[0]; + p[0] = p[2]; + p[2] = t; + p += 4; + } + } + } +} + +#define STBI__PNG_TYPE(a,b,c,d) (((unsigned) (a) << 24) + ((unsigned) (b) << 16) + ((unsigned) (c) << 8) + (unsigned) (d)) + +static int stbi__parse_png_file(stbi__png *z, int scan, int req_comp) +{ + stbi_uc palette[1024], pal_img_n=0; + stbi_uc has_trans=0, tc[3]={0}; + stbi__uint16 tc16[3]; + stbi__uint32 ioff=0, idata_limit=0, i, pal_len=0; + int first=1,k,interlace=0, color=0, is_iphone=0; + stbi__context *s = z->s; + + z->expanded = NULL; + z->idata = NULL; + z->out = NULL; + + if (!stbi__check_png_header(s)) return 0; + + if (scan == STBI__SCAN_type) return 1; + + for (;;) { + stbi__pngchunk c = stbi__get_chunk_header(s); + switch (c.type) { + case STBI__PNG_TYPE('C','g','B','I'): + is_iphone = 1; + stbi__skip(s, c.length); + break; + case STBI__PNG_TYPE('I','H','D','R'): { + int comp,filter; + if (!first) return stbi__err("multiple IHDR","Corrupt PNG"); + first = 0; + if (c.length != 13) return stbi__err("bad IHDR len","Corrupt PNG"); + s->img_x = stbi__get32be(s); + s->img_y = stbi__get32be(s); + if (s->img_y > STBI_MAX_DIMENSIONS) return stbi__err("too large","Very large image (corrupt?)"); + if (s->img_x > STBI_MAX_DIMENSIONS) return stbi__err("too large","Very large image (corrupt?)"); + z->depth = stbi__get8(s); if (z->depth != 1 && z->depth != 2 && z->depth != 4 && z->depth != 8 && z->depth != 16) return stbi__err("1/2/4/8/16-bit only","PNG not supported: 1/2/4/8/16-bit only"); + color = stbi__get8(s); if (color > 6) return stbi__err("bad ctype","Corrupt PNG"); + if (color == 3 && z->depth == 16) return stbi__err("bad ctype","Corrupt PNG"); + if (color == 3) pal_img_n = 3; else if (color & 1) return stbi__err("bad ctype","Corrupt PNG"); + comp = stbi__get8(s); if (comp) return stbi__err("bad comp method","Corrupt PNG"); + filter= stbi__get8(s); if (filter) return stbi__err("bad filter method","Corrupt PNG"); + interlace = stbi__get8(s); if (interlace>1) return stbi__err("bad interlace method","Corrupt PNG"); + if (!s->img_x || !s->img_y) return stbi__err("0-pixel image","Corrupt PNG"); + if (!pal_img_n) { + s->img_n = (color & 2 ? 3 : 1) + (color & 4 ? 1 : 0); + if ((1 << 30) / s->img_x / s->img_n < s->img_y) return stbi__err("too large", "Image too large to decode"); + if (scan == STBI__SCAN_header) return 1; + } else { + // if paletted, then pal_n is our final components, and + // img_n is # components to decompress/filter. + s->img_n = 1; + if ((1 << 30) / s->img_x / 4 < s->img_y) return stbi__err("too large","Corrupt PNG"); + // if SCAN_header, have to scan to see if we have a tRNS + } + break; + } + + case STBI__PNG_TYPE('P','L','T','E'): { + if (first) return stbi__err("first not IHDR", "Corrupt PNG"); + if (c.length > 256*3) return stbi__err("invalid PLTE","Corrupt PNG"); + pal_len = c.length / 3; + if (pal_len * 3 != c.length) return stbi__err("invalid PLTE","Corrupt PNG"); + for (i=0; i < pal_len; ++i) { + palette[i*4+0] = stbi__get8(s); + palette[i*4+1] = stbi__get8(s); + palette[i*4+2] = stbi__get8(s); + palette[i*4+3] = 255; + } + break; + } + + case STBI__PNG_TYPE('t','R','N','S'): { + if (first) return stbi__err("first not IHDR", "Corrupt PNG"); + if (z->idata) return stbi__err("tRNS after IDAT","Corrupt PNG"); + if (pal_img_n) { + if (scan == STBI__SCAN_header) { s->img_n = 4; return 1; } + if (pal_len == 0) return stbi__err("tRNS before PLTE","Corrupt PNG"); + if (c.length > pal_len) return stbi__err("bad tRNS len","Corrupt PNG"); + pal_img_n = 4; + for (i=0; i < c.length; ++i) + palette[i*4+3] = stbi__get8(s); + } else { + if (!(s->img_n & 1)) return stbi__err("tRNS with alpha","Corrupt PNG"); + if (c.length != (stbi__uint32) s->img_n*2) return stbi__err("bad tRNS len","Corrupt PNG"); + has_trans = 1; + if (z->depth == 16) { + for (k = 0; k < s->img_n; ++k) tc16[k] = (stbi__uint16)stbi__get16be(s); // copy the values as-is + } else { + for (k = 0; k < s->img_n; ++k) tc[k] = (stbi_uc)(stbi__get16be(s) & 255) * stbi__depth_scale_table[z->depth]; // non 8-bit images will be larger + } + } + break; + } + + case STBI__PNG_TYPE('I','D','A','T'): { + if (first) return stbi__err("first not IHDR", "Corrupt PNG"); + if (pal_img_n && !pal_len) return stbi__err("no PLTE","Corrupt PNG"); + if (scan == STBI__SCAN_header) { s->img_n = pal_img_n; return 1; } + if ((int)(ioff + c.length) < (int)ioff) return 0; + if (ioff + c.length > idata_limit) { + stbi__uint32 idata_limit_old = idata_limit; + stbi_uc *p; + if (idata_limit == 0) idata_limit = c.length > 4096 ? c.length : 4096; + while (ioff + c.length > idata_limit) + idata_limit *= 2; + STBI_NOTUSED(idata_limit_old); + p = (stbi_uc *) STBI_REALLOC_SIZED(z->idata, idata_limit_old, idata_limit); if (p == NULL) return stbi__err("outofmem", "Out of memory"); + z->idata = p; + } + if (!stbi__getn(s, z->idata+ioff,c.length)) return stbi__err("outofdata","Corrupt PNG"); + ioff += c.length; + break; + } + + case STBI__PNG_TYPE('I','E','N','D'): { + stbi__uint32 raw_len, bpl; + if (first) return stbi__err("first not IHDR", "Corrupt PNG"); + if (scan != STBI__SCAN_load) return 1; + if (z->idata == NULL) return stbi__err("no IDAT","Corrupt PNG"); + // initial guess for decoded data size to avoid unnecessary reallocs + bpl = (s->img_x * z->depth + 7) / 8; // bytes per line, per component + raw_len = bpl * s->img_y * s->img_n /* pixels */ + s->img_y /* filter mode per row */; + z->expanded = (stbi_uc *) stbi_zlib_decode_malloc_guesssize_headerflag((char *) z->idata, ioff, raw_len, (int *) &raw_len, !is_iphone); + if (z->expanded == NULL) return 0; // zlib should set error + STBI_FREE(z->idata); z->idata = NULL; + if ((req_comp == s->img_n+1 && req_comp != 3 && !pal_img_n) || has_trans) + s->img_out_n = s->img_n+1; + else + s->img_out_n = s->img_n; + if (!stbi__create_png_image(z, z->expanded, raw_len, s->img_out_n, z->depth, color, interlace)) return 0; + if (has_trans) { + if (z->depth == 16) { + if (!stbi__compute_transparency16(z, tc16, s->img_out_n)) return 0; + } else { + if (!stbi__compute_transparency(z, tc, s->img_out_n)) return 0; + } + } + if (is_iphone && stbi__de_iphone_flag && s->img_out_n > 2) + stbi__de_iphone(z); + if (pal_img_n) { + // pal_img_n == 3 or 4 + s->img_n = pal_img_n; // record the actual colors we had + s->img_out_n = pal_img_n; + if (req_comp >= 3) s->img_out_n = req_comp; + if (!stbi__expand_png_palette(z, palette, pal_len, s->img_out_n)) + return 0; + } else if (has_trans) { + // non-paletted image with tRNS -> source image has (constant) alpha + ++s->img_n; + } + STBI_FREE(z->expanded); z->expanded = NULL; + // end of PNG chunk, read and skip CRC + stbi__get32be(s); + return 1; + } + + default: + // if critical, fail + if (first) return stbi__err("first not IHDR", "Corrupt PNG"); + if ((c.type & (1 << 29)) == 0) { + #ifndef STBI_NO_FAILURE_STRINGS + // not threadsafe + static char invalid_chunk[] = "XXXX PNG chunk not known"; + invalid_chunk[0] = STBI__BYTECAST(c.type >> 24); + invalid_chunk[1] = STBI__BYTECAST(c.type >> 16); + invalid_chunk[2] = STBI__BYTECAST(c.type >> 8); + invalid_chunk[3] = STBI__BYTECAST(c.type >> 0); + #endif + return stbi__err(invalid_chunk, "PNG not supported: unknown PNG chunk type"); + } + stbi__skip(s, c.length); + break; + } + // end of PNG chunk, read and skip CRC + stbi__get32be(s); + } +} + +static void *stbi__do_png(stbi__png *p, int *x, int *y, int *n, int req_comp, stbi__result_info *ri) +{ + void *result=NULL; + if (req_comp < 0 || req_comp > 4) return stbi__errpuc("bad req_comp", "Internal error"); + if (stbi__parse_png_file(p, STBI__SCAN_load, req_comp)) { + if (p->depth <= 8) + ri->bits_per_channel = 8; + else if (p->depth == 16) + ri->bits_per_channel = 16; + else + return stbi__errpuc("bad bits_per_channel", "PNG not supported: unsupported color depth"); + result = p->out; + p->out = NULL; + if (req_comp && req_comp != p->s->img_out_n) { + if (ri->bits_per_channel == 8) + result = stbi__convert_format((unsigned char *) result, p->s->img_out_n, req_comp, p->s->img_x, p->s->img_y); + else + result = stbi__convert_format16((stbi__uint16 *) result, p->s->img_out_n, req_comp, p->s->img_x, p->s->img_y); + p->s->img_out_n = req_comp; + if (result == NULL) return result; + } + *x = p->s->img_x; + *y = p->s->img_y; + if (n) *n = p->s->img_n; + } + STBI_FREE(p->out); p->out = NULL; + STBI_FREE(p->expanded); p->expanded = NULL; + STBI_FREE(p->idata); p->idata = NULL; + + return result; +} + +static void *stbi__png_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + stbi__png p; + p.s = s; + return stbi__do_png(&p, x,y,comp,req_comp, ri); +} + +static int stbi__png_test(stbi__context *s) +{ + int r; + r = stbi__check_png_header(s); + stbi__rewind(s); + return r; +} + +static int stbi__png_info_raw(stbi__png *p, int *x, int *y, int *comp) +{ + if (!stbi__parse_png_file(p, STBI__SCAN_header, 0)) { + stbi__rewind( p->s ); + return 0; + } + if (x) *x = p->s->img_x; + if (y) *y = p->s->img_y; + if (comp) *comp = p->s->img_n; + return 1; +} + +static int stbi__png_info(stbi__context *s, int *x, int *y, int *comp) +{ + stbi__png p; + p.s = s; + return stbi__png_info_raw(&p, x, y, comp); +} + +static int stbi__png_is16(stbi__context *s) +{ + stbi__png p; + p.s = s; + if (!stbi__png_info_raw(&p, NULL, NULL, NULL)) + return 0; + if (p.depth != 16) { + stbi__rewind(p.s); + return 0; + } + return 1; +} +#endif + +// Microsoft/Windows BMP image + +#ifndef STBI_NO_BMP +static int stbi__bmp_test_raw(stbi__context *s) +{ + int r; + int sz; + if (stbi__get8(s) != 'B') return 0; + if (stbi__get8(s) != 'M') return 0; + stbi__get32le(s); // discard filesize + stbi__get16le(s); // discard reserved + stbi__get16le(s); // discard reserved + stbi__get32le(s); // discard data offset + sz = stbi__get32le(s); + r = (sz == 12 || sz == 40 || sz == 56 || sz == 108 || sz == 124); + return r; +} + +static int stbi__bmp_test(stbi__context *s) +{ + int r = stbi__bmp_test_raw(s); + stbi__rewind(s); + return r; +} + + +// returns 0..31 for the highest set bit +static int stbi__high_bit(unsigned int z) +{ + int n=0; + if (z == 0) return -1; + if (z >= 0x10000) { n += 16; z >>= 16; } + if (z >= 0x00100) { n += 8; z >>= 8; } + if (z >= 0x00010) { n += 4; z >>= 4; } + if (z >= 0x00004) { n += 2; z >>= 2; } + if (z >= 0x00002) { n += 1;/* >>= 1;*/ } + return n; +} + +static int stbi__bitcount(unsigned int a) +{ + a = (a & 0x55555555) + ((a >> 1) & 0x55555555); // max 2 + a = (a & 0x33333333) + ((a >> 2) & 0x33333333); // max 4 + a = (a + (a >> 4)) & 0x0f0f0f0f; // max 8 per 4, now 8 bits + a = (a + (a >> 8)); // max 16 per 8 bits + a = (a + (a >> 16)); // max 32 per 8 bits + return a & 0xff; +} + +// extract an arbitrarily-aligned N-bit value (N=bits) +// from v, and then make it 8-bits long and fractionally +// extend it to full full range. +static int stbi__shiftsigned(unsigned int v, int shift, int bits) +{ + static unsigned int mul_table[9] = { + 0, + 0xff/*0b11111111*/, 0x55/*0b01010101*/, 0x49/*0b01001001*/, 0x11/*0b00010001*/, + 0x21/*0b00100001*/, 0x41/*0b01000001*/, 0x81/*0b10000001*/, 0x01/*0b00000001*/, + }; + static unsigned int shift_table[9] = { + 0, 0,0,1,0,2,4,6,0, + }; + if (shift < 0) + v <<= -shift; + else + v >>= shift; + STBI_ASSERT(v < 256); + v >>= (8-bits); + STBI_ASSERT(bits >= 0 && bits <= 8); + return (int) ((unsigned) v * mul_table[bits]) >> shift_table[bits]; +} + +typedef struct +{ + int bpp, offset, hsz; + unsigned int mr,mg,mb,ma, all_a; + int extra_read; +} stbi__bmp_data; + +static int stbi__bmp_set_mask_defaults(stbi__bmp_data *info, int compress) +{ + // BI_BITFIELDS specifies masks explicitly, don't override + if (compress == 3) + return 1; + + if (compress == 0) { + if (info->bpp == 16) { + info->mr = 31u << 10; + info->mg = 31u << 5; + info->mb = 31u << 0; + } else if (info->bpp == 32) { + info->mr = 0xffu << 16; + info->mg = 0xffu << 8; + info->mb = 0xffu << 0; + info->ma = 0xffu << 24; + info->all_a = 0; // if all_a is 0 at end, then we loaded alpha channel but it was all 0 + } else { + // otherwise, use defaults, which is all-0 + info->mr = info->mg = info->mb = info->ma = 0; + } + return 1; + } + return 0; // error +} + +static void *stbi__bmp_parse_header(stbi__context *s, stbi__bmp_data *info) +{ + int hsz; + if (stbi__get8(s) != 'B' || stbi__get8(s) != 'M') return stbi__errpuc("not BMP", "Corrupt BMP"); + stbi__get32le(s); // discard filesize + stbi__get16le(s); // discard reserved + stbi__get16le(s); // discard reserved + info->offset = stbi__get32le(s); + info->hsz = hsz = stbi__get32le(s); + info->mr = info->mg = info->mb = info->ma = 0; + info->extra_read = 14; + + if (info->offset < 0) return stbi__errpuc("bad BMP", "bad BMP"); + + if (hsz != 12 && hsz != 40 && hsz != 56 && hsz != 108 && hsz != 124) return stbi__errpuc("unknown BMP", "BMP type not supported: unknown"); + if (hsz == 12) { + s->img_x = stbi__get16le(s); + s->img_y = stbi__get16le(s); + } else { + s->img_x = stbi__get32le(s); + s->img_y = stbi__get32le(s); + } + if (stbi__get16le(s) != 1) return stbi__errpuc("bad BMP", "bad BMP"); + info->bpp = stbi__get16le(s); + if (hsz != 12) { + int compress = stbi__get32le(s); + if (compress == 1 || compress == 2) return stbi__errpuc("BMP RLE", "BMP type not supported: RLE"); + if (compress >= 4) return stbi__errpuc("BMP JPEG/PNG", "BMP type not supported: unsupported compression"); // this includes PNG/JPEG modes + if (compress == 3 && info->bpp != 16 && info->bpp != 32) return stbi__errpuc("bad BMP", "bad BMP"); // bitfields requires 16 or 32 bits/pixel + stbi__get32le(s); // discard sizeof + stbi__get32le(s); // discard hres + stbi__get32le(s); // discard vres + stbi__get32le(s); // discard colorsused + stbi__get32le(s); // discard max important + if (hsz == 40 || hsz == 56) { + if (hsz == 56) { + stbi__get32le(s); + stbi__get32le(s); + stbi__get32le(s); + stbi__get32le(s); + } + if (info->bpp == 16 || info->bpp == 32) { + if (compress == 0) { + stbi__bmp_set_mask_defaults(info, compress); + } else if (compress == 3) { + info->mr = stbi__get32le(s); + info->mg = stbi__get32le(s); + info->mb = stbi__get32le(s); + info->extra_read += 12; + // not documented, but generated by photoshop and handled by mspaint + if (info->mr == info->mg && info->mg == info->mb) { + // ?!?!? + return stbi__errpuc("bad BMP", "bad BMP"); + } + } else + return stbi__errpuc("bad BMP", "bad BMP"); + } + } else { + // V4/V5 header + int i; + if (hsz != 108 && hsz != 124) + return stbi__errpuc("bad BMP", "bad BMP"); + info->mr = stbi__get32le(s); + info->mg = stbi__get32le(s); + info->mb = stbi__get32le(s); + info->ma = stbi__get32le(s); + if (compress != 3) // override mr/mg/mb unless in BI_BITFIELDS mode, as per docs + stbi__bmp_set_mask_defaults(info, compress); + stbi__get32le(s); // discard color space + for (i=0; i < 12; ++i) + stbi__get32le(s); // discard color space parameters + if (hsz == 124) { + stbi__get32le(s); // discard rendering intent + stbi__get32le(s); // discard offset of profile data + stbi__get32le(s); // discard size of profile data + stbi__get32le(s); // discard reserved + } + } + } + return (void *) 1; +} + + +static void *stbi__bmp_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + stbi_uc *out; + unsigned int mr=0,mg=0,mb=0,ma=0, all_a; + stbi_uc pal[256][4]; + int psize=0,i,j,width; + int flip_vertically, pad, target; + stbi__bmp_data info; + STBI_NOTUSED(ri); + + info.all_a = 255; + if (stbi__bmp_parse_header(s, &info) == NULL) + return NULL; // error code already set + + flip_vertically = ((int) s->img_y) > 0; + s->img_y = abs((int) s->img_y); + + if (s->img_y > STBI_MAX_DIMENSIONS) return stbi__errpuc("too large","Very large image (corrupt?)"); + if (s->img_x > STBI_MAX_DIMENSIONS) return stbi__errpuc("too large","Very large image (corrupt?)"); + + mr = info.mr; + mg = info.mg; + mb = info.mb; + ma = info.ma; + all_a = info.all_a; + + if (info.hsz == 12) { + if (info.bpp < 24) + psize = (info.offset - info.extra_read - 24) / 3; + } else { + if (info.bpp < 16) + psize = (info.offset - info.extra_read - info.hsz) >> 2; + } + if (psize == 0) { + if (info.offset != s->callback_already_read + (s->img_buffer - s->img_buffer_original)) { + return stbi__errpuc("bad offset", "Corrupt BMP"); + } + } + + if (info.bpp == 24 && ma == 0xff000000) + s->img_n = 3; + else + s->img_n = ma ? 4 : 3; + if (req_comp && req_comp >= 3) // we can directly decode 3 or 4 + target = req_comp; + else + target = s->img_n; // if they want monochrome, we'll post-convert + + // sanity-check size + if (!stbi__mad3sizes_valid(target, s->img_x, s->img_y, 0)) + return stbi__errpuc("too large", "Corrupt BMP"); + + out = (stbi_uc *) stbi__malloc_mad3(target, s->img_x, s->img_y, 0); + if (!out) return stbi__errpuc("outofmem", "Out of memory"); + if (info.bpp < 16) { + int z=0; + if (psize == 0 || psize > 256) { STBI_FREE(out); return stbi__errpuc("invalid", "Corrupt BMP"); } + for (i=0; i < psize; ++i) { + pal[i][2] = stbi__get8(s); + pal[i][1] = stbi__get8(s); + pal[i][0] = stbi__get8(s); + if (info.hsz != 12) stbi__get8(s); + pal[i][3] = 255; + } + stbi__skip(s, info.offset - info.extra_read - info.hsz - psize * (info.hsz == 12 ? 3 : 4)); + if (info.bpp == 1) width = (s->img_x + 7) >> 3; + else if (info.bpp == 4) width = (s->img_x + 1) >> 1; + else if (info.bpp == 8) width = s->img_x; + else { STBI_FREE(out); return stbi__errpuc("bad bpp", "Corrupt BMP"); } + pad = (-width)&3; + if (info.bpp == 1) { + for (j=0; j < (int) s->img_y; ++j) { + int bit_offset = 7, v = stbi__get8(s); + for (i=0; i < (int) s->img_x; ++i) { + int color = (v>>bit_offset)&0x1; + out[z++] = pal[color][0]; + out[z++] = pal[color][1]; + out[z++] = pal[color][2]; + if (target == 4) out[z++] = 255; + if (i+1 == (int) s->img_x) break; + if((--bit_offset) < 0) { + bit_offset = 7; + v = stbi__get8(s); + } + } + stbi__skip(s, pad); + } + } else { + for (j=0; j < (int) s->img_y; ++j) { + for (i=0; i < (int) s->img_x; i += 2) { + int v=stbi__get8(s),v2=0; + if (info.bpp == 4) { + v2 = v & 15; + v >>= 4; + } + out[z++] = pal[v][0]; + out[z++] = pal[v][1]; + out[z++] = pal[v][2]; + if (target == 4) out[z++] = 255; + if (i+1 == (int) s->img_x) break; + v = (info.bpp == 8) ? stbi__get8(s) : v2; + out[z++] = pal[v][0]; + out[z++] = pal[v][1]; + out[z++] = pal[v][2]; + if (target == 4) out[z++] = 255; + } + stbi__skip(s, pad); + } + } + } else { + int rshift=0,gshift=0,bshift=0,ashift=0,rcount=0,gcount=0,bcount=0,acount=0; + int z = 0; + int easy=0; + stbi__skip(s, info.offset - info.extra_read - info.hsz); + if (info.bpp == 24) width = 3 * s->img_x; + else if (info.bpp == 16) width = 2*s->img_x; + else /* bpp = 32 and pad = 0 */ width=0; + pad = (-width) & 3; + if (info.bpp == 24) { + easy = 1; + } else if (info.bpp == 32) { + if (mb == 0xff && mg == 0xff00 && mr == 0x00ff0000 && ma == 0xff000000) + easy = 2; + } + if (!easy) { + if (!mr || !mg || !mb) { STBI_FREE(out); return stbi__errpuc("bad masks", "Corrupt BMP"); } + // right shift amt to put high bit in position #7 + rshift = stbi__high_bit(mr)-7; rcount = stbi__bitcount(mr); + gshift = stbi__high_bit(mg)-7; gcount = stbi__bitcount(mg); + bshift = stbi__high_bit(mb)-7; bcount = stbi__bitcount(mb); + ashift = stbi__high_bit(ma)-7; acount = stbi__bitcount(ma); + if (rcount > 8 || gcount > 8 || bcount > 8 || acount > 8) { STBI_FREE(out); return stbi__errpuc("bad masks", "Corrupt BMP"); } + } + for (j=0; j < (int) s->img_y; ++j) { + if (easy) { + for (i=0; i < (int) s->img_x; ++i) { + unsigned char a; + out[z+2] = stbi__get8(s); + out[z+1] = stbi__get8(s); + out[z+0] = stbi__get8(s); + z += 3; + a = (easy == 2 ? stbi__get8(s) : 255); + all_a |= a; + if (target == 4) out[z++] = a; + } + } else { + int bpp = info.bpp; + for (i=0; i < (int) s->img_x; ++i) { + stbi__uint32 v = (bpp == 16 ? (stbi__uint32) stbi__get16le(s) : stbi__get32le(s)); + unsigned int a; + out[z++] = STBI__BYTECAST(stbi__shiftsigned(v & mr, rshift, rcount)); + out[z++] = STBI__BYTECAST(stbi__shiftsigned(v & mg, gshift, gcount)); + out[z++] = STBI__BYTECAST(stbi__shiftsigned(v & mb, bshift, bcount)); + a = (ma ? stbi__shiftsigned(v & ma, ashift, acount) : 255); + all_a |= a; + if (target == 4) out[z++] = STBI__BYTECAST(a); + } + } + stbi__skip(s, pad); + } + } + + // if alpha channel is all 0s, replace with all 255s + if (target == 4 && all_a == 0) + for (i=4*s->img_x*s->img_y-1; i >= 0; i -= 4) + out[i] = 255; + + if (flip_vertically) { + stbi_uc t; + for (j=0; j < (int) s->img_y>>1; ++j) { + stbi_uc *p1 = out + j *s->img_x*target; + stbi_uc *p2 = out + (s->img_y-1-j)*s->img_x*target; + for (i=0; i < (int) s->img_x*target; ++i) { + t = p1[i]; p1[i] = p2[i]; p2[i] = t; + } + } + } + + if (req_comp && req_comp != target) { + out = stbi__convert_format(out, target, req_comp, s->img_x, s->img_y); + if (out == NULL) return out; // stbi__convert_format frees input on failure + } + + *x = s->img_x; + *y = s->img_y; + if (comp) *comp = s->img_n; + return out; +} +#endif + +// Targa Truevision - TGA +// by Jonathan Dummer +#ifndef STBI_NO_TGA +// returns STBI_rgb or whatever, 0 on error +static int stbi__tga_get_comp(int bits_per_pixel, int is_grey, int* is_rgb16) +{ + // only RGB or RGBA (incl. 16bit) or grey allowed + if (is_rgb16) *is_rgb16 = 0; + switch(bits_per_pixel) { + case 8: return STBI_grey; + case 16: if(is_grey) return STBI_grey_alpha; + // fallthrough + case 15: if(is_rgb16) *is_rgb16 = 1; + return STBI_rgb; + case 24: // fallthrough + case 32: return bits_per_pixel/8; + default: return 0; + } +} + +static int stbi__tga_info(stbi__context *s, int *x, int *y, int *comp) +{ + int tga_w, tga_h, tga_comp, tga_image_type, tga_bits_per_pixel, tga_colormap_bpp; + int sz, tga_colormap_type; + stbi__get8(s); // discard Offset + tga_colormap_type = stbi__get8(s); // colormap type + if( tga_colormap_type > 1 ) { + stbi__rewind(s); + return 0; // only RGB or indexed allowed + } + tga_image_type = stbi__get8(s); // image type + if ( tga_colormap_type == 1 ) { // colormapped (paletted) image + if (tga_image_type != 1 && tga_image_type != 9) { + stbi__rewind(s); + return 0; + } + stbi__skip(s,4); // skip index of first colormap entry and number of entries + sz = stbi__get8(s); // check bits per palette color entry + if ( (sz != 8) && (sz != 15) && (sz != 16) && (sz != 24) && (sz != 32) ) { + stbi__rewind(s); + return 0; + } + stbi__skip(s,4); // skip image x and y origin + tga_colormap_bpp = sz; + } else { // "normal" image w/o colormap - only RGB or grey allowed, +/- RLE + if ( (tga_image_type != 2) && (tga_image_type != 3) && (tga_image_type != 10) && (tga_image_type != 11) ) { + stbi__rewind(s); + return 0; // only RGB or grey allowed, +/- RLE + } + stbi__skip(s,9); // skip colormap specification and image x/y origin + tga_colormap_bpp = 0; + } + tga_w = stbi__get16le(s); + if( tga_w < 1 ) { + stbi__rewind(s); + return 0; // test width + } + tga_h = stbi__get16le(s); + if( tga_h < 1 ) { + stbi__rewind(s); + return 0; // test height + } + tga_bits_per_pixel = stbi__get8(s); // bits per pixel + stbi__get8(s); // ignore alpha bits + if (tga_colormap_bpp != 0) { + if((tga_bits_per_pixel != 8) && (tga_bits_per_pixel != 16)) { + // when using a colormap, tga_bits_per_pixel is the size of the indexes + // I don't think anything but 8 or 16bit indexes makes sense + stbi__rewind(s); + return 0; + } + tga_comp = stbi__tga_get_comp(tga_colormap_bpp, 0, NULL); + } else { + tga_comp = stbi__tga_get_comp(tga_bits_per_pixel, (tga_image_type == 3) || (tga_image_type == 11), NULL); + } + if(!tga_comp) { + stbi__rewind(s); + return 0; + } + if (x) *x = tga_w; + if (y) *y = tga_h; + if (comp) *comp = tga_comp; + return 1; // seems to have passed everything +} + +static int stbi__tga_test(stbi__context *s) +{ + int res = 0; + int sz, tga_color_type; + stbi__get8(s); // discard Offset + tga_color_type = stbi__get8(s); // color type + if ( tga_color_type > 1 ) goto errorEnd; // only RGB or indexed allowed + sz = stbi__get8(s); // image type + if ( tga_color_type == 1 ) { // colormapped (paletted) image + if (sz != 1 && sz != 9) goto errorEnd; // colortype 1 demands image type 1 or 9 + stbi__skip(s,4); // skip index of first colormap entry and number of entries + sz = stbi__get8(s); // check bits per palette color entry + if ( (sz != 8) && (sz != 15) && (sz != 16) && (sz != 24) && (sz != 32) ) goto errorEnd; + stbi__skip(s,4); // skip image x and y origin + } else { // "normal" image w/o colormap + if ( (sz != 2) && (sz != 3) && (sz != 10) && (sz != 11) ) goto errorEnd; // only RGB or grey allowed, +/- RLE + stbi__skip(s,9); // skip colormap specification and image x/y origin + } + if ( stbi__get16le(s) < 1 ) goto errorEnd; // test width + if ( stbi__get16le(s) < 1 ) goto errorEnd; // test height + sz = stbi__get8(s); // bits per pixel + if ( (tga_color_type == 1) && (sz != 8) && (sz != 16) ) goto errorEnd; // for colormapped images, bpp is size of an index + if ( (sz != 8) && (sz != 15) && (sz != 16) && (sz != 24) && (sz != 32) ) goto errorEnd; + + res = 1; // if we got this far, everything's good and we can return 1 instead of 0 + +errorEnd: + stbi__rewind(s); + return res; +} + +// read 16bit value and convert to 24bit RGB +static void stbi__tga_read_rgb16(stbi__context *s, stbi_uc* out) +{ + stbi__uint16 px = (stbi__uint16)stbi__get16le(s); + stbi__uint16 fiveBitMask = 31; + // we have 3 channels with 5bits each + int r = (px >> 10) & fiveBitMask; + int g = (px >> 5) & fiveBitMask; + int b = px & fiveBitMask; + // Note that this saves the data in RGB(A) order, so it doesn't need to be swapped later + out[0] = (stbi_uc)((r * 255)/31); + out[1] = (stbi_uc)((g * 255)/31); + out[2] = (stbi_uc)((b * 255)/31); + + // some people claim that the most significant bit might be used for alpha + // (possibly if an alpha-bit is set in the "image descriptor byte") + // but that only made 16bit test images completely translucent.. + // so let's treat all 15 and 16bit TGAs as RGB with no alpha. +} + +static void *stbi__tga_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + // read in the TGA header stuff + int tga_offset = stbi__get8(s); + int tga_indexed = stbi__get8(s); + int tga_image_type = stbi__get8(s); + int tga_is_RLE = 0; + int tga_palette_start = stbi__get16le(s); + int tga_palette_len = stbi__get16le(s); + int tga_palette_bits = stbi__get8(s); + int tga_x_origin = stbi__get16le(s); + int tga_y_origin = stbi__get16le(s); + int tga_width = stbi__get16le(s); + int tga_height = stbi__get16le(s); + int tga_bits_per_pixel = stbi__get8(s); + int tga_comp, tga_rgb16=0; + int tga_inverted = stbi__get8(s); + // int tga_alpha_bits = tga_inverted & 15; // the 4 lowest bits - unused (useless?) + // image data + unsigned char *tga_data; + unsigned char *tga_palette = NULL; + int i, j; + unsigned char raw_data[4] = {0}; + int RLE_count = 0; + int RLE_repeating = 0; + int read_next_pixel = 1; + STBI_NOTUSED(ri); + STBI_NOTUSED(tga_x_origin); // @TODO + STBI_NOTUSED(tga_y_origin); // @TODO + + if (tga_height > STBI_MAX_DIMENSIONS) return stbi__errpuc("too large","Very large image (corrupt?)"); + if (tga_width > STBI_MAX_DIMENSIONS) return stbi__errpuc("too large","Very large image (corrupt?)"); + + // do a tiny bit of precessing + if ( tga_image_type >= 8 ) + { + tga_image_type -= 8; + tga_is_RLE = 1; + } + tga_inverted = 1 - ((tga_inverted >> 5) & 1); + + // If I'm paletted, then I'll use the number of bits from the palette + if ( tga_indexed ) tga_comp = stbi__tga_get_comp(tga_palette_bits, 0, &tga_rgb16); + else tga_comp = stbi__tga_get_comp(tga_bits_per_pixel, (tga_image_type == 3), &tga_rgb16); + + if(!tga_comp) // shouldn't really happen, stbi__tga_test() should have ensured basic consistency + return stbi__errpuc("bad format", "Can't find out TGA pixelformat"); + + // tga info + *x = tga_width; + *y = tga_height; + if (comp) *comp = tga_comp; + + if (!stbi__mad3sizes_valid(tga_width, tga_height, tga_comp, 0)) + return stbi__errpuc("too large", "Corrupt TGA"); + + tga_data = (unsigned char*)stbi__malloc_mad3(tga_width, tga_height, tga_comp, 0); + if (!tga_data) return stbi__errpuc("outofmem", "Out of memory"); + + // skip to the data's starting position (offset usually = 0) + stbi__skip(s, tga_offset ); + + if ( !tga_indexed && !tga_is_RLE && !tga_rgb16 ) { + for (i=0; i < tga_height; ++i) { + int row = tga_inverted ? tga_height -i - 1 : i; + stbi_uc *tga_row = tga_data + row*tga_width*tga_comp; + stbi__getn(s, tga_row, tga_width * tga_comp); + } + } else { + // do I need to load a palette? + if ( tga_indexed) + { + if (tga_palette_len == 0) { /* you have to have at least one entry! */ + STBI_FREE(tga_data); + return stbi__errpuc("bad palette", "Corrupt TGA"); + } + + // any data to skip? (offset usually = 0) + stbi__skip(s, tga_palette_start ); + // load the palette + tga_palette = (unsigned char*)stbi__malloc_mad2(tga_palette_len, tga_comp, 0); + if (!tga_palette) { + STBI_FREE(tga_data); + return stbi__errpuc("outofmem", "Out of memory"); + } + if (tga_rgb16) { + stbi_uc *pal_entry = tga_palette; + STBI_ASSERT(tga_comp == STBI_rgb); + for (i=0; i < tga_palette_len; ++i) { + stbi__tga_read_rgb16(s, pal_entry); + pal_entry += tga_comp; + } + } else if (!stbi__getn(s, tga_palette, tga_palette_len * tga_comp)) { + STBI_FREE(tga_data); + STBI_FREE(tga_palette); + return stbi__errpuc("bad palette", "Corrupt TGA"); + } + } + // load the data + for (i=0; i < tga_width * tga_height; ++i) + { + // if I'm in RLE mode, do I need to get a RLE stbi__pngchunk? + if ( tga_is_RLE ) + { + if ( RLE_count == 0 ) + { + // yep, get the next byte as a RLE command + int RLE_cmd = stbi__get8(s); + RLE_count = 1 + (RLE_cmd & 127); + RLE_repeating = RLE_cmd >> 7; + read_next_pixel = 1; + } else if ( !RLE_repeating ) + { + read_next_pixel = 1; + } + } else + { + read_next_pixel = 1; + } + // OK, if I need to read a pixel, do it now + if ( read_next_pixel ) + { + // load however much data we did have + if ( tga_indexed ) + { + // read in index, then perform the lookup + int pal_idx = (tga_bits_per_pixel == 8) ? stbi__get8(s) : stbi__get16le(s); + if ( pal_idx >= tga_palette_len ) { + // invalid index + pal_idx = 0; + } + pal_idx *= tga_comp; + for (j = 0; j < tga_comp; ++j) { + raw_data[j] = tga_palette[pal_idx+j]; + } + } else if(tga_rgb16) { + STBI_ASSERT(tga_comp == STBI_rgb); + stbi__tga_read_rgb16(s, raw_data); + } else { + // read in the data raw + for (j = 0; j < tga_comp; ++j) { + raw_data[j] = stbi__get8(s); + } + } + // clear the reading flag for the next pixel + read_next_pixel = 0; + } // end of reading a pixel + + // copy data + for (j = 0; j < tga_comp; ++j) + tga_data[i*tga_comp+j] = raw_data[j]; + + // in case we're in RLE mode, keep counting down + --RLE_count; + } + // do I need to invert the image? + if ( tga_inverted ) + { + for (j = 0; j*2 < tga_height; ++j) + { + int index1 = j * tga_width * tga_comp; + int index2 = (tga_height - 1 - j) * tga_width * tga_comp; + for (i = tga_width * tga_comp; i > 0; --i) + { + unsigned char temp = tga_data[index1]; + tga_data[index1] = tga_data[index2]; + tga_data[index2] = temp; + ++index1; + ++index2; + } + } + } + // clear my palette, if I had one + if ( tga_palette != NULL ) + { + STBI_FREE( tga_palette ); + } + } + + // swap RGB - if the source data was RGB16, it already is in the right order + if (tga_comp >= 3 && !tga_rgb16) + { + unsigned char* tga_pixel = tga_data; + for (i=0; i < tga_width * tga_height; ++i) + { + unsigned char temp = tga_pixel[0]; + tga_pixel[0] = tga_pixel[2]; + tga_pixel[2] = temp; + tga_pixel += tga_comp; + } + } + + // convert to target component count + if (req_comp && req_comp != tga_comp) + tga_data = stbi__convert_format(tga_data, tga_comp, req_comp, tga_width, tga_height); + + // the things I do to get rid of an error message, and yet keep + // Microsoft's C compilers happy... [8^( + tga_palette_start = tga_palette_len = tga_palette_bits = + tga_x_origin = tga_y_origin = 0; + STBI_NOTUSED(tga_palette_start); + // OK, done + return tga_data; +} +#endif + +// ************************************************************************************************* +// Photoshop PSD loader -- PD by Thatcher Ulrich, integration by Nicolas Schulz, tweaked by STB + +#ifndef STBI_NO_PSD +static int stbi__psd_test(stbi__context *s) +{ + int r = (stbi__get32be(s) == 0x38425053); + stbi__rewind(s); + return r; +} + +static int stbi__psd_decode_rle(stbi__context *s, stbi_uc *p, int pixelCount) +{ + int count, nleft, len; + + count = 0; + while ((nleft = pixelCount - count) > 0) { + len = stbi__get8(s); + if (len == 128) { + // No-op. + } else if (len < 128) { + // Copy next len+1 bytes literally. + len++; + if (len > nleft) return 0; // corrupt data + count += len; + while (len) { + *p = stbi__get8(s); + p += 4; + len--; + } + } else if (len > 128) { + stbi_uc val; + // Next -len+1 bytes in the dest are replicated from next source byte. + // (Interpret len as a negative 8-bit int.) + len = 257 - len; + if (len > nleft) return 0; // corrupt data + val = stbi__get8(s); + count += len; + while (len) { + *p = val; + p += 4; + len--; + } + } + } + + return 1; +} + +static void *stbi__psd_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri, int bpc) +{ + int pixelCount; + int channelCount, compression; + int channel, i; + int bitdepth; + int w,h; + stbi_uc *out; + STBI_NOTUSED(ri); + + // Check identifier + if (stbi__get32be(s) != 0x38425053) // "8BPS" + return stbi__errpuc("not PSD", "Corrupt PSD image"); + + // Check file type version. + if (stbi__get16be(s) != 1) + return stbi__errpuc("wrong version", "Unsupported version of PSD image"); + + // Skip 6 reserved bytes. + stbi__skip(s, 6 ); + + // Read the number of channels (R, G, B, A, etc). + channelCount = stbi__get16be(s); + if (channelCount < 0 || channelCount > 16) + return stbi__errpuc("wrong channel count", "Unsupported number of channels in PSD image"); + + // Read the rows and columns of the image. + h = stbi__get32be(s); + w = stbi__get32be(s); + + if (h > STBI_MAX_DIMENSIONS) return stbi__errpuc("too large","Very large image (corrupt?)"); + if (w > STBI_MAX_DIMENSIONS) return stbi__errpuc("too large","Very large image (corrupt?)"); + + // Make sure the depth is 8 bits. + bitdepth = stbi__get16be(s); + if (bitdepth != 8 && bitdepth != 16) + return stbi__errpuc("unsupported bit depth", "PSD bit depth is not 8 or 16 bit"); + + // Make sure the color mode is RGB. + // Valid options are: + // 0: Bitmap + // 1: Grayscale + // 2: Indexed color + // 3: RGB color + // 4: CMYK color + // 7: Multichannel + // 8: Duotone + // 9: Lab color + if (stbi__get16be(s) != 3) + return stbi__errpuc("wrong color format", "PSD is not in RGB color format"); + + // Skip the Mode Data. (It's the palette for indexed color; other info for other modes.) + stbi__skip(s,stbi__get32be(s) ); + + // Skip the image resources. (resolution, pen tool paths, etc) + stbi__skip(s, stbi__get32be(s) ); + + // Skip the reserved data. + stbi__skip(s, stbi__get32be(s) ); + + // Find out if the data is compressed. + // Known values: + // 0: no compression + // 1: RLE compressed + compression = stbi__get16be(s); + if (compression > 1) + return stbi__errpuc("bad compression", "PSD has an unknown compression format"); + + // Check size + if (!stbi__mad3sizes_valid(4, w, h, 0)) + return stbi__errpuc("too large", "Corrupt PSD"); + + // Create the destination image. + + if (!compression && bitdepth == 16 && bpc == 16) { + out = (stbi_uc *) stbi__malloc_mad3(8, w, h, 0); + ri->bits_per_channel = 16; + } else + out = (stbi_uc *) stbi__malloc(4 * w*h); + + if (!out) return stbi__errpuc("outofmem", "Out of memory"); + pixelCount = w*h; + + // Initialize the data to zero. + //memset( out, 0, pixelCount * 4 ); + + // Finally, the image data. + if (compression) { + // RLE as used by .PSD and .TIFF + // Loop until you get the number of unpacked bytes you are expecting: + // Read the next source byte into n. + // If n is between 0 and 127 inclusive, copy the next n+1 bytes literally. + // Else if n is between -127 and -1 inclusive, copy the next byte -n+1 times. + // Else if n is 128, noop. + // Endloop + + // The RLE-compressed data is preceded by a 2-byte data count for each row in the data, + // which we're going to just skip. + stbi__skip(s, h * channelCount * 2 ); + + // Read the RLE data by channel. + for (channel = 0; channel < 4; channel++) { + stbi_uc *p; + + p = out+channel; + if (channel >= channelCount) { + // Fill this channel with default data. + for (i = 0; i < pixelCount; i++, p += 4) + *p = (channel == 3 ? 255 : 0); + } else { + // Read the RLE data. + if (!stbi__psd_decode_rle(s, p, pixelCount)) { + STBI_FREE(out); + return stbi__errpuc("corrupt", "bad RLE data"); + } + } + } + + } else { + // We're at the raw image data. It's each channel in order (Red, Green, Blue, Alpha, ...) + // where each channel consists of an 8-bit (or 16-bit) value for each pixel in the image. + + // Read the data by channel. + for (channel = 0; channel < 4; channel++) { + if (channel >= channelCount) { + // Fill this channel with default data. + if (bitdepth == 16 && bpc == 16) { + stbi__uint16 *q = ((stbi__uint16 *) out) + channel; + stbi__uint16 val = channel == 3 ? 65535 : 0; + for (i = 0; i < pixelCount; i++, q += 4) + *q = val; + } else { + stbi_uc *p = out+channel; + stbi_uc val = channel == 3 ? 255 : 0; + for (i = 0; i < pixelCount; i++, p += 4) + *p = val; + } + } else { + if (ri->bits_per_channel == 16) { // output bpc + stbi__uint16 *q = ((stbi__uint16 *) out) + channel; + for (i = 0; i < pixelCount; i++, q += 4) + *q = (stbi__uint16) stbi__get16be(s); + } else { + stbi_uc *p = out+channel; + if (bitdepth == 16) { // input bpc + for (i = 0; i < pixelCount; i++, p += 4) + *p = (stbi_uc) (stbi__get16be(s) >> 8); + } else { + for (i = 0; i < pixelCount; i++, p += 4) + *p = stbi__get8(s); + } + } + } + } + } + + // remove weird white matte from PSD + if (channelCount >= 4) { + if (ri->bits_per_channel == 16) { + for (i=0; i < w*h; ++i) { + stbi__uint16 *pixel = (stbi__uint16 *) out + 4*i; + if (pixel[3] != 0 && pixel[3] != 65535) { + float a = pixel[3] / 65535.0f; + float ra = 1.0f / a; + float inv_a = 65535.0f * (1 - ra); + pixel[0] = (stbi__uint16) (pixel[0]*ra + inv_a); + pixel[1] = (stbi__uint16) (pixel[1]*ra + inv_a); + pixel[2] = (stbi__uint16) (pixel[2]*ra + inv_a); + } + } + } else { + for (i=0; i < w*h; ++i) { + unsigned char *pixel = out + 4*i; + if (pixel[3] != 0 && pixel[3] != 255) { + float a = pixel[3] / 255.0f; + float ra = 1.0f / a; + float inv_a = 255.0f * (1 - ra); + pixel[0] = (unsigned char) (pixel[0]*ra + inv_a); + pixel[1] = (unsigned char) (pixel[1]*ra + inv_a); + pixel[2] = (unsigned char) (pixel[2]*ra + inv_a); + } + } + } + } + + // convert to desired output format + if (req_comp && req_comp != 4) { + if (ri->bits_per_channel == 16) + out = (stbi_uc *) stbi__convert_format16((stbi__uint16 *) out, 4, req_comp, w, h); + else + out = stbi__convert_format(out, 4, req_comp, w, h); + if (out == NULL) return out; // stbi__convert_format frees input on failure + } + + if (comp) *comp = 4; + *y = h; + *x = w; + + return out; +} +#endif + +// ************************************************************************************************* +// Softimage PIC loader +// by Tom Seddon +// +// See http://softimage.wiki.softimage.com/index.php/INFO:_PIC_file_format +// See http://ozviz.wasp.uwa.edu.au/~pbourke/dataformats/softimagepic/ + +#ifndef STBI_NO_PIC +static int stbi__pic_is4(stbi__context *s,const char *str) +{ + int i; + for (i=0; i<4; ++i) + if (stbi__get8(s) != (stbi_uc)str[i]) + return 0; + + return 1; +} + +static int stbi__pic_test_core(stbi__context *s) +{ + int i; + + if (!stbi__pic_is4(s,"\x53\x80\xF6\x34")) + return 0; + + for(i=0;i<84;++i) + stbi__get8(s); + + if (!stbi__pic_is4(s,"PICT")) + return 0; + + return 1; +} + +typedef struct +{ + stbi_uc size,type,channel; +} stbi__pic_packet; + +static stbi_uc *stbi__readval(stbi__context *s, int channel, stbi_uc *dest) +{ + int mask=0x80, i; + + for (i=0; i<4; ++i, mask>>=1) { + if (channel & mask) { + if (stbi__at_eof(s)) return stbi__errpuc("bad file","PIC file too short"); + dest[i]=stbi__get8(s); + } + } + + return dest; +} + +static void stbi__copyval(int channel,stbi_uc *dest,const stbi_uc *src) +{ + int mask=0x80,i; + + for (i=0;i<4; ++i, mask>>=1) + if (channel&mask) + dest[i]=src[i]; +} + +static stbi_uc *stbi__pic_load_core(stbi__context *s,int width,int height,int *comp, stbi_uc *result) +{ + int act_comp=0,num_packets=0,y,chained; + stbi__pic_packet packets[10]; + + // this will (should...) cater for even some bizarre stuff like having data + // for the same channel in multiple packets. + do { + stbi__pic_packet *packet; + + if (num_packets==sizeof(packets)/sizeof(packets[0])) + return stbi__errpuc("bad format","too many packets"); + + packet = &packets[num_packets++]; + + chained = stbi__get8(s); + packet->size = stbi__get8(s); + packet->type = stbi__get8(s); + packet->channel = stbi__get8(s); + + act_comp |= packet->channel; + + if (stbi__at_eof(s)) return stbi__errpuc("bad file","file too short (reading packets)"); + if (packet->size != 8) return stbi__errpuc("bad format","packet isn't 8bpp"); + } while (chained); + + *comp = (act_comp & 0x10 ? 4 : 3); // has alpha channel? + + for(y=0; ytype) { + default: + return stbi__errpuc("bad format","packet has bad compression type"); + + case 0: {//uncompressed + int x; + + for(x=0;xchannel,dest)) + return 0; + break; + } + + case 1://Pure RLE + { + int left=width, i; + + while (left>0) { + stbi_uc count,value[4]; + + count=stbi__get8(s); + if (stbi__at_eof(s)) return stbi__errpuc("bad file","file too short (pure read count)"); + + if (count > left) + count = (stbi_uc) left; + + if (!stbi__readval(s,packet->channel,value)) return 0; + + for(i=0; ichannel,dest,value); + left -= count; + } + } + break; + + case 2: {//Mixed RLE + int left=width; + while (left>0) { + int count = stbi__get8(s), i; + if (stbi__at_eof(s)) return stbi__errpuc("bad file","file too short (mixed read count)"); + + if (count >= 128) { // Repeated + stbi_uc value[4]; + + if (count==128) + count = stbi__get16be(s); + else + count -= 127; + if (count > left) + return stbi__errpuc("bad file","scanline overrun"); + + if (!stbi__readval(s,packet->channel,value)) + return 0; + + for(i=0;ichannel,dest,value); + } else { // Raw + ++count; + if (count>left) return stbi__errpuc("bad file","scanline overrun"); + + for(i=0;ichannel,dest)) + return 0; + } + left-=count; + } + break; + } + } + } + } + + return result; +} + +static void *stbi__pic_load(stbi__context *s,int *px,int *py,int *comp,int req_comp, stbi__result_info *ri) +{ + stbi_uc *result; + int i, x,y, internal_comp; + STBI_NOTUSED(ri); + + if (!comp) comp = &internal_comp; + + for (i=0; i<92; ++i) + stbi__get8(s); + + x = stbi__get16be(s); + y = stbi__get16be(s); + + if (y > STBI_MAX_DIMENSIONS) return stbi__errpuc("too large","Very large image (corrupt?)"); + if (x > STBI_MAX_DIMENSIONS) return stbi__errpuc("too large","Very large image (corrupt?)"); + + if (stbi__at_eof(s)) return stbi__errpuc("bad file","file too short (pic header)"); + if (!stbi__mad3sizes_valid(x, y, 4, 0)) return stbi__errpuc("too large", "PIC image too large to decode"); + + stbi__get32be(s); //skip `ratio' + stbi__get16be(s); //skip `fields' + stbi__get16be(s); //skip `pad' + + // intermediate buffer is RGBA + result = (stbi_uc *) stbi__malloc_mad3(x, y, 4, 0); + if (!result) return stbi__errpuc("outofmem", "Out of memory"); + memset(result, 0xff, x*y*4); + + if (!stbi__pic_load_core(s,x,y,comp, result)) { + STBI_FREE(result); + result=0; + } + *px = x; + *py = y; + if (req_comp == 0) req_comp = *comp; + result=stbi__convert_format(result,4,req_comp,x,y); + + return result; +} + +static int stbi__pic_test(stbi__context *s) +{ + int r = stbi__pic_test_core(s); + stbi__rewind(s); + return r; +} +#endif + +// ************************************************************************************************* +// GIF loader -- public domain by Jean-Marc Lienher -- simplified/shrunk by stb + +#ifndef STBI_NO_GIF +typedef struct +{ + stbi__int16 prefix; + stbi_uc first; + stbi_uc suffix; +} stbi__gif_lzw; + +typedef struct +{ + int w,h; + stbi_uc *out; // output buffer (always 4 components) + stbi_uc *background; // The current "background" as far as a gif is concerned + stbi_uc *history; + int flags, bgindex, ratio, transparent, eflags; + stbi_uc pal[256][4]; + stbi_uc lpal[256][4]; + stbi__gif_lzw codes[8192]; + stbi_uc *color_table; + int parse, step; + int lflags; + int start_x, start_y; + int max_x, max_y; + int cur_x, cur_y; + int line_size; + int delay; +} stbi__gif; + +static int stbi__gif_test_raw(stbi__context *s) +{ + int sz; + if (stbi__get8(s) != 'G' || stbi__get8(s) != 'I' || stbi__get8(s) != 'F' || stbi__get8(s) != '8') return 0; + sz = stbi__get8(s); + if (sz != '9' && sz != '7') return 0; + if (stbi__get8(s) != 'a') return 0; + return 1; +} + +static int stbi__gif_test(stbi__context *s) +{ + int r = stbi__gif_test_raw(s); + stbi__rewind(s); + return r; +} + +static void stbi__gif_parse_colortable(stbi__context *s, stbi_uc pal[256][4], int num_entries, int transp) +{ + int i; + for (i=0; i < num_entries; ++i) { + pal[i][2] = stbi__get8(s); + pal[i][1] = stbi__get8(s); + pal[i][0] = stbi__get8(s); + pal[i][3] = transp == i ? 0 : 255; + } +} + +static int stbi__gif_header(stbi__context *s, stbi__gif *g, int *comp, int is_info) +{ + stbi_uc version; + if (stbi__get8(s) != 'G' || stbi__get8(s) != 'I' || stbi__get8(s) != 'F' || stbi__get8(s) != '8') + return stbi__err("not GIF", "Corrupt GIF"); + + version = stbi__get8(s); + if (version != '7' && version != '9') return stbi__err("not GIF", "Corrupt GIF"); + if (stbi__get8(s) != 'a') return stbi__err("not GIF", "Corrupt GIF"); + + stbi__g_failure_reason = ""; + g->w = stbi__get16le(s); + g->h = stbi__get16le(s); + g->flags = stbi__get8(s); + g->bgindex = stbi__get8(s); + g->ratio = stbi__get8(s); + g->transparent = -1; + + if (g->w > STBI_MAX_DIMENSIONS) return stbi__err("too large","Very large image (corrupt?)"); + if (g->h > STBI_MAX_DIMENSIONS) return stbi__err("too large","Very large image (corrupt?)"); + + if (comp != 0) *comp = 4; // can't actually tell whether it's 3 or 4 until we parse the comments + + if (is_info) return 1; + + if (g->flags & 0x80) + stbi__gif_parse_colortable(s,g->pal, 2 << (g->flags & 7), -1); + + return 1; +} + +static int stbi__gif_info_raw(stbi__context *s, int *x, int *y, int *comp) +{ + stbi__gif* g = (stbi__gif*) stbi__malloc(sizeof(stbi__gif)); + if (!g) return stbi__err("outofmem", "Out of memory"); + if (!stbi__gif_header(s, g, comp, 1)) { + STBI_FREE(g); + stbi__rewind( s ); + return 0; + } + if (x) *x = g->w; + if (y) *y = g->h; + STBI_FREE(g); + return 1; +} + +static void stbi__out_gif_code(stbi__gif *g, stbi__uint16 code) +{ + stbi_uc *p, *c; + int idx; + + // recurse to decode the prefixes, since the linked-list is backwards, + // and working backwards through an interleaved image would be nasty + if (g->codes[code].prefix >= 0) + stbi__out_gif_code(g, g->codes[code].prefix); + + if (g->cur_y >= g->max_y) return; + + idx = g->cur_x + g->cur_y; + p = &g->out[idx]; + g->history[idx / 4] = 1; + + c = &g->color_table[g->codes[code].suffix * 4]; + if (c[3] > 128) { // don't render transparent pixels; + p[0] = c[2]; + p[1] = c[1]; + p[2] = c[0]; + p[3] = c[3]; + } + g->cur_x += 4; + + if (g->cur_x >= g->max_x) { + g->cur_x = g->start_x; + g->cur_y += g->step; + + while (g->cur_y >= g->max_y && g->parse > 0) { + g->step = (1 << g->parse) * g->line_size; + g->cur_y = g->start_y + (g->step >> 1); + --g->parse; + } + } +} + +static stbi_uc *stbi__process_gif_raster(stbi__context *s, stbi__gif *g) +{ + stbi_uc lzw_cs; + stbi__int32 len, init_code; + stbi__uint32 first; + stbi__int32 codesize, codemask, avail, oldcode, bits, valid_bits, clear; + stbi__gif_lzw *p; + + lzw_cs = stbi__get8(s); + if (lzw_cs > 12) return NULL; + clear = 1 << lzw_cs; + first = 1; + codesize = lzw_cs + 1; + codemask = (1 << codesize) - 1; + bits = 0; + valid_bits = 0; + for (init_code = 0; init_code < clear; init_code++) { + g->codes[init_code].prefix = -1; + g->codes[init_code].first = (stbi_uc) init_code; + g->codes[init_code].suffix = (stbi_uc) init_code; + } + + // support no starting clear code + avail = clear+2; + oldcode = -1; + + len = 0; + for(;;) { + if (valid_bits < codesize) { + if (len == 0) { + len = stbi__get8(s); // start new block + if (len == 0) + return g->out; + } + --len; + bits |= (stbi__int32) stbi__get8(s) << valid_bits; + valid_bits += 8; + } else { + stbi__int32 code = bits & codemask; + bits >>= codesize; + valid_bits -= codesize; + // @OPTIMIZE: is there some way we can accelerate the non-clear path? + if (code == clear) { // clear code + codesize = lzw_cs + 1; + codemask = (1 << codesize) - 1; + avail = clear + 2; + oldcode = -1; + first = 0; + } else if (code == clear + 1) { // end of stream code + stbi__skip(s, len); + while ((len = stbi__get8(s)) > 0) + stbi__skip(s,len); + return g->out; + } else if (code <= avail) { + if (first) { + return stbi__errpuc("no clear code", "Corrupt GIF"); + } + + if (oldcode >= 0) { + p = &g->codes[avail++]; + if (avail > 8192) { + return stbi__errpuc("too many codes", "Corrupt GIF"); + } + + p->prefix = (stbi__int16) oldcode; + p->first = g->codes[oldcode].first; + p->suffix = (code == avail) ? p->first : g->codes[code].first; + } else if (code == avail) + return stbi__errpuc("illegal code in raster", "Corrupt GIF"); + + stbi__out_gif_code(g, (stbi__uint16) code); + + if ((avail & codemask) == 0 && avail <= 0x0FFF) { + codesize++; + codemask = (1 << codesize) - 1; + } + + oldcode = code; + } else { + return stbi__errpuc("illegal code in raster", "Corrupt GIF"); + } + } + } +} + +// this function is designed to support animated gifs, although stb_image doesn't support it +// two back is the image from two frames ago, used for a very specific disposal format +static stbi_uc *stbi__gif_load_next(stbi__context *s, stbi__gif *g, int *comp, int req_comp, stbi_uc *two_back) +{ + int dispose; + int first_frame; + int pi; + int pcount; + STBI_NOTUSED(req_comp); + + // on first frame, any non-written pixels get the background colour (non-transparent) + first_frame = 0; + if (g->out == 0) { + if (!stbi__gif_header(s, g, comp,0)) return 0; // stbi__g_failure_reason set by stbi__gif_header + if (!stbi__mad3sizes_valid(4, g->w, g->h, 0)) + return stbi__errpuc("too large", "GIF image is too large"); + pcount = g->w * g->h; + g->out = (stbi_uc *) stbi__malloc(4 * pcount); + g->background = (stbi_uc *) stbi__malloc(4 * pcount); + g->history = (stbi_uc *) stbi__malloc(pcount); + if (!g->out || !g->background || !g->history) + return stbi__errpuc("outofmem", "Out of memory"); + + // image is treated as "transparent" at the start - ie, nothing overwrites the current background; + // background colour is only used for pixels that are not rendered first frame, after that "background" + // color refers to the color that was there the previous frame. + memset(g->out, 0x00, 4 * pcount); + memset(g->background, 0x00, 4 * pcount); // state of the background (starts transparent) + memset(g->history, 0x00, pcount); // pixels that were affected previous frame + first_frame = 1; + } else { + // second frame - how do we dispose of the previous one? + dispose = (g->eflags & 0x1C) >> 2; + pcount = g->w * g->h; + + if ((dispose == 3) && (two_back == 0)) { + dispose = 2; // if I don't have an image to revert back to, default to the old background + } + + if (dispose == 3) { // use previous graphic + for (pi = 0; pi < pcount; ++pi) { + if (g->history[pi]) { + memcpy( &g->out[pi * 4], &two_back[pi * 4], 4 ); + } + } + } else if (dispose == 2) { + // restore what was changed last frame to background before that frame; + for (pi = 0; pi < pcount; ++pi) { + if (g->history[pi]) { + memcpy( &g->out[pi * 4], &g->background[pi * 4], 4 ); + } + } + } else { + // This is a non-disposal case eithe way, so just + // leave the pixels as is, and they will become the new background + // 1: do not dispose + // 0: not specified. + } + + // background is what out is after the undoing of the previou frame; + memcpy( g->background, g->out, 4 * g->w * g->h ); + } + + // clear my history; + memset( g->history, 0x00, g->w * g->h ); // pixels that were affected previous frame + + for (;;) { + int tag = stbi__get8(s); + switch (tag) { + case 0x2C: /* Image Descriptor */ + { + stbi__int32 x, y, w, h; + stbi_uc *o; + + x = stbi__get16le(s); + y = stbi__get16le(s); + w = stbi__get16le(s); + h = stbi__get16le(s); + if (((x + w) > (g->w)) || ((y + h) > (g->h))) + return stbi__errpuc("bad Image Descriptor", "Corrupt GIF"); + + g->line_size = g->w * 4; + g->start_x = x * 4; + g->start_y = y * g->line_size; + g->max_x = g->start_x + w * 4; + g->max_y = g->start_y + h * g->line_size; + g->cur_x = g->start_x; + g->cur_y = g->start_y; + + // if the width of the specified rectangle is 0, that means + // we may not see *any* pixels or the image is malformed; + // to make sure this is caught, move the current y down to + // max_y (which is what out_gif_code checks). + if (w == 0) + g->cur_y = g->max_y; + + g->lflags = stbi__get8(s); + + if (g->lflags & 0x40) { + g->step = 8 * g->line_size; // first interlaced spacing + g->parse = 3; + } else { + g->step = g->line_size; + g->parse = 0; + } + + if (g->lflags & 0x80) { + stbi__gif_parse_colortable(s,g->lpal, 2 << (g->lflags & 7), g->eflags & 0x01 ? g->transparent : -1); + g->color_table = (stbi_uc *) g->lpal; + } else if (g->flags & 0x80) { + g->color_table = (stbi_uc *) g->pal; + } else + return stbi__errpuc("missing color table", "Corrupt GIF"); + + o = stbi__process_gif_raster(s, g); + if (!o) return NULL; + + // if this was the first frame, + pcount = g->w * g->h; + if (first_frame && (g->bgindex > 0)) { + // if first frame, any pixel not drawn to gets the background color + for (pi = 0; pi < pcount; ++pi) { + if (g->history[pi] == 0) { + g->pal[g->bgindex][3] = 255; // just in case it was made transparent, undo that; It will be reset next frame if need be; + memcpy( &g->out[pi * 4], &g->pal[g->bgindex], 4 ); + } + } + } + + return o; + } + + case 0x21: // Comment Extension. + { + int len; + int ext = stbi__get8(s); + if (ext == 0xF9) { // Graphic Control Extension. + len = stbi__get8(s); + if (len == 4) { + g->eflags = stbi__get8(s); + g->delay = 10 * stbi__get16le(s); // delay - 1/100th of a second, saving as 1/1000ths. + + // unset old transparent + if (g->transparent >= 0) { + g->pal[g->transparent][3] = 255; + } + if (g->eflags & 0x01) { + g->transparent = stbi__get8(s); + if (g->transparent >= 0) { + g->pal[g->transparent][3] = 0; + } + } else { + // don't need transparent + stbi__skip(s, 1); + g->transparent = -1; + } + } else { + stbi__skip(s, len); + break; + } + } + while ((len = stbi__get8(s)) != 0) { + stbi__skip(s, len); + } + break; + } + + case 0x3B: // gif stream termination code + return (stbi_uc *) s; // using '1' causes warning on some compilers + + default: + return stbi__errpuc("unknown code", "Corrupt GIF"); + } + } +} + +static void *stbi__load_gif_main_outofmem(stbi__gif *g, stbi_uc *out, int **delays) +{ + STBI_FREE(g->out); + STBI_FREE(g->history); + STBI_FREE(g->background); + + if (out) STBI_FREE(out); + if (delays && *delays) STBI_FREE(*delays); + return stbi__errpuc("outofmem", "Out of memory"); +} + +static void *stbi__load_gif_main(stbi__context *s, int **delays, int *x, int *y, int *z, int *comp, int req_comp) +{ + if (stbi__gif_test(s)) { + int layers = 0; + stbi_uc *u = 0; + stbi_uc *out = 0; + stbi_uc *two_back = 0; + stbi__gif g; + int stride; + int out_size = 0; + int delays_size = 0; + + STBI_NOTUSED(out_size); + STBI_NOTUSED(delays_size); + + memset(&g, 0, sizeof(g)); + if (delays) { + *delays = 0; + } + + do { + u = stbi__gif_load_next(s, &g, comp, req_comp, two_back); + if (u == (stbi_uc *) s) u = 0; // end of animated gif marker + + if (u) { + *x = g.w; + *y = g.h; + ++layers; + stride = g.w * g.h * 4; + + if (out) { + void *tmp = (stbi_uc*) STBI_REALLOC_SIZED( out, out_size, layers * stride ); + if (!tmp) + return stbi__load_gif_main_outofmem(&g, out, delays); + else { + out = (stbi_uc*) tmp; + out_size = layers * stride; + } + + if (delays) { + int *new_delays = (int*) STBI_REALLOC_SIZED( *delays, delays_size, sizeof(int) * layers ); + if (!new_delays) + return stbi__load_gif_main_outofmem(&g, out, delays); + *delays = new_delays; + delays_size = layers * sizeof(int); + } + } else { + out = (stbi_uc*)stbi__malloc( layers * stride ); + if (!out) + return stbi__load_gif_main_outofmem(&g, out, delays); + out_size = layers * stride; + if (delays) { + *delays = (int*) stbi__malloc( layers * sizeof(int) ); + if (!*delays) + return stbi__load_gif_main_outofmem(&g, out, delays); + delays_size = layers * sizeof(int); + } + } + memcpy( out + ((layers - 1) * stride), u, stride ); + if (layers >= 2) { + two_back = out - 2 * stride; + } + + if (delays) { + (*delays)[layers - 1U] = g.delay; + } + } + } while (u != 0); + + // free temp buffer; + STBI_FREE(g.out); + STBI_FREE(g.history); + STBI_FREE(g.background); + + // do the final conversion after loading everything; + if (req_comp && req_comp != 4) + out = stbi__convert_format(out, 4, req_comp, layers * g.w, g.h); + + *z = layers; + return out; + } else { + return stbi__errpuc("not GIF", "Image was not as a gif type."); + } +} + +static void *stbi__gif_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + stbi_uc *u = 0; + stbi__gif g; + memset(&g, 0, sizeof(g)); + STBI_NOTUSED(ri); + + u = stbi__gif_load_next(s, &g, comp, req_comp, 0); + if (u == (stbi_uc *) s) u = 0; // end of animated gif marker + if (u) { + *x = g.w; + *y = g.h; + + // moved conversion to after successful load so that the same + // can be done for multiple frames. + if (req_comp && req_comp != 4) + u = stbi__convert_format(u, 4, req_comp, g.w, g.h); + } else if (g.out) { + // if there was an error and we allocated an image buffer, free it! + STBI_FREE(g.out); + } + + // free buffers needed for multiple frame loading; + STBI_FREE(g.history); + STBI_FREE(g.background); + + return u; +} + +static int stbi__gif_info(stbi__context *s, int *x, int *y, int *comp) +{ + return stbi__gif_info_raw(s,x,y,comp); +} +#endif + +// ************************************************************************************************* +// Radiance RGBE HDR loader +// originally by Nicolas Schulz +#ifndef STBI_NO_HDR +static int stbi__hdr_test_core(stbi__context *s, const char *signature) +{ + int i; + for (i=0; signature[i]; ++i) + if (stbi__get8(s) != signature[i]) + return 0; + stbi__rewind(s); + return 1; +} + +static int stbi__hdr_test(stbi__context* s) +{ + int r = stbi__hdr_test_core(s, "#?RADIANCE\n"); + stbi__rewind(s); + if(!r) { + r = stbi__hdr_test_core(s, "#?RGBE\n"); + stbi__rewind(s); + } + return r; +} + +#define STBI__HDR_BUFLEN 1024 +static char *stbi__hdr_gettoken(stbi__context *z, char *buffer) +{ + int len=0; + char c = '\0'; + + c = (char) stbi__get8(z); + + while (!stbi__at_eof(z) && c != '\n') { + buffer[len++] = c; + if (len == STBI__HDR_BUFLEN-1) { + // flush to end of line + while (!stbi__at_eof(z) && stbi__get8(z) != '\n') + ; + break; + } + c = (char) stbi__get8(z); + } + + buffer[len] = 0; + return buffer; +} + +static void stbi__hdr_convert(float *output, stbi_uc *input, int req_comp) +{ + if ( input[3] != 0 ) { + float f1; + // Exponent + f1 = (float) ldexp(1.0f, input[3] - (int)(128 + 8)); + if (req_comp <= 2) + output[0] = (input[0] + input[1] + input[2]) * f1 / 3; + else { + output[0] = input[0] * f1; + output[1] = input[1] * f1; + output[2] = input[2] * f1; + } + if (req_comp == 2) output[1] = 1; + if (req_comp == 4) output[3] = 1; + } else { + switch (req_comp) { + case 4: output[3] = 1; /* fallthrough */ + case 3: output[0] = output[1] = output[2] = 0; + break; + case 2: output[1] = 1; /* fallthrough */ + case 1: output[0] = 0; + break; + } + } +} + +static float *stbi__hdr_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + char buffer[STBI__HDR_BUFLEN]; + char *token; + int valid = 0; + int width, height; + stbi_uc *scanline; + float *hdr_data; + int len; + unsigned char count, value; + int i, j, k, c1,c2, z; + const char *headerToken; + STBI_NOTUSED(ri); + + // Check identifier + headerToken = stbi__hdr_gettoken(s,buffer); + if (strcmp(headerToken, "#?RADIANCE") != 0 && strcmp(headerToken, "#?RGBE") != 0) + return stbi__errpf("not HDR", "Corrupt HDR image"); + + // Parse header + for(;;) { + token = stbi__hdr_gettoken(s,buffer); + if (token[0] == 0) break; + if (strcmp(token, "FORMAT=32-bit_rle_rgbe") == 0) valid = 1; + } + + if (!valid) return stbi__errpf("unsupported format", "Unsupported HDR format"); + + // Parse width and height + // can't use sscanf() if we're not using stdio! + token = stbi__hdr_gettoken(s,buffer); + if (strncmp(token, "-Y ", 3)) return stbi__errpf("unsupported data layout", "Unsupported HDR format"); + token += 3; + height = (int) strtol(token, &token, 10); + while (*token == ' ') ++token; + if (strncmp(token, "+X ", 3)) return stbi__errpf("unsupported data layout", "Unsupported HDR format"); + token += 3; + width = (int) strtol(token, NULL, 10); + + if (height > STBI_MAX_DIMENSIONS) return stbi__errpf("too large","Very large image (corrupt?)"); + if (width > STBI_MAX_DIMENSIONS) return stbi__errpf("too large","Very large image (corrupt?)"); + + *x = width; + *y = height; + + if (comp) *comp = 3; + if (req_comp == 0) req_comp = 3; + + if (!stbi__mad4sizes_valid(width, height, req_comp, sizeof(float), 0)) + return stbi__errpf("too large", "HDR image is too large"); + + // Read data + hdr_data = (float *) stbi__malloc_mad4(width, height, req_comp, sizeof(float), 0); + if (!hdr_data) + return stbi__errpf("outofmem", "Out of memory"); + + // Load image data + // image data is stored as some number of sca + if ( width < 8 || width >= 32768) { + // Read flat data + for (j=0; j < height; ++j) { + for (i=0; i < width; ++i) { + stbi_uc rgbe[4]; + main_decode_loop: + stbi__getn(s, rgbe, 4); + stbi__hdr_convert(hdr_data + j * width * req_comp + i * req_comp, rgbe, req_comp); + } + } + } else { + // Read RLE-encoded data + scanline = NULL; + + for (j = 0; j < height; ++j) { + c1 = stbi__get8(s); + c2 = stbi__get8(s); + len = stbi__get8(s); + if (c1 != 2 || c2 != 2 || (len & 0x80)) { + // not run-length encoded, so we have to actually use THIS data as a decoded + // pixel (note this can't be a valid pixel--one of RGB must be >= 128) + stbi_uc rgbe[4]; + rgbe[0] = (stbi_uc) c1; + rgbe[1] = (stbi_uc) c2; + rgbe[2] = (stbi_uc) len; + rgbe[3] = (stbi_uc) stbi__get8(s); + stbi__hdr_convert(hdr_data, rgbe, req_comp); + i = 1; + j = 0; + STBI_FREE(scanline); + goto main_decode_loop; // yes, this makes no sense + } + len <<= 8; + len |= stbi__get8(s); + if (len != width) { STBI_FREE(hdr_data); STBI_FREE(scanline); return stbi__errpf("invalid decoded scanline length", "corrupt HDR"); } + if (scanline == NULL) { + scanline = (stbi_uc *) stbi__malloc_mad2(width, 4, 0); + if (!scanline) { + STBI_FREE(hdr_data); + return stbi__errpf("outofmem", "Out of memory"); + } + } + + for (k = 0; k < 4; ++k) { + int nleft; + i = 0; + while ((nleft = width - i) > 0) { + count = stbi__get8(s); + if (count > 128) { + // Run + value = stbi__get8(s); + count -= 128; + if (count > nleft) { STBI_FREE(hdr_data); STBI_FREE(scanline); return stbi__errpf("corrupt", "bad RLE data in HDR"); } + for (z = 0; z < count; ++z) + scanline[i++ * 4 + k] = value; + } else { + // Dump + if (count > nleft) { STBI_FREE(hdr_data); STBI_FREE(scanline); return stbi__errpf("corrupt", "bad RLE data in HDR"); } + for (z = 0; z < count; ++z) + scanline[i++ * 4 + k] = stbi__get8(s); + } + } + } + for (i=0; i < width; ++i) + stbi__hdr_convert(hdr_data+(j*width + i)*req_comp, scanline + i*4, req_comp); + } + if (scanline) + STBI_FREE(scanline); + } + + return hdr_data; +} + +static int stbi__hdr_info(stbi__context *s, int *x, int *y, int *comp) +{ + char buffer[STBI__HDR_BUFLEN]; + char *token; + int valid = 0; + int dummy; + + if (!x) x = &dummy; + if (!y) y = &dummy; + if (!comp) comp = &dummy; + + if (stbi__hdr_test(s) == 0) { + stbi__rewind( s ); + return 0; + } + + for(;;) { + token = stbi__hdr_gettoken(s,buffer); + if (token[0] == 0) break; + if (strcmp(token, "FORMAT=32-bit_rle_rgbe") == 0) valid = 1; + } + + if (!valid) { + stbi__rewind( s ); + return 0; + } + token = stbi__hdr_gettoken(s,buffer); + if (strncmp(token, "-Y ", 3)) { + stbi__rewind( s ); + return 0; + } + token += 3; + *y = (int) strtol(token, &token, 10); + while (*token == ' ') ++token; + if (strncmp(token, "+X ", 3)) { + stbi__rewind( s ); + return 0; + } + token += 3; + *x = (int) strtol(token, NULL, 10); + *comp = 3; + return 1; +} +#endif // STBI_NO_HDR + +#ifndef STBI_NO_BMP +static int stbi__bmp_info(stbi__context *s, int *x, int *y, int *comp) +{ + void *p; + stbi__bmp_data info; + + info.all_a = 255; + p = stbi__bmp_parse_header(s, &info); + if (p == NULL) { + stbi__rewind( s ); + return 0; + } + if (x) *x = s->img_x; + if (y) *y = s->img_y; + if (comp) { + if (info.bpp == 24 && info.ma == 0xff000000) + *comp = 3; + else + *comp = info.ma ? 4 : 3; + } + return 1; +} +#endif + +#ifndef STBI_NO_PSD +static int stbi__psd_info(stbi__context *s, int *x, int *y, int *comp) +{ + int channelCount, dummy, depth; + if (!x) x = &dummy; + if (!y) y = &dummy; + if (!comp) comp = &dummy; + if (stbi__get32be(s) != 0x38425053) { + stbi__rewind( s ); + return 0; + } + if (stbi__get16be(s) != 1) { + stbi__rewind( s ); + return 0; + } + stbi__skip(s, 6); + channelCount = stbi__get16be(s); + if (channelCount < 0 || channelCount > 16) { + stbi__rewind( s ); + return 0; + } + *y = stbi__get32be(s); + *x = stbi__get32be(s); + depth = stbi__get16be(s); + if (depth != 8 && depth != 16) { + stbi__rewind( s ); + return 0; + } + if (stbi__get16be(s) != 3) { + stbi__rewind( s ); + return 0; + } + *comp = 4; + return 1; +} + +static int stbi__psd_is16(stbi__context *s) +{ + int channelCount, depth; + if (stbi__get32be(s) != 0x38425053) { + stbi__rewind( s ); + return 0; + } + if (stbi__get16be(s) != 1) { + stbi__rewind( s ); + return 0; + } + stbi__skip(s, 6); + channelCount = stbi__get16be(s); + if (channelCount < 0 || channelCount > 16) { + stbi__rewind( s ); + return 0; + } + STBI_NOTUSED(stbi__get32be(s)); + STBI_NOTUSED(stbi__get32be(s)); + depth = stbi__get16be(s); + if (depth != 16) { + stbi__rewind( s ); + return 0; + } + return 1; +} +#endif + +#ifndef STBI_NO_PIC +static int stbi__pic_info(stbi__context *s, int *x, int *y, int *comp) +{ + int act_comp=0,num_packets=0,chained,dummy; + stbi__pic_packet packets[10]; + + if (!x) x = &dummy; + if (!y) y = &dummy; + if (!comp) comp = &dummy; + + if (!stbi__pic_is4(s,"\x53\x80\xF6\x34")) { + stbi__rewind(s); + return 0; + } + + stbi__skip(s, 88); + + *x = stbi__get16be(s); + *y = stbi__get16be(s); + if (stbi__at_eof(s)) { + stbi__rewind( s); + return 0; + } + if ( (*x) != 0 && (1 << 28) / (*x) < (*y)) { + stbi__rewind( s ); + return 0; + } + + stbi__skip(s, 8); + + do { + stbi__pic_packet *packet; + + if (num_packets==sizeof(packets)/sizeof(packets[0])) + return 0; + + packet = &packets[num_packets++]; + chained = stbi__get8(s); + packet->size = stbi__get8(s); + packet->type = stbi__get8(s); + packet->channel = stbi__get8(s); + act_comp |= packet->channel; + + if (stbi__at_eof(s)) { + stbi__rewind( s ); + return 0; + } + if (packet->size != 8) { + stbi__rewind( s ); + return 0; + } + } while (chained); + + *comp = (act_comp & 0x10 ? 4 : 3); + + return 1; +} +#endif + +// ************************************************************************************************* +// Portable Gray Map and Portable Pixel Map loader +// by Ken Miller +// +// PGM: http://netpbm.sourceforge.net/doc/pgm.html +// PPM: http://netpbm.sourceforge.net/doc/ppm.html +// +// Known limitations: +// Does not support comments in the header section +// Does not support ASCII image data (formats P2 and P3) + +#ifndef STBI_NO_PNM + +static int stbi__pnm_test(stbi__context *s) +{ + char p, t; + p = (char) stbi__get8(s); + t = (char) stbi__get8(s); + if (p != 'P' || (t != '5' && t != '6')) { + stbi__rewind( s ); + return 0; + } + return 1; +} + +static void *stbi__pnm_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + stbi_uc *out; + STBI_NOTUSED(ri); + + ri->bits_per_channel = stbi__pnm_info(s, (int *)&s->img_x, (int *)&s->img_y, (int *)&s->img_n); + if (ri->bits_per_channel == 0) + return 0; + + if (s->img_y > STBI_MAX_DIMENSIONS) return stbi__errpuc("too large","Very large image (corrupt?)"); + if (s->img_x > STBI_MAX_DIMENSIONS) return stbi__errpuc("too large","Very large image (corrupt?)"); + + *x = s->img_x; + *y = s->img_y; + if (comp) *comp = s->img_n; + + if (!stbi__mad4sizes_valid(s->img_n, s->img_x, s->img_y, ri->bits_per_channel / 8, 0)) + return stbi__errpuc("too large", "PNM too large"); + + out = (stbi_uc *) stbi__malloc_mad4(s->img_n, s->img_x, s->img_y, ri->bits_per_channel / 8, 0); + if (!out) return stbi__errpuc("outofmem", "Out of memory"); + stbi__getn(s, out, s->img_n * s->img_x * s->img_y * (ri->bits_per_channel / 8)); + + if (req_comp && req_comp != s->img_n) { + out = stbi__convert_format(out, s->img_n, req_comp, s->img_x, s->img_y); + if (out == NULL) return out; // stbi__convert_format frees input on failure + } + return out; +} + +static int stbi__pnm_isspace(char c) +{ + return c == ' ' || c == '\t' || c == '\n' || c == '\v' || c == '\f' || c == '\r'; +} + +static void stbi__pnm_skip_whitespace(stbi__context *s, char *c) +{ + for (;;) { + while (!stbi__at_eof(s) && stbi__pnm_isspace(*c)) + *c = (char) stbi__get8(s); + + if (stbi__at_eof(s) || *c != '#') + break; + + while (!stbi__at_eof(s) && *c != '\n' && *c != '\r' ) + *c = (char) stbi__get8(s); + } +} + +static int stbi__pnm_isdigit(char c) +{ + return c >= '0' && c <= '9'; +} + +static int stbi__pnm_getinteger(stbi__context *s, char *c) +{ + int value = 0; + + while (!stbi__at_eof(s) && stbi__pnm_isdigit(*c)) { + value = value*10 + (*c - '0'); + *c = (char) stbi__get8(s); + } + + return value; +} + +static int stbi__pnm_info(stbi__context *s, int *x, int *y, int *comp) +{ + int maxv, dummy; + char c, p, t; + + if (!x) x = &dummy; + if (!y) y = &dummy; + if (!comp) comp = &dummy; + + stbi__rewind(s); + + // Get identifier + p = (char) stbi__get8(s); + t = (char) stbi__get8(s); + if (p != 'P' || (t != '5' && t != '6')) { + stbi__rewind(s); + return 0; + } + + *comp = (t == '6') ? 3 : 1; // '5' is 1-component .pgm; '6' is 3-component .ppm + + c = (char) stbi__get8(s); + stbi__pnm_skip_whitespace(s, &c); + + *x = stbi__pnm_getinteger(s, &c); // read width + stbi__pnm_skip_whitespace(s, &c); + + *y = stbi__pnm_getinteger(s, &c); // read height + stbi__pnm_skip_whitespace(s, &c); + + maxv = stbi__pnm_getinteger(s, &c); // read max value + if (maxv > 65535) + return stbi__err("max value > 65535", "PPM image supports only 8-bit and 16-bit images"); + else if (maxv > 255) + return 16; + else + return 8; +} + +static int stbi__pnm_is16(stbi__context *s) +{ + if (stbi__pnm_info(s, NULL, NULL, NULL) == 16) + return 1; + return 0; +} +#endif + +static int stbi__info_main(stbi__context *s, int *x, int *y, int *comp) +{ + #ifndef STBI_NO_JPEG + if (stbi__jpeg_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_PNG + if (stbi__png_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_GIF + if (stbi__gif_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_BMP + if (stbi__bmp_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_PSD + if (stbi__psd_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_PIC + if (stbi__pic_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_PNM + if (stbi__pnm_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_HDR + if (stbi__hdr_info(s, x, y, comp)) return 1; + #endif + + // test tga last because it's a crappy test! + #ifndef STBI_NO_TGA + if (stbi__tga_info(s, x, y, comp)) + return 1; + #endif + return stbi__err("unknown image type", "Image not of any known type, or corrupt"); +} + +static int stbi__is_16_main(stbi__context *s) +{ + #ifndef STBI_NO_PNG + if (stbi__png_is16(s)) return 1; + #endif + + #ifndef STBI_NO_PSD + if (stbi__psd_is16(s)) return 1; + #endif + + #ifndef STBI_NO_PNM + if (stbi__pnm_is16(s)) return 1; + #endif + return 0; +} + +#ifndef STBI_NO_STDIO +STBIDEF int stbi_info(char const *filename, int *x, int *y, int *comp) +{ + FILE *f = stbi__fopen(filename, "rb"); + int result; + if (!f) return stbi__err("can't fopen", "Unable to open file"); + result = stbi_info_from_file(f, x, y, comp); + fclose(f); + return result; +} + +STBIDEF int stbi_info_from_file(FILE *f, int *x, int *y, int *comp) +{ + int r; + stbi__context s; + long pos = ftell(f); + stbi__start_file(&s, f); + r = stbi__info_main(&s,x,y,comp); + fseek(f,pos,SEEK_SET); + return r; +} + +STBIDEF int stbi_is_16_bit(char const *filename) +{ + FILE *f = stbi__fopen(filename, "rb"); + int result; + if (!f) return stbi__err("can't fopen", "Unable to open file"); + result = stbi_is_16_bit_from_file(f); + fclose(f); + return result; +} + +STBIDEF int stbi_is_16_bit_from_file(FILE *f) +{ + int r; + stbi__context s; + long pos = ftell(f); + stbi__start_file(&s, f); + r = stbi__is_16_main(&s); + fseek(f,pos,SEEK_SET); + return r; +} +#endif // !STBI_NO_STDIO + +STBIDEF int stbi_info_from_memory(stbi_uc const *buffer, int len, int *x, int *y, int *comp) +{ + stbi__context s; + stbi__start_mem(&s,buffer,len); + return stbi__info_main(&s,x,y,comp); +} + +STBIDEF int stbi_info_from_callbacks(stbi_io_callbacks const *c, void *user, int *x, int *y, int *comp) +{ + stbi__context s; + stbi__start_callbacks(&s, (stbi_io_callbacks *) c, user); + return stbi__info_main(&s,x,y,comp); +} + +STBIDEF int stbi_is_16_bit_from_memory(stbi_uc const *buffer, int len) +{ + stbi__context s; + stbi__start_mem(&s,buffer,len); + return stbi__is_16_main(&s); +} + +STBIDEF int stbi_is_16_bit_from_callbacks(stbi_io_callbacks const *c, void *user) +{ + stbi__context s; + stbi__start_callbacks(&s, (stbi_io_callbacks *) c, user); + return stbi__is_16_main(&s); +} + +#endif // STB_IMAGE_IMPLEMENTATION + +/* + revision history: + 2.20 (2019-02-07) support utf8 filenames in Windows; fix warnings and platform ifdefs + 2.19 (2018-02-11) fix warning + 2.18 (2018-01-30) fix warnings + 2.17 (2018-01-29) change sbti__shiftsigned to avoid clang -O2 bug + 1-bit BMP + *_is_16_bit api + avoid warnings + 2.16 (2017-07-23) all functions have 16-bit variants; + STBI_NO_STDIO works again; + compilation fixes; + fix rounding in unpremultiply; + optimize vertical flip; + disable raw_len validation; + documentation fixes + 2.15 (2017-03-18) fix png-1,2,4 bug; now all Imagenet JPGs decode; + warning fixes; disable run-time SSE detection on gcc; + uniform handling of optional "return" values; + thread-safe initialization of zlib tables + 2.14 (2017-03-03) remove deprecated STBI_JPEG_OLD; fixes for Imagenet JPGs + 2.13 (2016-11-29) add 16-bit API, only supported for PNG right now + 2.12 (2016-04-02) fix typo in 2.11 PSD fix that caused crashes + 2.11 (2016-04-02) allocate large structures on the stack + remove white matting for transparent PSD + fix reported channel count for PNG & BMP + re-enable SSE2 in non-gcc 64-bit + support RGB-formatted JPEG + read 16-bit PNGs (only as 8-bit) + 2.10 (2016-01-22) avoid warning introduced in 2.09 by STBI_REALLOC_SIZED + 2.09 (2016-01-16) allow comments in PNM files + 16-bit-per-pixel TGA (not bit-per-component) + info() for TGA could break due to .hdr handling + info() for BMP to shares code instead of sloppy parse + can use STBI_REALLOC_SIZED if allocator doesn't support realloc + code cleanup + 2.08 (2015-09-13) fix to 2.07 cleanup, reading RGB PSD as RGBA + 2.07 (2015-09-13) fix compiler warnings + partial animated GIF support + limited 16-bpc PSD support + #ifdef unused functions + bug with < 92 byte PIC,PNM,HDR,TGA + 2.06 (2015-04-19) fix bug where PSD returns wrong '*comp' value + 2.05 (2015-04-19) fix bug in progressive JPEG handling, fix warning + 2.04 (2015-04-15) try to re-enable SIMD on MinGW 64-bit + 2.03 (2015-04-12) extra corruption checking (mmozeiko) + stbi_set_flip_vertically_on_load (nguillemot) + fix NEON support; fix mingw support + 2.02 (2015-01-19) fix incorrect assert, fix warning + 2.01 (2015-01-17) fix various warnings; suppress SIMD on gcc 32-bit without -msse2 + 2.00b (2014-12-25) fix STBI_MALLOC in progressive JPEG + 2.00 (2014-12-25) optimize JPG, including x86 SSE2 & NEON SIMD (ryg) + progressive JPEG (stb) + PGM/PPM support (Ken Miller) + STBI_MALLOC,STBI_REALLOC,STBI_FREE + GIF bugfix -- seemingly never worked + STBI_NO_*, STBI_ONLY_* + 1.48 (2014-12-14) fix incorrectly-named assert() + 1.47 (2014-12-14) 1/2/4-bit PNG support, both direct and paletted (Omar Cornut & stb) + optimize PNG (ryg) + fix bug in interlaced PNG with user-specified channel count (stb) + 1.46 (2014-08-26) + fix broken tRNS chunk (colorkey-style transparency) in non-paletted PNG + 1.45 (2014-08-16) + fix MSVC-ARM internal compiler error by wrapping malloc + 1.44 (2014-08-07) + various warning fixes from Ronny Chevalier + 1.43 (2014-07-15) + fix MSVC-only compiler problem in code changed in 1.42 + 1.42 (2014-07-09) + don't define _CRT_SECURE_NO_WARNINGS (affects user code) + fixes to stbi__cleanup_jpeg path + added STBI_ASSERT to avoid requiring assert.h + 1.41 (2014-06-25) + fix search&replace from 1.36 that messed up comments/error messages + 1.40 (2014-06-22) + fix gcc struct-initialization warning + 1.39 (2014-06-15) + fix to TGA optimization when req_comp != number of components in TGA; + fix to GIF loading because BMP wasn't rewinding (whoops, no GIFs in my test suite) + add support for BMP version 5 (more ignored fields) + 1.38 (2014-06-06) + suppress MSVC warnings on integer casts truncating values + fix accidental rename of 'skip' field of I/O + 1.37 (2014-06-04) + remove duplicate typedef + 1.36 (2014-06-03) + convert to header file single-file library + if de-iphone isn't set, load iphone images color-swapped instead of returning NULL + 1.35 (2014-05-27) + various warnings + fix broken STBI_SIMD path + fix bug where stbi_load_from_file no longer left file pointer in correct place + fix broken non-easy path for 32-bit BMP (possibly never used) + TGA optimization by Arseny Kapoulkine + 1.34 (unknown) + use STBI_NOTUSED in stbi__resample_row_generic(), fix one more leak in tga failure case + 1.33 (2011-07-14) + make stbi_is_hdr work in STBI_NO_HDR (as specified), minor compiler-friendly improvements + 1.32 (2011-07-13) + support for "info" function for all supported filetypes (SpartanJ) + 1.31 (2011-06-20) + a few more leak fixes, bug in PNG handling (SpartanJ) + 1.30 (2011-06-11) + added ability to load files via callbacks to accomidate custom input streams (Ben Wenger) + removed deprecated format-specific test/load functions + removed support for installable file formats (stbi_loader) -- would have been broken for IO callbacks anyway + error cases in bmp and tga give messages and don't leak (Raymond Barbiero, grisha) + fix inefficiency in decoding 32-bit BMP (David Woo) + 1.29 (2010-08-16) + various warning fixes from Aurelien Pocheville + 1.28 (2010-08-01) + fix bug in GIF palette transparency (SpartanJ) + 1.27 (2010-08-01) + cast-to-stbi_uc to fix warnings + 1.26 (2010-07-24) + fix bug in file buffering for PNG reported by SpartanJ + 1.25 (2010-07-17) + refix trans_data warning (Won Chun) + 1.24 (2010-07-12) + perf improvements reading from files on platforms with lock-heavy fgetc() + minor perf improvements for jpeg + deprecated type-specific functions so we'll get feedback if they're needed + attempt to fix trans_data warning (Won Chun) + 1.23 fixed bug in iPhone support + 1.22 (2010-07-10) + removed image *writing* support + stbi_info support from Jetro Lauha + GIF support from Jean-Marc Lienher + iPhone PNG-extensions from James Brown + warning-fixes from Nicolas Schulz and Janez Zemva (i.stbi__err. Janez (U+017D)emva) + 1.21 fix use of 'stbi_uc' in header (reported by jon blow) + 1.20 added support for Softimage PIC, by Tom Seddon + 1.19 bug in interlaced PNG corruption check (found by ryg) + 1.18 (2008-08-02) + fix a threading bug (local mutable static) + 1.17 support interlaced PNG + 1.16 major bugfix - stbi__convert_format converted one too many pixels + 1.15 initialize some fields for thread safety + 1.14 fix threadsafe conversion bug + header-file-only version (#define STBI_HEADER_FILE_ONLY before including) + 1.13 threadsafe + 1.12 const qualifiers in the API + 1.11 Support installable IDCT, colorspace conversion routines + 1.10 Fixes for 64-bit (don't use "unsigned long") + optimized upsampling by Fabian "ryg" Giesen + 1.09 Fix format-conversion for PSD code (bad global variables!) + 1.08 Thatcher Ulrich's PSD code integrated by Nicolas Schulz + 1.07 attempt to fix C++ warning/errors again + 1.06 attempt to fix C++ warning/errors again + 1.05 fix TGA loading to return correct *comp and use good luminance calc + 1.04 default float alpha is 1, not 255; use 'void *' for stbi_image_free + 1.03 bugfixes to STBI_NO_STDIO, STBI_NO_HDR + 1.02 support for (subset of) HDR files, float interface for preferred access to them + 1.01 fix bug: possible bug in handling right-side up bmps... not sure + fix bug: the stbi__bmp_load() and stbi__tga_load() functions didn't work at all + 1.00 interface to zlib that skips zlib header + 0.99 correct handling of alpha in palette + 0.98 TGA loader by lonesock; dynamically add loaders (untested) + 0.97 jpeg errors on too large a file; also catch another malloc failure + 0.96 fix detection of invalid v value - particleman@mollyrocket forum + 0.95 during header scan, seek to markers in case of padding + 0.94 STBI_NO_STDIO to disable stdio usage; rename all #defines the same + 0.93 handle jpegtran output; verbose errors + 0.92 read 4,8,16,24,32-bit BMP files of several formats + 0.91 output 24-bit Windows 3.0 BMP files + 0.90 fix a few more warnings; bump version number to approach 1.0 + 0.61 bugfixes due to Marc LeBlanc, Christopher Lloyd + 0.60 fix compiling as c++ + 0.59 fix warnings: merge Dave Moore's -Wall fixes + 0.58 fix bug: zlib uncompressed mode len/nlen was wrong endian + 0.57 fix bug: jpg last huffman symbol before marker was >9 bits but less than 16 available + 0.56 fix bug: zlib uncompressed mode len vs. nlen + 0.55 fix bug: restart_interval not initialized to 0 + 0.54 allow NULL for 'int *comp' + 0.53 fix bug in png 3->4; speedup png decoding + 0.52 png handles req_comp=3,4 directly; minor cleanup; jpeg comments + 0.51 obey req_comp requests, 1-component jpegs return as 1-component, + on 'test' only check type, not whether we support this variant + 0.50 (2006-11-19) + first released version +*/ + + +/* +------------------------------------------------------------------------------ +This software is available under 2 licenses -- choose whichever you prefer. +------------------------------------------------------------------------------ +ALTERNATIVE A - MIT License +Copyright (c) 2017 Sean Barrett +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. +------------------------------------------------------------------------------ +ALTERNATIVE B - Public Domain (www.unlicense.org) +This is free and unencumbered software released into the public domain. +Anyone is free to copy, modify, publish, use, compile, sell, or distribute this +software, either in source code form or as a compiled binary, for any purpose, +commercial or non-commercial, and by any means. +In jurisdictions that recognize copyright laws, the author or authors of this +software dedicate any and all copyright interest in the software to the public +domain. We make this dedication for the benefit of the public at large and to +the detriment of our heirs and successors. We intend this dedication to be an +overt act of relinquishment in perpetuity of all present and future rights to +this software under copyright law. +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 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. +------------------------------------------------------------------------------ +*/ diff --git a/LICENSE b/LICENSE new file mode 100644 index 0000000..e4aa019 --- /dev/null +++ b/LICENSE @@ -0,0 +1,55 @@ +# +# Roo/E, the Kangaroo Punch Portable GUI Toolkit +# Copyright (C) 2026 Scott Duensing +# +# http://kangaroopunch.com +# +# +# This file is part of Roo/E. +# +# Roo/E is free software: you can redistribute it and/or modify it under the +# terms of the GNU Affero General Public License as published by the Free +# Software Foundation, either version 3 of the License, or (at your option) +# any later version. +# +# Roo/E 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 Affero General Public License +# along with Roo/E. If not, see . +# + + +Third Party Library Licenses: + + +MemWatch +-------- +http://www.linkdata.se/sourcecode/memwatch/ +GPL2 + + +PThreads +-------- +https://www.delorie.com/pub/djgpp/current/v2tk/ +LGPL2 + + +SDL2 +---- +https://www.libsdl.org/ +BSD 3-Clause + + +stb_ds.h +-------- +https://github.com/nothings/stb +Public Domain + + +stb_image.h +----------- +https://github.com/nothings/stb +Public Domain diff --git a/Makefile b/Makefile new file mode 100644 index 0000000..c6dbfc3 --- /dev/null +++ b/Makefile @@ -0,0 +1,129 @@ +# +# Roo/E, the Kangaroo Punch Portable GUI Toolkit +# Copyright (C) 2026 Scott Duensing +# +# http://kangaroopunch.com +# +# +# This file is part of Roo/E. +# +# Roo/E is free software: you can redistribute it and/or modify it under the +# terms of the GNU Affero General Public License as published by the Free +# Software Foundation, either version 3 of the License, or (at your option) +# any later version. +# +# Roo/E 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 Affero General Public License +# along with Roo/E. If not, see . +# + + +# Tools +CC := gcc +AR := ar +DXE3GEN := dxe3gen +EXE2COFF := exe2coff + +# Compiler Settings +CFLAGS := -DPLATFORM_DOS -Wall +LDFLAGS := + +# Output Directories +OBJ := obj +BIN := bin +DYN := $(BIN)/dyn +APP := $(BIN)/app +SDK := $(BIN)/sdk +FNT := $(BIN)/fonts +FIN := font/in + +# Include Paths +INC := ../include include 3rdparty 3rdparty/pthreads/include + +# Font Compiler Source and Target - NOTE: This is a Linux binary! +FONT_ELF := font +FONT_SRC := $(shell find font/src -name '*.c') +FONT_OBJ := $(FONT_SRC:%=$(OBJ)/%.o) +FONT_LIB := -lm + +# Roo/E Dynamic Libraries +DYNS_EXE := stbimage.dyn +DYNS_SRC := $(shell find dyn/stbimage -name '*.c') +DYNS_OBJ := $(DYNS_SRC:%=$(OBJ)/%.o) +DYNS_LIB := + +# Roo/E Source and Target +ROOE_EXE := roo_e.exe +ROOE_SRC := $(shell find roo_e -name '*.c') +ROOE_OBJ := $(ROOE_SRC:%=$(OBJ)/%.o) +ROOE_LIB := 3rdparty/pthreads/lib/libgthreads.a -lgcc + +# MultiPlayer Game Client +MPGC_EXE := kpsmpgc.app +MPGC_SRC := $(shell find kpsmpgc -name '*.c') +MPGC_OBJ := $(MPGC_SRC:%=$(OBJ)/%.o) +MPGC_LIB := + +FONTS := $(FNT)/vga4x8.fnt $(FNT)/vga8x8.fnt $(FNT)/vga8x14.fnt $(FNT)/vga8x16.fnt + +# Wiring +FONT_PNGS := $(subst $(FNT),$(FIN),$(FONTS)) +FONT_PNGS := $(subst .fnt,.png,$(FONTS_PNGS)) +INC_FLAGS := $(addprefix -I,$(INC)) +CFLAGS += $(INC_FLAGS) + +# Force User to Run This Using Toolchains +.PHONY: use_build_script +use_build_script: + @echo "Use build.sh to cross-compile, not make." + +# Build Everything +.PHONY: dos +dos: $(BIN)/$(ROOE_EXE) $(APP)/$(MPGC_EXE) $(DYN)/$(DYNS_EXE) + +.PHONY: linux +linux: $(FONTS) + +# Remove Everything +.PHONY: clean +clean: + rm -rf $(OBJ) $(BIN) + +# Font Compiler Target +$(BIN)/$(FONT_ELF): $(FONT_OBJ) Makefile + mkdir -p $(dir $@) $(BIN)/fonts + $(CC) $(FONT_OBJ) -o $@ $(LDFLAGS) $(FONT_LIB) + +$(FONTS) &: $(FIN)/vga4x8.png $(FIN)/vga4x8.png $(FIN)/vga4x8.png $(FIN)/vga4x8.png $(BIN)/$(FONT_ELF) + cd $(BIN) && ./$(FONT_ELF) + +# Roo/E Target +$(BIN)/$(ROOE_EXE): $(ROOE_OBJ) Makefile + mkdir -p $(dir $@) + $(CC) $(ROOE_OBJ) -o $@ $(LDFLAGS) $(ROOE_LIB) + # Embed the DPMI Server + $(EXE2COFF) $(BIN)/$(ROOE_EXE) + cat 3rdparty/CWSDPMI/CWSDSTUB.EXE $(basename $(BIN)/$(ROOE_EXE)) > $(BIN)/$(ROOE_EXE) + rm $(basename $(BIN)/$(ROOE_EXE)) + +# MPGC Target +$(APP)/$(MPGC_EXE): $(MPGC_OBJ) Makefile + mkdir -p $(dir $@) + $(DXE3GEN) -o $@ $(MPGC_OBJ) -U $(LDFLAGS) $(MPGC_LIB) + +# DYNS Target +$(DYN)/$(DYNS_EXE): $(DYNS_OBJ) Makefile + mkdir -p $(dir $@) + $(AR) rcs $(OBJ)/$(DYNS_EXE).a $(DYNS_OBJ) + $(DXE3GEN) -o $@ -Y $@.a --whole-archive -U $(OBJ)/$(DYNS_EXE).a $(LDFLAGS) $(DYNS_LIB) + mkdir -p $(SDK) + mv $(DYN)/$(DYNS_EXE).a $(SDK)/. + +# Build C Files +$(OBJ)/%.c.o: %.c + mkdir -p $(dir $@) + $(CC) $(CFLAGS) -c $< -o $@ diff --git a/agpl30.txt b/agpl30.txt new file mode 100644 index 0000000..be3f7b2 --- /dev/null +++ b/agpl30.txt @@ -0,0 +1,661 @@ + GNU AFFERO GENERAL PUBLIC LICENSE + Version 3, 19 November 2007 + + Copyright (C) 2007 Free Software Foundation, Inc. + Everyone is permitted to copy and distribute verbatim copies + of this license document, but changing it is not allowed. + + Preamble + + The GNU Affero General Public License is a free, copyleft license for +software and other kinds of works, specifically designed to ensure +cooperation with the community in the case of network server software. + + The licenses for most software and other practical works are designed +to take away your freedom to share and change the works. By contrast, +our General Public Licenses are 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. + + 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. + + Developers that use our General Public Licenses protect your rights +with two steps: (1) assert copyright on the software, and (2) offer +you this License which gives you legal permission to copy, distribute +and/or modify the software. + + A secondary benefit of defending all users' freedom is that +improvements made in alternate versions of the program, if they +receive widespread use, become available for other developers to +incorporate. Many developers of free software are heartened and +encouraged by the resulting cooperation. However, in the case of +software used on network servers, this result may fail to come about. +The GNU General Public License permits making a modified version and +letting the public access it on a server without ever releasing its +source code to the public. + + The GNU Affero General Public License is designed specifically to +ensure that, in such cases, the modified source code becomes available +to the community. It requires the operator of a network server to +provide the source code of the modified version running there to the +users of that server. Therefore, public use of a modified version, on +a publicly accessible server, gives the public access to the source +code of the modified version. + + An older license, called the Affero General Public License and +published by Affero, was designed to accomplish similar goals. This is +a different license, not a version of the Affero GPL, but Affero has +released a new version of the Affero GPL which permits relicensing under +this license. + + 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 Affero 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. Remote Network Interaction; Use with the GNU General Public License. + + Notwithstanding any other provision of this License, if you modify the +Program, your modified version must prominently offer all users +interacting with it remotely through a computer network (if your version +supports such interaction) an opportunity to receive the Corresponding +Source of your version by providing access to the Corresponding Source +from a network server at no charge, through some standard or customary +means of facilitating copying of software. This Corresponding Source +shall include the Corresponding Source for any work covered by version 3 +of the GNU General Public License that is incorporated pursuant to the +following paragraph. + + 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 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 work with which it is combined will remain governed by version +3 of the GNU General Public License. + + 14. Revised Versions of this License. + + The Free Software Foundation may publish revised and/or new versions of +the GNU Affero 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 Affero 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 Affero 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 Affero General Public License can be used, that proxy's +public statement of acceptance of a version permanently authorizes you +to choose that version for the Program. + + Later license versions may give you additional or different +permissions. However, no additional obligations are imposed on any +author or copyright holder as a result of your choosing to follow a +later version. + + 15. Disclaimer of Warranty. + + THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY +APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT +HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY +OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, +THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR +PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM +IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF +ALL NECESSARY SERVICING, REPAIR OR CORRECTION. + + 16. Limitation of Liability. + + IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING +WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS +THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY +GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE +USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF +DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD +PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), +EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF +SUCH DAMAGES. + + 17. Interpretation of Sections 15 and 16. + + If the disclaimer of warranty and limitation of liability provided +above cannot be given local legal effect according to their terms, +reviewing courts shall apply local law that most closely approximates +an absolute waiver of all civil liability in connection with the +Program, unless a warranty or assumption of liability accompanies a +copy of the Program in return for a fee. + + END OF TERMS AND CONDITIONS + + How to Apply These Terms to Your New Programs + + If you develop a new program, and you want it to be of the greatest +possible use to the public, the best way to achieve this is to make it +free software which everyone can redistribute and change under these terms. + + To do so, attach the following notices to the program. It is safest +to attach them to the start of each source file to most effectively +state the exclusion of warranty; and each file should have at least +the "copyright" line and a pointer to where the full notice is found. + + + Copyright (C) + + This program is free software: you can redistribute it and/or modify + it under the terms of the GNU Affero General Public License as published by + the Free Software Foundation, either version 3 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU Affero General Public License for more details. + + You should have received a copy of the GNU Affero General Public License + along with this program. If not, see . + +Also add information on how to contact you by electronic and paper mail. + + If your software can interact with users remotely through a computer +network, you should also make sure that it provides a way for users to +get its source. For example, if your program is a web application, its +interface could display a "Source" link that leads users to an archive +of the code. There are many ways you could offer source, and different +solutions will be better for different programs; see section 13 for the +specific requirements. + + You should also get your employer (if you work as a programmer) or school, +if any, to sign a "copyright disclaimer" for the program, if necessary. +For more information on this, and how to apply and follow the GNU AGPL, see +. diff --git a/build.sh b/build.sh new file mode 100755 index 0000000..dda3402 --- /dev/null +++ b/build.sh @@ -0,0 +1,32 @@ +#!/bin/bash + +# +# Roo/E, the Kangaroo Punch Portable GUI Toolkit +# Copyright (C) 2026 Scott Duensing +# +# http://kangaroopunch.com +# +# +# This file is part of Roo/E. +# +# Roo/E is free software: you can redistribute it and/or modify it under the +# terms of the GNU Affero General Public License as published by the Free +# Software Foundation, either version 3 of the License, or (at your option) +# any later version. +# +# Roo/E 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 Affero General Public License +# along with Roo/E. If not, see . +# + + +# Linux side first. +make linux + +# Now do the DOS part. +eval "$(../toolchains/toolchains.sh use x86 dos)" +make dos diff --git a/dyn/stbimage/stbimage.c b/dyn/stbimage/stbimage.c new file mode 100644 index 0000000..4ccec0e --- /dev/null +++ b/dyn/stbimage/stbimage.c @@ -0,0 +1,39 @@ +/* Roo/E, the Kangaroo Punch Portable GUI Toolkit +* Copyright (C) 2026 Scott Duensing +* +* http://kangaroopunch.com +* +* +* This file is part of Roo/E. +* +* Roo/E is free software: you can redistribute it and/or modify it under the +* terms of the GNU Affero General Public License as published by the Free +* Software Foundation, either version 3 of the License, or (at your option) +* any later version. +* +* Roo/E 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 Affero General Public License +* along with Roo/E. If not, see . +*/ + + +#include + +#define STB_IMAGE_IMPLEMENTATION +#include "stb_image.h" + + +int dynStart(void) { + printf("stb_image starting!\n"); + return 0; +} + + +int dynStop(void) { + printf("stb_image stopping!\n"); + return 0; +} diff --git a/env.sh b/env.sh new file mode 100755 index 0000000..10d9977 --- /dev/null +++ b/env.sh @@ -0,0 +1,28 @@ +#!/bin/bash + +# +# Roo/E, the Kangaroo Punch Portable GUI Toolkit +# Copyright (C) 2026 Scott Duensing +# +# http://kangaroopunch.com +# +# +# This file is part of Roo/E. +# +# Roo/E is free software: you can redistribute it and/or modify it under the +# terms of the GNU Affero General Public License as published by the Free +# Software Foundation, either version 3 of the License, or (at your option) +# any later version. +# +# Roo/E 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 Affero General Public License +# along with Roo/E. If not, see . +# + + +eval "$(../toolchains/toolchains.sh use x86 dos)" +export C_INCLUDE_PATH=${HOME}/code/toolchains/x-tools/djgpp/i586-pc-msdosdjgpp/sys-include diff --git a/font/in/vga4x8.png b/font/in/vga4x8.png new file mode 100644 index 0000000..579fd48 --- /dev/null +++ b/font/in/vga4x8.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b4c2a6fe844525cdd4abddd2c7aa2d6d73ec667dd82aee10e8a9c62e2f367d8b +size 18029 diff --git a/font/in/vga8x14.png b/font/in/vga8x14.png new file mode 100644 index 0000000..a52f226 --- /dev/null +++ b/font/in/vga8x14.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5330f2c589cd0094131d7b686947fbc1da0a6c3df4388a09f3bb77ba57d53a0c +size 22459 diff --git a/font/in/vga8x16.png b/font/in/vga8x16.png new file mode 100644 index 0000000..f49a9d9 --- /dev/null +++ b/font/in/vga8x16.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:170c659c489d3cd1d141dbf12ae37bbf40975941ced925ce6e50817128e5bf97 +size 20257 diff --git a/font/in/vga8x8.png b/font/in/vga8x8.png new file mode 100644 index 0000000..64e316d --- /dev/null +++ b/font/in/vga8x8.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:20b093388e649e7f0eb3c3ca143a592d00c89579b6390cac7bd86287be19bd48 +size 3512 diff --git a/font/src/main.c b/font/src/main.c new file mode 100644 index 0000000..5292969 --- /dev/null +++ b/font/src/main.c @@ -0,0 +1,100 @@ +/* Roo/E, the Kangaroo Punch Portable GUI Toolkit +* Copyright (C) 2026 Scott Duensing +* +* http://kangaroopunch.com +* +* +* This file is part of Roo/E. +* +* Roo/E is free software: you can redistribute it and/or modify it under the +* terms of the GNU Affero General Public License as published by the Free +* Software Foundation, either version 3 of the License, or (at your option) +* any later version. +* +* Roo/E 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 Affero General Public License +* along with Roo/E. If not, see . +*/ + + +#include + +#ifdef PLATFORM_LINUX +#define MEMWATCH +#include "memwatch/memwatch.h" +#endif + +#define STB_IMAGE_IMPLEMENTATION +#define STBI_ONLY_PNG +#include "stb_image.h" + +#include "stddclmr.h" + + +void makeFont(char *source, char *target, int pixelsW, int pixelsH, int charsW, int charCount); + + +void makeFont(char *source, char *target, int pixelsW, int pixelsH, int charsW, int charCount) { + unsigned char *font = NULL; + unsigned char data = 0; + FILE *out = NULL; + int bits = 0; + int x; + int y; + int n; + int w; + int h; + + // Load font atlas from disk. + font = stbi_load(source, (int *)&w, (int *)&h, (int *)&n, 3); + if (!font) return; + + // Create data file for font. + out = fopen(target, "wb"); + if (!out) { + stbi_image_free(font); + return; + } + + // Provide some metadata for enhancement later. + fputc(pixelsW, out); // Width of characters in pixels + fputc(pixelsH, out); // Height of characters in pixels + fputc(charsW, out); // Number of characters per row + fputc(charCount - 1, out); // Number of characters - 1 + + // Convert bitmap to actual bits. + for (y=0; y 7) { + bits = 0; + fputc(data, out); + } + } + } + + fclose(out); + + // Clean up. + stbi_image_free(font); +} + + +int main(int argc, char *argv[]) { + (void)argc; + (void)argv; + + // Run this in DOS from the BIN folder. + makeFont("../font/in/vga4x8.png", "fonts/vga4x8.fnt", 4, 8, 16, 256); + makeFont("../font/in/vga8x8.png", "fonts/vga8x8.fnt", 8, 8, 16, 256); + makeFont("../font/in/vga8x14.png", "fonts/vga8x14.fnt", 8, 14, 16, 256); + makeFont("../font/in/vga8x16.png", "fonts/vga8x16.fnt", 8, 16, 16, 256); + + return 0; +} diff --git a/icon.png b/icon.png new file mode 100644 index 0000000..6d0b444 --- /dev/null +++ b/icon.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1af1efed78cd9bf6dc046ebc934aa2e327cbc5244eca6f55b51dac9ffdfa25af +size 103464 diff --git a/include/stddclmr.h b/include/stddclmr.h new file mode 100644 index 0000000..9502ae6 --- /dev/null +++ b/include/stddclmr.h @@ -0,0 +1,120 @@ +/* + * Roo/E, the Kangaroo Punch Portable GUI Toolkit + * Copyright (C) 2026 Scott Duensing + * + * http://kangaroopunch.com + * + * + * This file is part of Roo/E. + * + * Roo/E is free software: you can redistribute it and/or modify it under the + * terms of the GNU Affero General Public License as published by the Free + * Software Foundation, either version 3 of the License, or (at your option) + * any later version. + * + * Roo/E 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 Affero General Public License + * along with Roo/E. If not, see . + * + */ + + +#ifndef STDDCLMR_H +#define STDDCLMR_H + +/* +Action figures sold separately. Add toner. All models over 18 years of age. +All rights reserved. Allow four to six weeks for delivery. An equal +opportunity employer. Any resemblance to actual persons, living or dead, is +unintentional and purely coincidental. Apply only to affected area. Approved +for veterans. As seen on TV. At participating locations only. Avoid contact +with mucous membranes. Avoid contact with skin. Avoid extreme temperatures +and store in a cool dry place. Batteries not included. Be sure each item is +properly endorsed. Beware of dog. Booths for two or more. Breaking seal +constitutes acceptance of agreement. Call toll free number before digging. +Caveat emptor. Check here if tax deductible. Close cover before striking +Colors may fade. Contains a substantial amount of non-tobacco ingredients. +Contents may settle during shipment. Contestants have been briefed on some +questions before the show. Copyright 1995 Joker's Wild. Disclaimer does +not cover hurricane, lightning, tornado, tsunami, volcanic eruption, +earthquake, flood, and other Acts of God, misuse, neglect, unauthorized +repair, damage from improper installation, broken antenna or marred cabinet, +incorrect line voltage, missing or altered serial numbers, sonic boom +vibrations, electromagnetic radiation from nuclear blasts, customer +adjustments that are not covered in the joke list, and incidents owing to +airplane crash, ship sinking, motor vehicle accidents, leaky roof, broken +glass, falling rocks, mud slides, forest fire, flying projectiles, or +dropping the item. Do not bend, fold, mutilate, or spindle. Do not place +near flammable or magnetic source. Do not puncture, incinerate, or store +above 120 degrees Fahrenheit. Do not stamp. Use other side for additional +listings. Do not use while operating a motor vehicle or heavy equipment. Do +not write below this line. Documents are provided "as is" without any +warranties expressed or implied. Don't quote me on anything. Don't quote me +on that. Driver does not carry cash. Drop in any mailbox. Edited for +television. Employees and their families are not eligible. Falling rock. +First pull up, then pull down. Flames redirected to /dev/null. For a +limited time only. For external use only. For off-road use only. For office +use only. For recreational use only. Do not disturb. Freshest if eaten +before date on carton. Hand wash only, tumble dry on low heat. If a rash, +redness, irritation, or swelling develops, discontinue use. If condition +persists, consult your physician. If defects are discovered, do not attempt +to fix them yourself, but return to an authorized service center. If +ingested, do not induce vomiting, if symptoms persist, consult a doctor. +Keep away from open flames and avoid inhaling fumes. Keep away from +sunlight, pets, and small children. Keep cool; process promptly. Limit +one-per-family please. Limited time offer, call now to ensure prompt +delivery. List at least two alternate dates. List each check separately by +bank number. List was current at time of printing. Lost ticket pays maximum +rate. May be too intense for some viewers. Must be 18 to enter. No Canadian +coins. No alcohol, dogs or horses. No anchovies unless otherwise specified. +No animals were harmed in the production of these documents. No money down. +No other warranty expressed or implied. No passes accepted for this +engagement. No postage necessary if mailed in the United States. No +preservatives added. No purchase necessary. No salt, MSG, artificial color +or flavor added. No shoes, no shirt, no service, no kidding. No solicitors. +No substitutions allowed. No transfers issued until the bus comes to a +complete stop. No user-serviceable parts inside. Not affiliated with the +American Red Cross. Not liable for damages due to use or misuse. Not +recommended for children. Not responsible for direct, indirect, incidental +or consequential damages resulting from any defect, error or failure to +perform. Not the Beatles. Objects in mirror may be closer than they appear. +One size fits all. Many suitcases look alike. Other copyright laws for +specific entries apply wherever noted. Other restrictions may apply. Package +sold by weight, not volume. Parental advisory - explicit lyrics. Penalty for +private use. Place stamp here. Please remain seated until the ride has come +to a complete stop. Possible penalties for early withdrawal. Post office will +not deliver without postage. Postage will be paid by addressee. Prerecorded +for this time zone. Price does not include taxes. Processed at location +stamped in code at top of carton. Quantities are limited while supplies last. +Read at your own risk. Record additional transactions on back of previous +stub. Replace with same type. Reproduction strictly prohibited. Restaurant +package, not for resale. Return to sender, no forwarding order on file, +unable to forward. Safety goggles may be required during use. Sanitized for +your protection. Sealed for your protection, do not use if the safety seal is +broken. See label for sequence. Shading within a garment may occur. Sign here +without admitting guilt. Simulated picture. Slightly enlarged to show detail. +Slightly higher west of the Rockies. Slippery when wet. Smoking these may be +hazardous to your health. Some assembly required. Some equipment shown is +optional. Some of the trademarks mentioned in this product appear for +identification purposes only. Subject to FCC approval. Subject to change +without notice. Substantial penalty for early withdrawal. Text may contain +material some readers may find objectionable, parental guidance is advised. +Text used in these documents is made from 100% recycled electrons and magnetic +particles. These documents do not reflect the thoughts or opinions of either +myself, my company, my friends, or my rabbit. This is not an offer to sell +securities. This offer is void where prohibited, taxed, or otherwise +restricted. This product is meant for educational purposes only. Times +approximate. Unix is a registered trademark of AT&T. Use only as directed. Use +only in a well-ventilated are. User assumes full liabilities. Void where +prohibited. We have sent the forms which seem right for you. You must be +present to win. You need not be present to win. Your canceled check is your +receipt. Your mileage may vary. I didn't do it. You can't prove anything. + +This supersedes all previous notices. +*/ + +#endif // STDDCLMR_H diff --git a/kpsmpgc/main.c b/kpsmpgc/main.c new file mode 100644 index 0000000..2bb26bd --- /dev/null +++ b/kpsmpgc/main.c @@ -0,0 +1,45 @@ +/* Roo/E, the Kangaroo Punch Portable GUI Toolkit +* Copyright (C) 2026 Scott Duensing +* +* http://kangaroopunch.com +* +* +* This file is part of Roo/E. +* +* Roo/E is free software: you can redistribute it and/or modify it under the +* terms of the GNU Affero General Public License as published by the Free +* Software Foundation, either version 3 of the License, or (at your option) +* any later version. +* +* Roo/E 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 Affero General Public License +* along with Roo/E. If not, see . +*/ + + +#include +#include "stb_image.h" + + +void __attribute__((constructor)) registerApp(void) { + printf("App Loaded!\n"); +} + + +void __attribute__((destructor)) unregisterApp(void) { + printf("App Unloaded!\n"); +} + + +int appMain(const int argc, char* argv[]) { + (void)argc; + (void)argv; + + printf("App running!\n"); + printf("stbi_failure_reason: [%s]\n", stbi_failure_reason()); + return 0; +} diff --git a/notes.txt b/notes.txt new file mode 100644 index 0000000..098c6eb --- /dev/null +++ b/notes.txt @@ -0,0 +1,47 @@ +Roo/E, the Kangaroo Punch Portable GUI Toolkit +Copyright (C) 2022-2025 Scott Duensing + +http://kangaroopunch.com + + +GOALS +~~~~~ +Roo/E is designed to run on any platform that provides at least keyboard and +mouse input with a bitmapped display. The current mimimum target is a 486 +running MS-DOS with a VESA 2.0 compliant SVGA display. + +Unlike most low-end windowing systems, Roo/E should provide "solid" rendering +at all times. Dragging and resizing windows should not result in wireframe +representations of the window, nor should window operations interrupt or pause +the content in the windows. + +Minimized windows display a live view of the entire window contents so users +can continue to monitor their activity. + +Widgets and other window contents should instantly appear. No watching them +being drawn to the display. + +Roo/E should be thread friendly. + + +IMPLEMENTATION +~~~~~~~~~~~~~~ +In order to reach these goals, Roo/E trades memory for speed. Many surface +caches are used to reduce the amount of redraw needed at any given time. + +When new windows are defined, they specify the maximum bounds of their +content. This means no endlessly scrolling windows as is common in word +processors and web browsers. On creation, a new off-screen surface is +created with these maximum bounds. This allows widgets to be drawn in the +window without consideration for clipping. Just keep it inside the window! + +The "window manager" handles window chrome and blitting window contents to yet +another surface. This cached surface is updated when anything in the window is +"damaged". + +Finally, these chromed surfaces (or icons if minimized) are positioned and +dirty rectangles are calculated. Dirty rectangles are then merged, if +possible, to reduce the number of blits required to update the display. This +rectangle list is then blitted from the apporpriate cached surfaces to the +actual display. This results in the entire display being redrawn but with no +overdraw. diff --git a/roo_e/main.c b/roo_e/main.c new file mode 100644 index 0000000..e0f6f4f --- /dev/null +++ b/roo_e/main.c @@ -0,0 +1,223 @@ +/* + * Roo/E, the Kangaroo Punch Portable GUI Toolkit + * Copyright (C) 2026 Scott Duensing + * + * http://kangaroopunch.com + * + * + * This file is part of Roo/E. + * + * Roo/E is free software: you can redistribute it and/or modify it under the + * terms of the GNU Affero General Public License as published by the Free + * Software Foundation, either version 3 of the License, or (at your option) + * any later version. + * + * Roo/E 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 Affero General Public License + * along with Roo/E. If not, see . + */ + + +#include +#include +#include +#include +#include +#include "pthread.h" + +// ReSharper disable once CppUnusedIncludeDirective +#include "stddclmr.h" +#include "util.h" + +#define STB_DS_IMPLEMENTATION +#include "stb_ds.h" + + +int startApp(const char *appName, int argc, char *argv[]); + + +#ifdef PLATFORM_DOS +#include +void *dxeResolver(const char *symbol); +static int lastResort(void); +typedef unsigned int word __attribute__((mode(word))); +typedef unsigned int pointer __attribute__((mode(pointer))); +#pragma GCC diagnostic push +#pragma GCC diagnostic ignored "-Wbuiltin-declaration-mismatch" +struct __emutls_object // NOLINT +{ + word size; + word align; + union { + pointer offset; + void *ptr; + } loc; + void *templ; +}; +void *__emutls_get_address(struct __emutls_object *); // NOLINT +#pragma GCC diagnostic pop +DXE_EXPORT_TABLE(exports) + // dxe + DXE_EXPORT(dlclose) + DXE_EXPORT(dlopen) + DXE_EXPORT(dlregsym) + DXE_EXPORT(dlstatbind) + DXE_EXPORT(dlstatunbind) + // math + DXE_EXPORT(ldexp) + DXE_EXPORT(pow) + // stb_ds + DXE_EXPORT(stbds_arrfreef) + DXE_EXPORT(stbds_arrgrowf) + DXE_EXPORT(stbds_hash_bytes) + DXE_EXPORT(stbds_hash_string) + DXE_EXPORT(stbds_hmdel_key) + DXE_EXPORT(stbds_hmfree_func) + DXE_EXPORT(stbds_hmget_key) + DXE_EXPORT(stbds_hmget_key_ts) + DXE_EXPORT(stbds_hmput_default) + DXE_EXPORT(stbds_hmput_key) + DXE_EXPORT(stbds_rand_seed) + DXE_EXPORT(stbds_shmode_func) + DXE_EXPORT(stbds_stralloc) + DXE_EXPORT(stbds_strreset) + // stdio + DXE_EXPORT(fclose) + DXE_EXPORT(feof) + DXE_EXPORT(ferror) + DXE_EXPORT(fgetc) + DXE_EXPORT(fopen) + DXE_EXPORT(fread) + DXE_EXPORT(fseek) + DXE_EXPORT(ftell) + DXE_EXPORT(printf) + DXE_EXPORT(puts) + DXE_EXPORT(ungetc) + // stdlib + DXE_EXPORT(free) + DXE_EXPORT(malloc) + DXE_EXPORT(realloc) + // string + DXE_EXPORT(memcpy) + DXE_EXPORT(memset) + DXE_EXPORT(strcmp) + DXE_EXPORT(strncmp) + DXE_EXPORT(strtol) + // Compiler stuff. + DXE_EXPORT(__dj_assert) + DXE_EXPORT(__emutls_get_address) +DXE_EXPORT_END + +void *dxeResolver(const char *symbol) { + union { + int (*from)(void); + void *to; + } funcPtrCast; + + printf("%s: undefined symbol in dynamic module!\n", symbol); + + funcPtrCast.from = lastResort; + return funcPtrCast.to; // NOLINT +} + +static int lastResort(void) { + printf("Last resort function called!\n"); + return 0; +} +#endif // PLATFORM_DOS + + +int startApp(const char *appName, const int argc, char *argv[]) { + void *dxe = NULL; + int result = -1; + char *filename = NULL; + int (*appMain)(const int argc, char *argv[]); + + union { + void *from; + int (*to)(const int argc, char *argv[]); + } appPtrCast; + + dlerror(); // Clear any existing error. + filename = utilCreateString("app/%s.app", appName); + dxe = dlopen(filename, RTLD_LAZY); + if (!dxe) { + printf("Unable to load %s! %s\n", filename, dlerror()); + } else { + appPtrCast.from = dlsym(dxe, "_appMain"); + appMain = appPtrCast.to; // NOLINT + result = appMain(argc, argv); + } + + dlclose(dxe); + + return result; +} + + +int main(const int argc, char *argv[]) { + + void *dxe = NULL; + DIR *dir = NULL; + struct dirent *dirent = NULL; + char *filename = NULL; + int (*dynInit)(void); + int result; + + union { + void *from; + int (*to)(void); + } dynPtrCast; + + (void)argc; + (void)argv; + +#ifdef PLATFORM_DOS + // Set the error callback function. + _dlsymresolver = dxeResolver; + // Register the symbols exported into dynamic modules. + dlregsym(exports); +#endif // PLATFORM_DOS + + // Load libraries. + debug("Loading dynamic libraries.\n"); + if ((dir = opendir("dyn/")) == NULL) { + printf("Unable to open dyn directory!\n"); + exit(1); + } + while ((dirent = readdir(dir))) { + if ((dirent->d_type == DT_REG) || (dirent->d_type == DT_LNK)) { + if (utilEndsWith(dirent->d_name, ".dyn", false)) { + filename = utilCreateString("dyn/%s", dirent->d_name); + dlerror(); // Clear any existing error. + dxe = dlopen(filename,RTLD_LAZY | RTLD_GLOBAL); + if (!dxe) { + printf("Unable to load %s! %s\n", filename, dlerror()); + exit(1); + } + debug("Loaded %s.\n", filename); + dynPtrCast.from = dlsym(dxe, "_dynStart"); + if (dynPtrCast.from != NULL) { + dynInit = dynPtrCast.to; // NOLINT + debug("Starting %s.\n", filename); + result = dynInit(); + if (result != 0) { + printf("%s failed to start!\n", filename); + exit(1); + } + } + free(filename); + } + } + } + closedir(dir); + + // Load the Shell. + result = startApp("kpsmpgc", 0, NULL); + + return result; +} diff --git a/roo_e/util.c b/roo_e/util.c new file mode 100644 index 0000000..0d4f44f --- /dev/null +++ b/roo_e/util.c @@ -0,0 +1,80 @@ +/* Roo/E, the Kangaroo Punch Portable GUI Toolkit +* Copyright (C) 2026 Scott Duensing +* +* http://kangaroopunch.com +* +* +* This file is part of Roo/E. +* +* Roo/E is free software: you can redistribute it and/or modify it under the +* terms of the GNU Affero General Public License as published by the Free +* Software Foundation, either version 3 of the License, or (at your option) +* any later version. +* +* Roo/E 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 Affero General Public License +* along with Roo/E. If not, see . +*/ + + +#include +#include +#include + +#include "util.h" + +#include + + +char *utilCreateString(char *format, ...) { + va_list args; + char *string; + + va_start(args, format); + string = utilCreateStringVArgs(format, args); + va_end(args); + + return string; +} + + +__attribute__((__format__(__printf__, 1, 0))) +char *utilCreateStringVArgs(char *format, va_list args) { + va_list argsCopy; + int32_t size = 0; + char *buffer = NULL; + + va_copy(argsCopy, args); + size = vsnprintf(NULL, 0, format, argsCopy) + 1; + va_end(argsCopy); + buffer = calloc(1, (size_t)size); + if (buffer) { + vsnprintf(buffer, (size_t)size, format, args); + } + + return buffer; +} + + +bool utilEndsWith(const char *haystack, const char *needle, const bool caseSensitive) { + size_t h = strlen(haystack); + size_t n = strlen(needle); + + if (h < n) return false; + + while (n > 0) { + h--; + n--; + if (caseSensitive) { + if (haystack[h] != needle[n]) return false; + } else { + if ((toupper(haystack[h]) != toupper(needle[n]))) return false; + } + } + + return true; +} diff --git a/roo_e/util.h b/roo_e/util.h new file mode 100644 index 0000000..5af9e53 --- /dev/null +++ b/roo_e/util.h @@ -0,0 +1,49 @@ +/* Roo/E, the Kangaroo Punch Portable GUI Toolkit +* Copyright (C) 2026 Scott Duensing +* +* http://kangaroopunch.com +* +* +* This file is part of Roo/E. +* +* Roo/E is free software: you can redistribute it and/or modify it under the +* terms of the GNU Affero General Public License as published by the Free +* Software Foundation, either version 3 of the License, or (at your option) +* any later version. +* +* Roo/E 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 Affero General Public License +* along with Roo/E. If not, see . +*/ + + +#ifndef UTIL_H +#define UTIL_H + + +#include +#include +#include + + +#define debug printf + +#ifndef va_copy +#if defined(__GNUC__) || defined(__clang__) +#define va_copy(dest, src) __builtin_va_copy(dest, src) +#else +#define va_copy(dest, src) (dest = src) +#endif +#endif + + +char *utilCreateString(char *format, ...); +char *utilCreateStringVArgs(char *format, va_list args); +bool utilEndsWith(const char *haystack, const char *needle, bool caseSensitive); + + +#endif // UTIL_H