1\input texinfo @c -*- texinfo -*- 2@documentencoding UTF-8 3 4@settitle Developer Documentation 5@titlepage 6@center @titlefont{Developer Documentation} 7@end titlepage 8 9@top 10 11@contents 12 13@chapter Notes for external developers 14 15This document is mostly useful for internal FFmpeg developers. 16External developers who need to use the API in their application should 17refer to the API doxygen documentation in the public headers, and 18check the examples in @file{doc/examples} and in the source code to 19see how the public API is employed. 20 21You can use the FFmpeg libraries in your commercial program, but you 22are encouraged to @emph{publish any patch you make}. In this case the 23best way to proceed is to send your patches to the ffmpeg-devel 24mailing list following the guidelines illustrated in the remainder of 25this document. 26 27For more detailed legal information about the use of FFmpeg in 28external programs read the @file{LICENSE} file in the source tree and 29consult @url{https://ffmpeg.org/legal.html}. 30 31@chapter Contributing 32 33There are 2 ways by which code gets into FFmpeg: 34@itemize @bullet 35@item Submitting patches to the ffmpeg-devel mailing list. 36 See @ref{Submitting patches} for details. 37@item Directly committing changes to the main tree. 38@end itemize 39 40Whichever way, changes should be reviewed by the maintainer of the code 41before they are committed. And they should follow the @ref{Coding Rules}. 42The developer making the commit and the author are responsible for their changes 43and should try to fix issues their commit causes. 44 45@anchor{Coding Rules} 46@chapter Coding Rules 47 48@section Code formatting conventions 49 50There are the following guidelines regarding the indentation in files: 51 52@itemize @bullet 53@item 54Indent size is 4. 55 56@item 57The TAB character is forbidden outside of Makefiles as is any 58form of trailing whitespace. Commits containing either will be 59rejected by the git repository. 60 61@item 62You should try to limit your code lines to 80 characters; however, do so if 63and only if this improves readability. 64 65@item 66K&R coding style is used. 67@end itemize 68The presentation is one inspired by 'indent -i4 -kr -nut'. 69 70The main priority in FFmpeg is simplicity and small code size in order to 71minimize the bug count. 72 73@section Comments 74Use the JavaDoc/Doxygen format (see examples below) so that code documentation 75can be generated automatically. All nontrivial functions should have a comment 76above them explaining what the function does, even if it is just one sentence. 77All structures and their member variables should be documented, too. 78 79Avoid Qt-style and similar Doxygen syntax with @code{!} in it, i.e. replace 80@code{//!} with @code{///} and similar. Also @@ syntax should be employed 81for markup commands, i.e. use @code{@@param} and not @code{\param}. 82 83@example 84/** 85 * @@file 86 * MPEG codec. 87 * @@author ... 88 */ 89 90/** 91 * Summary sentence. 92 * more text ... 93 * ... 94 */ 95typedef struct Foobar @{ 96 int var1; /**< var1 description */ 97 int var2; ///< var2 description 98 /** var3 description */ 99 int var3; 100@} Foobar; 101 102/** 103 * Summary sentence. 104 * more text ... 105 * ... 106 * @@param my_parameter description of my_parameter 107 * @@return return value description 108 */ 109int myfunc(int my_parameter) 110... 111@end example 112 113@section C language features 114 115FFmpeg is programmed in the ISO C90 language with a few additional 116features from ISO C99, namely: 117 118@itemize @bullet 119@item 120the @samp{inline} keyword; 121 122@item 123@samp{//} comments; 124 125@item 126designated struct initializers (@samp{struct s x = @{ .i = 17 @};}); 127 128@item 129compound literals (@samp{x = (struct s) @{ 17, 23 @};}). 130 131@item 132for loops with variable definition (@samp{for (int i = 0; i < 8; i++)}); 133 134@item 135Variadic macros (@samp{#define ARRAY(nb, ...) (int[nb + 1])@{ nb, __VA_ARGS__ @}}); 136 137@item 138Implementation defined behavior for signed integers is assumed to match the 139expected behavior for two's complement. Non representable values in integer 140casts are binary truncated. Shift right of signed values uses sign extension. 141@end itemize 142 143These features are supported by all compilers we care about, so we will not 144accept patches to remove their use unless they absolutely do not impair 145clarity and performance. 146 147All code must compile with recent versions of GCC and a number of other 148currently supported compilers. To ensure compatibility, please do not use 149additional C99 features or GCC extensions. Especially watch out for: 150 151@itemize @bullet 152@item 153mixing statements and declarations; 154 155@item 156@samp{long long} (use @samp{int64_t} instead); 157 158@item 159@samp{__attribute__} not protected by @samp{#ifdef __GNUC__} or similar; 160 161@item 162GCC statement expressions (@samp{(x = (@{ int y = 4; y; @})}). 163@end itemize 164 165@section Naming conventions 166All names should be composed with underscores (_), not CamelCase. For example, 167@samp{avfilter_get_video_buffer} is an acceptable function name and 168@samp{AVFilterGetVideo} is not. The exception from this are type names, like 169for example structs and enums; they should always be in CamelCase. 170 171There are the following conventions for naming variables and functions: 172 173@itemize @bullet 174@item 175For local variables no prefix is required. 176 177@item 178For file-scope variables and functions declared as @code{static}, no prefix 179is required. 180 181@item 182For variables and functions visible outside of file scope, but only used 183internally by a library, an @code{ff_} prefix should be used, 184e.g. @samp{ff_w64_demuxer}. 185 186@item 187For variables and functions visible outside of file scope, used internally 188across multiple libraries, use @code{avpriv_} as prefix, for example, 189@samp{avpriv_report_missing_feature}. 190 191@item 192Each library has its own prefix for public symbols, in addition to the 193commonly used @code{av_} (@code{avformat_} for libavformat, 194@code{avcodec_} for libavcodec, @code{swr_} for libswresample, etc). 195Check the existing code and choose names accordingly. 196Note that some symbols without these prefixes are also exported for 197retro-compatibility reasons. These exceptions are declared in the 198@code{lib<name>/lib<name>.v} files. 199@end itemize 200 201Furthermore, name space reserved for the system should not be invaded. 202Identifiers ending in @code{_t} are reserved by 203@url{http://pubs.opengroup.org/onlinepubs/007904975/functions/xsh_chap02_02.html#tag_02_02_02, POSIX}. 204Also avoid names starting with @code{__} or @code{_} followed by an uppercase 205letter as they are reserved by the C standard. Names starting with @code{_} 206are reserved at the file level and may not be used for externally visible 207symbols. If in doubt, just avoid names starting with @code{_} altogether. 208 209@section Miscellaneous conventions 210 211@itemize @bullet 212@item 213fprintf and printf are forbidden in libavformat and libavcodec, 214please use av_log() instead. 215 216@item 217Casts should be used only when necessary. Unneeded parentheses 218should also be avoided if they don't make the code easier to understand. 219@end itemize 220 221@section Editor configuration 222In order to configure Vim to follow FFmpeg formatting conventions, paste 223the following snippet into your @file{.vimrc}: 224@example 225" indentation rules for FFmpeg: 4 spaces, no tabs 226set expandtab 227set shiftwidth=4 228set softtabstop=4 229set cindent 230set cinoptions=(0 231" Allow tabs in Makefiles. 232autocmd FileType make,automake set noexpandtab shiftwidth=8 softtabstop=8 233" Trailing whitespace and tabs are forbidden, so highlight them. 234highlight ForbiddenWhitespace ctermbg=red guibg=red 235match ForbiddenWhitespace /\s\+$\|\t/ 236" Do not highlight spaces at the end of line while typing on that line. 237autocmd InsertEnter * match ForbiddenWhitespace /\t\|\s\+\%#\@@<!$/ 238@end example 239 240For Emacs, add these roughly equivalent lines to your @file{.emacs.d/init.el}: 241@lisp 242(c-add-style "ffmpeg" 243 '("k&r" 244 (c-basic-offset . 4) 245 (indent-tabs-mode . nil) 246 (show-trailing-whitespace . t) 247 (c-offsets-alist 248 (statement-cont . (c-lineup-assignments +))) 249 ) 250 ) 251(setq c-default-style "ffmpeg") 252@end lisp 253 254@chapter Development Policy 255 256@section Patches/Committing 257@subheading Licenses for patches must be compatible with FFmpeg. 258Contributions should be licensed under the 259@uref{http://www.gnu.org/licenses/lgpl-2.1.html, LGPL 2.1}, 260including an "or any later version" clause, or, if you prefer 261a gift-style license, the 262@uref{http://opensource.org/licenses/isc-license.txt, ISC} or 263@uref{http://mit-license.org/, MIT} license. 264@uref{http://www.gnu.org/licenses/gpl-2.0.html, GPL 2} including 265an "or any later version" clause is also acceptable, but LGPL is 266preferred. 267If you add a new file, give it a proper license header. Do not copy and 268paste it from a random place, use an existing file as template. 269 270@subheading You must not commit code which breaks FFmpeg! 271This means unfinished code which is enabled and breaks compilation, 272or compiles but does not work/breaks the regression tests. Code which 273is unfinished but disabled may be permitted under-circumstances, like 274missing samples or an implementation with a small subset of features. 275Always check the mailing list for any reviewers with issues and test 276FATE before you push. 277 278@subheading Keep the main commit message short with an extended description below. 279The commit message should have a short first line in the form of 280a @samp{topic: short description} as a header, separated by a newline 281from the body consisting of an explanation of why the change is necessary. 282If the commit fixes a known bug on the bug tracker, the commit message 283should include its bug ID. Referring to the issue on the bug tracker does 284not exempt you from writing an excerpt of the bug in the commit message. 285 286@subheading Testing must be adequate but not excessive. 287If it works for you, others, and passes FATE then it should be OK to commit 288it, provided it fits the other committing criteria. You should not worry about 289over-testing things. If your code has problems (portability, triggers 290compiler bugs, unusual environment etc) they will be reported and eventually 291fixed. 292 293@subheading Do not commit unrelated changes together. 294They should be split them into self-contained pieces. Also do not forget 295that if part B depends on part A, but A does not depend on B, then A can 296and should be committed first and separate from B. Keeping changes well 297split into self-contained parts makes reviewing and understanding them on 298the commit log mailing list easier. This also helps in case of debugging 299later on. 300Also if you have doubts about splitting or not splitting, do not hesitate to 301ask/discuss it on the developer mailing list. 302 303@subheading Ask before you change the build system (configure, etc). 304Do not commit changes to the build system (Makefiles, configure script) 305which change behavior, defaults etc, without asking first. The same 306applies to compiler warning fixes, trivial looking fixes and to code 307maintained by other developers. We usually have a reason for doing things 308the way we do. Send your changes as patches to the ffmpeg-devel mailing 309list, and if the code maintainers say OK, you may commit. This does not 310apply to files you wrote and/or maintain. 311 312@subheading Cosmetic changes should be kept in separate patches. 313We refuse source indentation and other cosmetic changes if they are mixed 314with functional changes, such commits will be rejected and removed. Every 315developer has his own indentation style, you should not change it. Of course 316if you (re)write something, you can use your own style, even though we would 317prefer if the indentation throughout FFmpeg was consistent (Many projects 318force a given indentation style - we do not.). If you really need to make 319indentation changes (try to avoid this), separate them strictly from real 320changes. 321 322NOTE: If you had to put if()@{ .. @} over a large (> 5 lines) chunk of code, 323then either do NOT change the indentation of the inner part within (do not 324move it to the right)! or do so in a separate commit 325 326@subheading Commit messages should always be filled out properly. 327Always fill out the commit log message. Describe in a few lines what you 328changed and why. You can refer to mailing list postings if you fix a 329particular bug. Comments such as "fixed!" or "Changed it." are unacceptable. 330Recommended format: 331 332@example 333area changed: Short 1 line description 334 335details describing what and why and giving references. 336@end example 337 338@subheading Credit the author of the patch. 339Make sure the author of the commit is set correctly. (see git commit --author) 340If you apply a patch, send an 341answer to ffmpeg-devel (or wherever you got the patch from) saying that 342you applied the patch. 343 344@subheading Complex patches should refer to discussion surrounding them. 345When applying patches that have been discussed (at length) on the mailing 346list, reference the thread in the log message. 347 348@subheading Always wait long enough before pushing changes 349Do NOT commit to code actively maintained by others without permission. 350Send a patch to ffmpeg-devel. If no one answers within a reasonable 351time-frame (12h for build failures and security fixes, 3 days small changes, 3521 week for big patches) then commit your patch if you think it is OK. 353Also note, the maintainer can simply ask for more time to review! 354 355@section Code 356@subheading API/ABI changes should be discussed before they are made. 357Do not change behavior of the programs (renaming options etc) or public 358API or ABI without first discussing it on the ffmpeg-devel mailing list. 359Do not remove widely used functionality or features (redundant code can be removed). 360 361@subheading Remember to check if you need to bump versions for libav*. 362Depending on the change, you may need to change the version integer. 363Incrementing the first component means no backward compatibility to 364previous versions (e.g. removal of a function from the public API). 365Incrementing the second component means backward compatible change 366(e.g. addition of a function to the public API or extension of an 367existing data structure). 368Incrementing the third component means a noteworthy binary compatible 369change (e.g. encoder bug fix that matters for the decoder). The third 370component always starts at 100 to distinguish FFmpeg from Libav. 371 372@subheading Warnings for correct code may be disabled if there is no other option. 373Compiler warnings indicate potential bugs or code with bad style. If a type of 374warning always points to correct and clean code, that warning should 375be disabled, not the code changed. 376Thus the remaining warnings can either be bugs or correct code. 377If it is a bug, the bug has to be fixed. If it is not, the code should 378be changed to not generate a warning unless that causes a slowdown 379or obfuscates the code. 380 381@subheading Check untrusted input properly. 382Never write to unallocated memory, never write over the end of arrays, 383always check values read from some untrusted source before using them 384as array index or other risky things. 385 386@section Documentation/Other 387@subheading Subscribe to the ffmpeg-devel mailing list. 388It is important to be subscribed to the 389@uref{https://lists.ffmpeg.org/mailman/listinfo/ffmpeg-devel, ffmpeg-devel} 390mailing list. Almost any non-trivial patch is to be sent there for review. 391Other developers may have comments about your contribution. We expect you see 392those comments, and to improve it if requested. (N.B. Experienced committers 393have other channels, and may sometimes skip review for trivial fixes.) Also, 394discussion here about bug fixes and FFmpeg improvements by other developers may 395be helpful information for you. Finally, by being a list subscriber, your 396contribution will be posted immediately to the list, without the moderation 397hold which messages from non-subscribers experience. 398 399However, it is more important to the project that we receive your patch than 400that you be subscribed to the ffmpeg-devel list. If you have a patch, and don't 401want to subscribe and discuss the patch, then please do send it to the list 402anyway. 403 404@subheading Subscribe to the ffmpeg-cvslog mailing list. 405Diffs of all commits are sent to the 406@uref{https://lists.ffmpeg.org/mailman/listinfo/ffmpeg-cvslog, ffmpeg-cvslog} 407mailing list. Some developers read this list to review all code base changes 408from all sources. Subscribing to this list is not mandatory. 409 410@subheading Keep the documentation up to date. 411Update the documentation if you change behavior or add features. If you are 412unsure how best to do this, send a patch to ffmpeg-devel, the documentation 413maintainer(s) will review and commit your stuff. 414 415@subheading Important discussions should be accessible to all. 416Try to keep important discussions and requests (also) on the public 417developer mailing list, so that all developers can benefit from them. 418 419@subheading Check your entries in MAINTAINERS. 420Make sure that no parts of the codebase that you maintain are missing from the 421@file{MAINTAINERS} file. If something that you want to maintain is missing add it with 422your name after it. 423If at some point you no longer want to maintain some code, then please help in 424finding a new maintainer and also don't forget to update the @file{MAINTAINERS} file. 425 426We think our rules are not too hard. If you have comments, contact us. 427 428@chapter Code of conduct 429 430Be friendly and respectful towards others and third parties. 431Treat others the way you yourself want to be treated. 432 433Be considerate. Not everyone shares the same viewpoint and priorities as you do. 434Different opinions and interpretations help the project. 435Looking at issues from a different perspective assists development. 436 437Do not assume malice for things that can be attributed to incompetence. Even if 438it is malice, it's rarely good to start with that as initial assumption. 439 440Stay friendly even if someone acts contrarily. Everyone has a bad day 441once in a while. 442If you yourself have a bad day or are angry then try to take a break and reply 443once you are calm and without anger if you have to. 444 445Try to help other team members and cooperate if you can. 446 447The goal of software development is to create technical excellence, not for any 448individual to be better and "win" against the others. Large software projects 449are only possible and successful through teamwork. 450 451If someone struggles do not put them down. Give them a helping hand 452instead and point them in the right direction. 453 454Finally, keep in mind the immortal words of Bill and Ted, 455"Be excellent to each other." 456 457@anchor{Submitting patches} 458@chapter Submitting patches 459 460First, read the @ref{Coding Rules} above if you did not yet, in particular 461the rules regarding patch submission. 462 463When you submit your patch, please use @code{git format-patch} or 464@code{git send-email}. We cannot read other diffs :-). 465 466Also please do not submit a patch which contains several unrelated changes. 467Split it into separate, self-contained pieces. This does not mean splitting 468file by file. Instead, make the patch as small as possible while still 469keeping it as a logical unit that contains an individual change, even 470if it spans multiple files. This makes reviewing your patches much easier 471for us and greatly increases your chances of getting your patch applied. 472 473Use the patcheck tool of FFmpeg to check your patch. 474The tool is located in the tools directory. 475 476Run the @ref{Regression tests} before submitting a patch in order to verify 477it does not cause unexpected problems. 478 479It also helps quite a bit if you tell us what the patch does (for example 480'replaces lrint by lrintf'), and why (for example '*BSD isn't C99 compliant 481and has no lrint()') 482 483Also please if you send several patches, send each patch as a separate mail, 484do not attach several unrelated patches to the same mail. 485 486Patches should be posted to the 487@uref{https://lists.ffmpeg.org/mailman/listinfo/ffmpeg-devel, ffmpeg-devel} 488mailing list. Use @code{git send-email} when possible since it will properly 489send patches without requiring extra care. If you cannot, then send patches 490as base64-encoded attachments, so your patch is not trashed during 491transmission. Also ensure the correct mime type is used 492(text/x-diff or text/x-patch or at least text/plain) and that only one 493patch is inline or attached per mail. 494You can check @url{https://patchwork.ffmpeg.org}, if your patch does not show up, its mime type 495likely was wrong. 496 497Your patch will be reviewed on the mailing list. You will likely be asked 498to make some changes and are expected to send in an improved version that 499incorporates the requests from the review. This process may go through 500several iterations. Once your patch is deemed good enough, some developer 501will pick it up and commit it to the official FFmpeg tree. 502 503Give us a few days to react. But if some time passes without reaction, 504send a reminder by email. Your patch should eventually be dealt with. 505 506 507@chapter New codecs or formats checklist 508 509@enumerate 510@item 511Did you use av_cold for codec initialization and close functions? 512 513@item 514Did you add a long_name under NULL_IF_CONFIG_SMALL to the AVCodec or 515AVInputFormat/AVOutputFormat struct? 516 517@item 518Did you bump the minor version number (and reset the micro version 519number) in @file{libavcodec/version.h} or @file{libavformat/version.h}? 520 521@item 522Did you register it in @file{allcodecs.c} or @file{allformats.c}? 523 524@item 525Did you add the AVCodecID to @file{avcodec.h}? 526When adding new codec IDs, also add an entry to the codec descriptor 527list in @file{libavcodec/codec_desc.c}. 528 529@item 530If it has a FourCC, did you add it to @file{libavformat/riff.c}, 531even if it is only a decoder? 532 533@item 534Did you add a rule to compile the appropriate files in the Makefile? 535Remember to do this even if you're just adding a format to a file that is 536already being compiled by some other rule, like a raw demuxer. 537 538@item 539Did you add an entry to the table of supported formats or codecs in 540@file{doc/general.texi}? 541 542@item 543Did you add an entry in the Changelog? 544 545@item 546If it depends on a parser or a library, did you add that dependency in 547configure? 548 549@item 550Did you @code{git add} the appropriate files before committing? 551 552@item 553Did you make sure it compiles standalone, i.e. with 554@code{configure --disable-everything --enable-decoder=foo} 555(or @code{--enable-demuxer} or whatever your component is)? 556@end enumerate 557 558 559@chapter Patch submission checklist 560 561@enumerate 562@item 563Does @code{make fate} pass with the patch applied? 564 565@item 566Was the patch generated with git format-patch or send-email? 567 568@item 569Did you sign-off your patch? (@code{git commit -s}) 570See @uref{https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/plain/Documentation/process/submitting-patches.rst, Sign your work} for the meaning 571of @dfn{sign-off}. 572 573@item 574Did you provide a clear git commit log message? 575 576@item 577Is the patch against latest FFmpeg git master branch? 578 579@item 580Are you subscribed to ffmpeg-devel? 581(the list is subscribers only due to spam) 582 583@item 584Have you checked that the changes are minimal, so that the same cannot be 585achieved with a smaller patch and/or simpler final code? 586 587@item 588If the change is to speed critical code, did you benchmark it? 589 590@item 591If you did any benchmarks, did you provide them in the mail? 592 593@item 594Have you checked that the patch does not introduce buffer overflows or 595other security issues? 596 597@item 598Did you test your decoder or demuxer against damaged data? If no, see 599tools/trasher, the noise bitstream filter, and 600@uref{http://caca.zoy.org/wiki/zzuf, zzuf}. Your decoder or demuxer 601should not crash, end in a (near) infinite loop, or allocate ridiculous 602amounts of memory when fed damaged data. 603 604@item 605Did you test your decoder or demuxer against sample files? 606Samples may be obtained at @url{https://samples.ffmpeg.org}. 607 608@item 609Does the patch not mix functional and cosmetic changes? 610 611@item 612Did you add tabs or trailing whitespace to the code? Both are forbidden. 613 614@item 615Is the patch attached to the email you send? 616 617@item 618Is the mime type of the patch correct? It should be text/x-diff or 619text/x-patch or at least text/plain and not application/octet-stream. 620 621@item 622If the patch fixes a bug, did you provide a verbose analysis of the bug? 623 624@item 625If the patch fixes a bug, did you provide enough information, including 626a sample, so the bug can be reproduced and the fix can be verified? 627Note please do not attach samples >100k to mails but rather provide a 628URL, you can upload to @url{https://streams.videolan.org/upload/}. 629 630@item 631Did you provide a verbose summary about what the patch does change? 632 633@item 634Did you provide a verbose explanation why it changes things like it does? 635 636@item 637Did you provide a verbose summary of the user visible advantages and 638disadvantages if the patch is applied? 639 640@item 641Did you provide an example so we can verify the new feature added by the 642patch easily? 643 644@item 645If you added a new file, did you insert a license header? It should be 646taken from FFmpeg, not randomly copied and pasted from somewhere else. 647 648@item 649You should maintain alphabetical order in alphabetically ordered lists as 650long as doing so does not break API/ABI compatibility. 651 652@item 653Lines with similar content should be aligned vertically when doing so 654improves readability. 655 656@item 657Consider adding a regression test for your code. 658 659@item 660If you added YASM code please check that things still work with --disable-yasm. 661 662@item 663Make sure you check the return values of function and return appropriate 664error codes. Especially memory allocation functions like @code{av_malloc()} 665are notoriously left unchecked, which is a serious problem. 666 667@item 668Test your code with valgrind and or Address Sanitizer to ensure it's free 669of leaks, out of array accesses, etc. 670@end enumerate 671 672@chapter Patch review process 673 674All patches posted to ffmpeg-devel will be reviewed, unless they contain a 675clear note that the patch is not for the git master branch. 676Reviews and comments will be posted as replies to the patch on the 677mailing list. The patch submitter then has to take care of every comment, 678that can be by resubmitting a changed patch or by discussion. Resubmitted 679patches will themselves be reviewed like any other patch. If at some point 680a patch passes review with no comments then it is approved, that can for 681simple and small patches happen immediately while large patches will generally 682have to be changed and reviewed many times before they are approved. 683After a patch is approved it will be committed to the repository. 684 685We will review all submitted patches, but sometimes we are quite busy so 686especially for large patches this can take several weeks. 687 688If you feel that the review process is too slow and you are willing to try to 689take over maintainership of the area of code you change then just clone 690git master and maintain the area of code there. We will merge each area from 691where its best maintained. 692 693When resubmitting patches, please do not make any significant changes 694not related to the comments received during review. Such patches will 695be rejected. Instead, submit significant changes or new features as 696separate patches. 697 698Everyone is welcome to review patches. Also if you are waiting for your patch 699to be reviewed, please consider helping to review other patches, that is a great 700way to get everyone's patches reviewed sooner. 701 702@anchor{Regression tests} 703@chapter Regression tests 704 705Before submitting a patch (or committing to the repository), you should at least 706test that you did not break anything. 707 708Running 'make fate' accomplishes this, please see @url{fate.html} for details. 709 710[Of course, some patches may change the results of the regression tests. In 711this case, the reference results of the regression tests shall be modified 712accordingly]. 713 714@section Adding files to the fate-suite dataset 715 716When there is no muxer or encoder available to generate test media for a 717specific test then the media has to be included in the fate-suite. 718First please make sure that the sample file is as small as possible to test the 719respective decoder or demuxer sufficiently. Large files increase network 720bandwidth and disk space requirements. 721Once you have a working fate test and fate sample, provide in the commit 722message or introductory message for the patch series that you post to 723the ffmpeg-devel mailing list, a direct link to download the sample media. 724 725@section Visualizing Test Coverage 726 727The FFmpeg build system allows visualizing the test coverage in an easy 728manner with the coverage tools @code{gcov}/@code{lcov}. This involves 729the following steps: 730 731@enumerate 732@item 733 Configure to compile with instrumentation enabled: 734 @code{configure --toolchain=gcov}. 735 736@item 737 Run your test case, either manually or via FATE. This can be either 738 the full FATE regression suite, or any arbitrary invocation of any 739 front-end tool provided by FFmpeg, in any combination. 740 741@item 742 Run @code{make lcov} to generate coverage data in HTML format. 743 744@item 745 View @code{lcov/index.html} in your preferred HTML viewer. 746@end enumerate 747 748You can use the command @code{make lcov-reset} to reset the coverage 749measurements. You will need to rerun @code{make lcov} after running a 750new test. 751 752@section Using Valgrind 753 754The configure script provides a shortcut for using valgrind to spot bugs 755related to memory handling. Just add the option 756@code{--toolchain=valgrind-memcheck} or @code{--toolchain=valgrind-massif} 757to your configure line, and reasonable defaults will be set for running 758FATE under the supervision of either the @strong{memcheck} or the 759@strong{massif} tool of the valgrind suite. 760 761In case you need finer control over how valgrind is invoked, use the 762@code{--target-exec='valgrind <your_custom_valgrind_options>} option in 763your configure line instead. 764 765@anchor{Release process} 766@chapter Release process 767 768FFmpeg maintains a set of @strong{release branches}, which are the 769recommended deliverable for system integrators and distributors (such as 770Linux distributions, etc.). At regular times, a @strong{release 771manager} prepares, tests and publishes tarballs on the 772@url{https://ffmpeg.org} website. 773 774There are two kinds of releases: 775 776@enumerate 777@item 778@strong{Major releases} always include the latest and greatest 779features and functionality. 780 781@item 782@strong{Point releases} are cut from @strong{release} branches, 783which are named @code{release/X}, with @code{X} being the release 784version number. 785@end enumerate 786 787Note that we promise to our users that shared libraries from any FFmpeg 788release never break programs that have been @strong{compiled} against 789previous versions of @strong{the same release series} in any case! 790 791However, from time to time, we do make API changes that require adaptations 792in applications. Such changes are only allowed in (new) major releases and 793require further steps such as bumping library version numbers and/or 794adjustments to the symbol versioning file. Please discuss such changes 795on the @strong{ffmpeg-devel} mailing list in time to allow forward planning. 796 797@anchor{Criteria for Point Releases} 798@section Criteria for Point Releases 799 800Changes that match the following criteria are valid candidates for 801inclusion into a point release: 802 803@enumerate 804@item 805Fixes a security issue, preferably identified by a @strong{CVE 806number} issued by @url{http://cve.mitre.org/}. 807 808@item 809Fixes a documented bug in @url{https://trac.ffmpeg.org}. 810 811@item 812Improves the included documentation. 813 814@item 815Retains both source code and binary compatibility with previous 816point releases of the same release branch. 817@end enumerate 818 819The order for checking the rules is (1 OR 2 OR 3) AND 4. 820 821 822@section Release Checklist 823 824The release process involves the following steps: 825 826@enumerate 827@item 828Ensure that the @file{RELEASE} file contains the version number for 829the upcoming release. 830 831@item 832Add the release at @url{https://trac.ffmpeg.org/admin/ticket/versions}. 833 834@item 835Announce the intent to do a release to the mailing list. 836 837@item 838Make sure all relevant security fixes have been backported. See 839@url{https://ffmpeg.org/security.html}. 840 841@item 842Ensure that the FATE regression suite still passes in the release 843branch on at least @strong{i386} and @strong{amd64} 844(cf. @ref{Regression tests}). 845 846@item 847Prepare the release tarballs in @code{bz2} and @code{gz} formats, and 848supplementing files that contain @code{gpg} signatures 849 850@item 851Publish the tarballs at @url{https://ffmpeg.org/releases}. Create and 852push an annotated tag in the form @code{nX}, with @code{X} 853containing the version number. 854 855@item 856Propose and send a patch to the @strong{ffmpeg-devel} mailing list 857with a news entry for the website. 858 859@item 860Publish the news entry. 861 862@item 863Send an announcement to the mailing list. 864@end enumerate 865 866@bye 867