1 \input texinfo @c -*- texinfo -*-
2 @documentencoding UTF-8
4 @settitle Developer Documentation
6 @center @titlefont{Developer Documentation}
13 @chapter Developers Guide
15 @section Notes for external developers
17 This document is mostly useful for internal FFmpeg developers.
18 External developers who need to use the API in their application should
19 refer to the API doxygen documentation in the public headers, and
20 check the examples in @file{doc/examples} and in the source code to
21 see how the public API is employed.
23 You can use the FFmpeg libraries in your commercial program, but you
24 are encouraged to @emph{publish any patch you make}. In this case the
25 best way to proceed is to send your patches to the ffmpeg-devel
26 mailing list following the guidelines illustrated in the remainder of
29 For more detailed legal information about the use of FFmpeg in
30 external programs read the @file{LICENSE} file in the source tree and
31 consult @url{http://ffmpeg.org/legal.html}.
35 There are 3 ways by which code gets into ffmpeg.
37 @item Submitting Patches to the main developer mailing list
38 see @ref{Submitting patches} for details.
39 @item Directly committing changes to the main tree.
40 @item Committing changes to a git clone, for example on github.com or
41 gitorious.org. And asking us to merge these changes.
44 Whichever way, changes should be reviewed by the maintainer of the code
45 before they are committed. And they should follow the @ref{Coding Rules}.
46 The developer making the commit and the author are responsible for their changes
47 and should try to fix issues their commit causes.
52 @subsection Code formatting conventions
54 There are the following guidelines regarding the indentation in files:
61 The TAB character is forbidden outside of Makefiles as is any
62 form of trailing whitespace. Commits containing either will be
63 rejected by the git repository.
66 You should try to limit your code lines to 80 characters; however, do so if
67 and only if this improves readability.
69 The presentation is one inspired by 'indent -i4 -kr -nut'.
71 The main priority in FFmpeg is simplicity and small code size in order to
72 minimize the bug count.
75 Use the JavaDoc/Doxygen format (see examples below) so that code documentation
76 can be generated automatically. All nontrivial functions should have a comment
77 above them explaining what the function does, even if it is just one sentence.
78 All structures and their member variables should be documented, too.
80 Avoid Qt-style and similar Doxygen syntax with @code{!} in it, i.e. replace
81 @code{//!} with @code{///} and similar. Also @@ syntax should be employed
82 for markup commands, i.e. use @code{@@param} and not @code{\param}.
96 typedef struct Foobar @{
97 int var1; /**< var1 description */
98 int var2; ///< var2 description
99 /** var3 description */
107 * @@param my_parameter description of my_parameter
108 * @@return return value description
110 int myfunc(int my_parameter)
114 @subsection C language features
116 FFmpeg is programmed in the ISO C90 language with a few additional
117 features from ISO C99, namely:
121 the @samp{inline} keyword;
127 designated struct initializers (@samp{struct s x = @{ .i = 17 @};})
130 compound literals (@samp{x = (struct s) @{ 17, 23 @};})
133 These features are supported by all compilers we care about, so we will not
134 accept patches to remove their use unless they absolutely do not impair
135 clarity and performance.
137 All code must compile with recent versions of GCC and a number of other
138 currently supported compilers. To ensure compatibility, please do not use
139 additional C99 features or GCC extensions. Especially watch out for:
143 mixing statements and declarations;
146 @samp{long long} (use @samp{int64_t} instead);
149 @samp{__attribute__} not protected by @samp{#ifdef __GNUC__} or similar;
152 GCC statement expressions (@samp{(x = (@{ int y = 4; y; @})}).
155 @subsection Naming conventions
156 All names should be composed with underscores (_), not CamelCase. For example,
157 @samp{avfilter_get_video_buffer} is an acceptable function name and
158 @samp{AVFilterGetVideo} is not. The exception from this are type names, like
159 for example structs and enums; they should always be in the CamelCase
161 There are the following conventions for naming variables and functions:
165 For local variables no prefix is required.
168 For file-scope variables and functions declared as @code{static}, no prefix
172 For variables and functions visible outside of file scope, but only used
173 internally by a library, an @code{ff_} prefix should be used,
174 e.g. @samp{ff_w64_demuxer}.
177 For variables and functions visible outside of file scope, used internally
178 across multiple libraries, use @code{avpriv_} as prefix, for example,
179 @samp{avpriv_aac_parse_header}.
182 Each library has its own prefix for public symbols, in addition to the
183 commonly used @code{av_} (@code{avformat_} for libavformat,
184 @code{avcodec_} for libavcodec, @code{swr_} for libswresample, etc).
185 Check the existing code and choose names accordingly.
186 Note that some symbols without these prefixes are also exported for
187 retro-compatibility reasons. These exceptions are declared in the
188 @code{lib<name>/lib<name>.v} files.
191 Furthermore, name space reserved for the system should not be invaded.
192 Identifiers ending in @code{_t} are reserved by
193 @url{http://pubs.opengroup.org/onlinepubs/007904975/functions/xsh_chap02_02.html#tag_02_02_02, POSIX}.
194 Also avoid names starting with @code{__} or @code{_} followed by an uppercase
195 letter as they are reserved by the C standard. Names starting with @code{_}
196 are reserved at the file level and may not be used for externally visible
197 symbols. If in doubt, just avoid names starting with @code{_} altogether.
199 @subsection Miscellaneous conventions
203 fprintf and printf are forbidden in libavformat and libavcodec,
204 please use av_log() instead.
207 Casts should be used only when necessary. Unneeded parentheses
208 should also be avoided if they don't make the code easier to understand.
211 @subsection Editor configuration
212 In order to configure Vim to follow FFmpeg formatting conventions, paste
213 the following snippet into your @file{.vimrc}:
215 " indentation rules for FFmpeg: 4 spaces, no tabs
221 " Allow tabs in Makefiles.
222 autocmd FileType make,automake set noexpandtab shiftwidth=8 softtabstop=8
223 " Trailing whitespace and tabs are forbidden, so highlight them.
224 highlight ForbiddenWhitespace ctermbg=red guibg=red
225 match ForbiddenWhitespace /\s\+$\|\t/
226 " Do not highlight spaces at the end of line while typing on that line.
227 autocmd InsertEnter * match ForbiddenWhitespace /\t\|\s\+\%#\@@<!$/
230 For Emacs, add these roughly equivalent lines to your @file{.emacs.d/init.el}:
232 (c-add-style "ffmpeg"
235 (indent-tabs-mode . nil)
236 (show-trailing-whitespace . t)
238 (statement-cont . (c-lineup-assignments +)))
241 (setq c-default-style "ffmpeg")
244 @section Development Policy
248 Contributions should be licensed under the
249 @uref{http://www.gnu.org/licenses/lgpl-2.1.html, LGPL 2.1},
250 including an "or any later version" clause, or, if you prefer
251 a gift-style license, the
252 @uref{http://opensource.org/licenses/isc-license.txt, ISC} or
253 @uref{http://mit-license.org/, MIT} license.
254 @uref{http://www.gnu.org/licenses/gpl-2.0.html, GPL 2} including
255 an "or any later version" clause is also acceptable, but LGPL is
257 If you add a new file, give it a proper license header. Do not copy and
258 paste it from a random place, use an existing file as template.
261 You must not commit code which breaks FFmpeg! (Meaning unfinished but
262 enabled code which breaks compilation or compiles but does not work or
263 breaks the regression tests)
264 You can commit unfinished stuff (for testing etc), but it must be disabled
265 (#ifdef etc) by default so it does not interfere with other developers'
269 The commit message should have a short first line in the form of
270 a @samp{topic: short description} as a header, separated by a newline
271 from the body consisting of an explanation of why the change is necessary.
272 If the commit fixes a known bug on the bug tracker, the commit message
273 should include its bug ID. Referring to the issue on the bug tracker does
274 not exempt you from writing an excerpt of the bug in the commit message.
277 You do not have to over-test things. If it works for you, and you think it
278 should work for others, then commit. If your code has problems
279 (portability, triggers compiler bugs, unusual environment etc) they will be
280 reported and eventually fixed.
283 Do not commit unrelated changes together, split them into self-contained
284 pieces. Also do not forget that if part B depends on part A, but A does not
285 depend on B, then A can and should be committed first and separate from B.
286 Keeping changes well split into self-contained parts makes reviewing and
287 understanding them on the commit log mailing list easier. This also helps
288 in case of debugging later on.
289 Also if you have doubts about splitting or not splitting, do not hesitate to
290 ask/discuss it on the developer mailing list.
293 Do not change behavior of the programs (renaming options etc) or public
294 API or ABI without first discussing it on the ffmpeg-devel mailing list.
295 Do not remove functionality from the code. Just improve!
297 Note: Redundant code can be removed.
300 Do not commit changes to the build system (Makefiles, configure script)
301 which change behavior, defaults etc, without asking first. The same
302 applies to compiler warning fixes, trivial looking fixes and to code
303 maintained by other developers. We usually have a reason for doing things
304 the way we do. Send your changes as patches to the ffmpeg-devel mailing
305 list, and if the code maintainers say OK, you may commit. This does not
306 apply to files you wrote and/or maintain.
309 We refuse source indentation and other cosmetic changes if they are mixed
310 with functional changes, such commits will be rejected and removed. Every
311 developer has his own indentation style, you should not change it. Of course
312 if you (re)write something, you can use your own style, even though we would
313 prefer if the indentation throughout FFmpeg was consistent (Many projects
314 force a given indentation style - we do not.). If you really need to make
315 indentation changes (try to avoid this), separate them strictly from real
318 NOTE: If you had to put if()@{ .. @} over a large (> 5 lines) chunk of code,
319 then either do NOT change the indentation of the inner part within (do not
320 move it to the right)! or do so in a separate commit
323 Always fill out the commit log message. Describe in a few lines what you
324 changed and why. You can refer to mailing list postings if you fix a
325 particular bug. Comments such as "fixed!" or "Changed it." are unacceptable.
329 area changed: Short 1 line description
331 details describing what and why and giving references.
335 Make sure the author of the commit is set correctly. (see git commit --author)
336 If you apply a patch, send an
337 answer to ffmpeg-devel (or wherever you got the patch from) saying that
338 you applied the patch.
341 When applying patches that have been discussed (at length) on the mailing
342 list, reference the thread in the log message.
345 Do NOT commit to code actively maintained by others without permission.
346 Send a patch to ffmpeg-devel instead. If no one answers within a reasonable
347 timeframe (12h for build failures and security fixes, 3 days small changes,
348 1 week for big patches) then commit your patch if you think it is OK.
349 Also note, the maintainer can simply ask for more time to review!
352 Subscribe to the ffmpeg-cvslog mailing list. The diffs of all commits
353 are sent there and reviewed by all the other developers. Bugs and possible
354 improvements or general questions regarding commits are discussed there. We
355 expect you to react if problems with your code are uncovered.
358 Update the documentation if you change behavior or add features. If you are
359 unsure how best to do this, send a patch to ffmpeg-devel, the documentation
360 maintainer(s) will review and commit your stuff.
363 Try to keep important discussions and requests (also) on the public
364 developer mailing list, so that all developers can benefit from them.
367 Never write to unallocated memory, never write over the end of arrays,
368 always check values read from some untrusted source before using them
369 as array index or other risky things.
372 Remember to check if you need to bump versions for the specific libav*
373 parts (libavutil, libavcodec, libavformat) you are changing. You need
374 to change the version integer.
375 Incrementing the first component means no backward compatibility to
376 previous versions (e.g. removal of a function from the public API).
377 Incrementing the second component means backward compatible change
378 (e.g. addition of a function to the public API or extension of an
379 existing data structure).
380 Incrementing the third component means a noteworthy binary compatible
381 change (e.g. encoder bug fix that matters for the decoder). The third
382 component always starts at 100 to distinguish FFmpeg from Libav.
385 Compiler warnings indicate potential bugs or code with bad style. If a type of
386 warning always points to correct and clean code, that warning should
387 be disabled, not the code changed.
388 Thus the remaining warnings can either be bugs or correct code.
389 If it is a bug, the bug has to be fixed. If it is not, the code should
390 be changed to not generate a warning unless that causes a slowdown
391 or obfuscates the code.
394 Make sure that no parts of the codebase that you maintain are missing from the
395 @file{MAINTAINERS} file. If something that you want to maintain is missing add it with
397 If at some point you no longer want to maintain some code, then please help
398 finding a new maintainer and also don't forget updating the @file{MAINTAINERS} file.
401 We think our rules are not too hard. If you have comments, contact us.
403 @anchor{Submitting patches}
404 @section Submitting patches
406 First, read the @ref{Coding Rules} above if you did not yet, in particular
407 the rules regarding patch submission.
409 When you submit your patch, please use @code{git format-patch} or
410 @code{git send-email}. We cannot read other diffs :-)
412 Also please do not submit a patch which contains several unrelated changes.
413 Split it into separate, self-contained pieces. This does not mean splitting
414 file by file. Instead, make the patch as small as possible while still
415 keeping it as a logical unit that contains an individual change, even
416 if it spans multiple files. This makes reviewing your patches much easier
417 for us and greatly increases your chances of getting your patch applied.
419 Use the patcheck tool of FFmpeg to check your patch.
420 The tool is located in the tools directory.
422 Run the @ref{Regression tests} before submitting a patch in order to verify
423 it does not cause unexpected problems.
425 It also helps quite a bit if you tell us what the patch does (for example
426 'replaces lrint by lrintf'), and why (for example '*BSD isn't C99 compliant
429 Also please if you send several patches, send each patch as a separate mail,
430 do not attach several unrelated patches to the same mail.
432 Patches should be posted to the
433 @uref{http://lists.ffmpeg.org/mailman/listinfo/ffmpeg-devel, ffmpeg-devel}
434 mailing list. Use @code{git send-email} when possible since it will properly
435 send patches without requiring extra care. If you cannot, then send patches
436 as base64-encoded attachments, so your patch is not trashed during
439 Your patch will be reviewed on the mailing list. You will likely be asked
440 to make some changes and are expected to send in an improved version that
441 incorporates the requests from the review. This process may go through
442 several iterations. Once your patch is deemed good enough, some developer
443 will pick it up and commit it to the official FFmpeg tree.
445 Give us a few days to react. But if some time passes without reaction,
446 send a reminder by email. Your patch should eventually be dealt with.
449 @section New codecs or formats checklist
453 Did you use av_cold for codec initialization and close functions?
456 Did you add a long_name under NULL_IF_CONFIG_SMALL to the AVCodec or
457 AVInputFormat/AVOutputFormat struct?
460 Did you bump the minor version number (and reset the micro version
461 number) in @file{libavcodec/version.h} or @file{libavformat/version.h}?
464 Did you register it in @file{allcodecs.c} or @file{allformats.c}?
467 Did you add the AVCodecID to @file{avcodec.h}?
468 When adding new codec IDs, also add an entry to the codec descriptor
469 list in @file{libavcodec/codec_desc.c}.
472 If it has a FourCC, did you add it to @file{libavformat/riff.c},
473 even if it is only a decoder?
476 Did you add a rule to compile the appropriate files in the Makefile?
477 Remember to do this even if you're just adding a format to a file that is
478 already being compiled by some other rule, like a raw demuxer.
481 Did you add an entry to the table of supported formats or codecs in
482 @file{doc/general.texi}?
485 Did you add an entry in the Changelog?
488 If it depends on a parser or a library, did you add that dependency in
492 Did you @code{git add} the appropriate files before committing?
495 Did you make sure it compiles standalone, i.e. with
496 @code{configure --disable-everything --enable-decoder=foo}
497 (or @code{--enable-demuxer} or whatever your component is)?
501 @section patch submission checklist
505 Does @code{make fate} pass with the patch applied?
508 Was the patch generated with git format-patch or send-email?
511 Did you sign off your patch? (git commit -s)
512 See @url{http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=blob_plain;f=Documentation/SubmittingPatches} for the meaning
516 Did you provide a clear git commit log message?
519 Is the patch against latest FFmpeg git master branch?
522 Are you subscribed to ffmpeg-devel?
523 (the list is subscribers only due to spam)
526 Have you checked that the changes are minimal, so that the same cannot be
527 achieved with a smaller patch and/or simpler final code?
530 If the change is to speed critical code, did you benchmark it?
533 If you did any benchmarks, did you provide them in the mail?
536 Have you checked that the patch does not introduce buffer overflows or
537 other security issues?
540 Did you test your decoder or demuxer against damaged data? If no, see
541 tools/trasher, the noise bitstream filter, and
542 @uref{http://caca.zoy.org/wiki/zzuf, zzuf}. Your decoder or demuxer
543 should not crash, end in a (near) infinite loop, or allocate ridiculous
544 amounts of memory when fed damaged data.
547 Did you test your decoder or demuxer against sample files?
548 Samples may be obtained at @url{http://samples.ffmpeg.org}.
551 Does the patch not mix functional and cosmetic changes?
554 Did you add tabs or trailing whitespace to the code? Both are forbidden.
557 Is the patch attached to the email you send?
560 Is the mime type of the patch correct? It should be text/x-diff or
561 text/x-patch or at least text/plain and not application/octet-stream.
564 If the patch fixes a bug, did you provide a verbose analysis of the bug?
567 If the patch fixes a bug, did you provide enough information, including
568 a sample, so the bug can be reproduced and the fix can be verified?
569 Note please do not attach samples >100k to mails but rather provide a
570 URL, you can upload to ftp://upload.ffmpeg.org
573 Did you provide a verbose summary about what the patch does change?
576 Did you provide a verbose explanation why it changes things like it does?
579 Did you provide a verbose summary of the user visible advantages and
580 disadvantages if the patch is applied?
583 Did you provide an example so we can verify the new feature added by the
587 If you added a new file, did you insert a license header? It should be
588 taken from FFmpeg, not randomly copied and pasted from somewhere else.
591 You should maintain alphabetical order in alphabetically ordered lists as
592 long as doing so does not break API/ABI compatibility.
595 Lines with similar content should be aligned vertically when doing so
596 improves readability.
599 Consider to add a regression test for your code.
602 If you added YASM code please check that things still work with --disable-yasm
605 Make sure you check the return values of function and return appropriate
606 error codes. Especially memory allocation functions like @code{av_malloc()}
607 are notoriously left unchecked, which is a serious problem.
610 Test your code with valgrind and or Address Sanitizer to ensure it's free
611 of leaks, out of array accesses, etc.
614 @section Patch review process
616 All patches posted to ffmpeg-devel will be reviewed, unless they contain a
617 clear note that the patch is not for the git master branch.
618 Reviews and comments will be posted as replies to the patch on the
619 mailing list. The patch submitter then has to take care of every comment,
620 that can be by resubmitting a changed patch or by discussion. Resubmitted
621 patches will themselves be reviewed like any other patch. If at some point
622 a patch passes review with no comments then it is approved, that can for
623 simple and small patches happen immediately while large patches will generally
624 have to be changed and reviewed many times before they are approved.
625 After a patch is approved it will be committed to the repository.
627 We will review all submitted patches, but sometimes we are quite busy so
628 especially for large patches this can take several weeks.
630 If you feel that the review process is too slow and you are willing to try to
631 take over maintainership of the area of code you change then just clone
632 git master and maintain the area of code there. We will merge each area from
633 where its best maintained.
635 When resubmitting patches, please do not make any significant changes
636 not related to the comments received during review. Such patches will
637 be rejected. Instead, submit significant changes or new features as
640 Everyone is welcome to review patches. Also if you are waiting for your patch
641 to be reviewed, please consider helping to review other patches, that is a great
642 way to get everyone's patches reviewed sooner.
644 @anchor{Regression tests}
645 @section Regression tests
647 Before submitting a patch (or committing to the repository), you should at least
648 test that you did not break anything.
650 Running 'make fate' accomplishes this, please see @url{fate.html} for details.
652 [Of course, some patches may change the results of the regression tests. In
653 this case, the reference results of the regression tests shall be modified
656 @subsection Adding files to the fate-suite dataset
658 When there is no muxer or encoder available to generate test media for a
659 specific test then the media has to be included in the fate-suite.
660 First please make sure that the sample file is as small as possible to test the
661 respective decoder or demuxer sufficiently. Large files increase network
662 bandwidth and disk space requirements.
663 Once you have a working fate test and fate sample, provide in the commit
664 message or introductory message for the patch series that you post to
665 the ffmpeg-devel mailing list, a direct link to download the sample media.
668 @subsection Visualizing Test Coverage
670 The FFmpeg build system allows visualizing the test coverage in an easy
671 manner with the coverage tools @code{gcov}/@code{lcov}. This involves
676 Configure to compile with instrumentation enabled:
677 @code{configure --toolchain=gcov}.
680 Run your test case, either manually or via FATE. This can be either
681 the full FATE regression suite, or any arbitrary invocation of any
682 front-end tool provided by FFmpeg, in any combination.
685 Run @code{make lcov} to generate coverage data in HTML format.
688 View @code{lcov/index.html} in your preferred HTML viewer.
691 You can use the command @code{make lcov-reset} to reset the coverage
692 measurements. You will need to rerun @code{make lcov} after running a
695 @subsection Using Valgrind
697 The configure script provides a shortcut for using valgrind to spot bugs
698 related to memory handling. Just add the option
699 @code{--toolchain=valgrind-memcheck} or @code{--toolchain=valgrind-massif}
700 to your configure line, and reasonable defaults will be set for running
701 FATE under the supervision of either the @strong{memcheck} or the
702 @strong{massif} tool of the valgrind suite.
704 In case you need finer control over how valgrind is invoked, use the
705 @code{--target-exec='valgrind <your_custom_valgrind_options>} option in
706 your configure line instead.
708 @anchor{Release process}
709 @section Release process
711 FFmpeg maintains a set of @strong{release branches}, which are the
712 recommended deliverable for system integrators and distributors (such as
713 Linux distributions, etc.). At regular times, a @strong{release
714 manager} prepares, tests and publishes tarballs on the
715 @url{http://ffmpeg.org} website.
717 There are two kinds of releases:
721 @strong{Major releases} always include the latest and greatest
722 features and functionality.
725 @strong{Point releases} are cut from @strong{release} branches,
726 which are named @code{release/X}, with @code{X} being the release
730 Note that we promise to our users that shared libraries from any FFmpeg
731 release never break programs that have been @strong{compiled} against
732 previous versions of @strong{the same release series} in any case!
734 However, from time to time, we do make API changes that require adaptations
735 in applications. Such changes are only allowed in (new) major releases and
736 require further steps such as bumping library version numbers and/or
737 adjustments to the symbol versioning file. Please discuss such changes
738 on the @strong{ffmpeg-devel} mailing list in time to allow forward planning.
740 @anchor{Criteria for Point Releases}
741 @subsection Criteria for Point Releases
743 Changes that match the following criteria are valid candidates for
744 inclusion into a point release:
748 Fixes a security issue, preferably identified by a @strong{CVE
749 number} issued by @url{http://cve.mitre.org/}.
752 Fixes a documented bug in @url{https://trac.ffmpeg.org}.
755 Improves the included documentation.
758 Retains both source code and binary compatibility with previous
759 point releases of the same release branch.
762 The order for checking the rules is (1 OR 2 OR 3) AND 4.
765 @subsection Release Checklist
767 The release process involves the following steps:
771 Ensure that the @file{RELEASE} file contains the version number for
772 the upcoming release.
775 Add the release at @url{https://trac.ffmpeg.org/admin/ticket/versions}.
778 Announce the intent to do a release to the mailing list.
781 Make sure all relevant security fixes have been backported. See
782 @url{https://ffmpeg.org/security.html}.
785 Ensure that the FATE regression suite still passes in the release
786 branch on at least @strong{i386} and @strong{amd64}
787 (cf. @ref{Regression tests}).
790 Prepare the release tarballs in @code{bz2} and @code{gz} formats, and
791 supplementing files that contain @code{gpg} signatures
794 Publish the tarballs at @url{http://ffmpeg.org/releases}. Create and
795 push an annotated tag in the form @code{nX}, with @code{X}
796 containing the version number.
799 Propose and send a patch to the @strong{ffmpeg-devel} mailing list
800 with a news entry for the website.
803 Publish the news entry.
806 Send announcement to the mailing list.