Developer Documentation

libavcodec is the library containing the codecs (both encoding and decoding). Look at ‘libavcodec/apiexample.c’ to see how to use it.
libavformat is the library containing the file format handling (mux and demux code for several formats). Look at ‘ffplay.c’ to use it in a player. See ‘libavformat/output-example.c’ to use it to generate audio or video streams.

1.2 Integrating libavcodec or libavformat in your program

You can integrate all the source code of the libraries to link them statically to avoid any version problem. All you need is to provide a ’config.mak’ and a ’config.h’ in the parent directory. See the defines generated by ./configure to understand what is needed.

You can use libavcodec or libavformat in your commercial program, but any patch you make must be published. The best way to proceed is to send your patches to the FFmpeg mailing list.

1.3 Coding Rules

FFmpeg is programmed in the ISO C90 language with a few additional features from ISO C99, namely:

the ‘inline’ keyword;
‘//’ comments;
designated struct initializers (‘struct s x = { .i = 17 };’)
compound literals (‘x = (struct s) { 17, 23 };’)

These features are supported by all compilers we care about, so we will not accept patches to remove their use unless they absolutely do not impair clarity and performance.

All code must compile with GCC 2.95 and GCC 3.3. Currently, FFmpeg also compiles with several other compilers, such as the Compaq ccc compiler or Sun Studio 9, and we would like to keep it that way unless it would be exceedingly involved. To ensure compatibility, please do not use any additional C99 features or GCC extensions. Especially watch out for:

mixing statements and declarations;
‘long long’ (use ‘int64_t’ instead);
‘__attribute__’ not protected by ‘#ifdef __GNUC__’ or similar;
GCC statement expressions (‘(x = ({ int y = 4; y; })’).

Indent size is 4. The presentation is one inspired by ’indent -i4 -kr -nut’. The TAB character is forbidden outside of Makefiles as is any form of trailing whitespace. Commits containing either will be rejected by the git repository.

The main priority in FFmpeg is simplicity and small code size in order to minimize the bug count.

Comments: Use the JavaDoc/Doxygen format (see examples below) so that code documentation can be generated automatically. All nontrivial functions should have a comment above them explaining what the function does, even if it is just one sentence. All structures and their member variables should be documented, too.

/**
 * @file mpeg.c
 * MPEG codec.
 * @author ...
 */

/**
 * Summary sentence.
 * more text ...
 * ...
 */
typedef struct Foobar{
    int var1; /**< var1 description */
    int var2; ///< var2 description
    /** var3 description */
    int var3;
} Foobar;

/**
 * Summary sentence.
 * more text ...
 * ...
 * @param my_parameter description of my_parameter
 * @return return value description
 */
int myfunc(int my_parameter)
...

fprintf and printf are forbidden in libavformat and libavcodec, please use av_log() instead.

Casts should be used only when necessary. Unneeded parentheses should also be avoided if they don’t make the code easier to understand.

1.4 Development Policy

Contributions should be licensed under the LGPL 2.1, including an "or any later version" clause, or the MIT license. GPL 2 including an "or any later version" clause is also acceptable, but LGPL is preferred.
You must not commit code which breaks FFmpeg! (Meaning unfinished but enabled code which breaks compilation or compiles but does not work or breaks the regression tests) You can commit unfinished stuff (for testing etc), but it must be disabled (#ifdef etc) by default so it does not interfere with other developers’ work.
You do not have to over-test things. If it works for you, and you think it should work for others, then commit. If your code has problems (portability, triggers compiler bugs, unusual environment etc) they will be reported and eventually fixed.
Do not commit unrelated changes together, split them into self-contained pieces. Also do not forget that if part B depends on part A, but A does not depend on B, then A can and should be committed first and separate from B. Keeping changes well split into self-contained parts makes reviewing and understanding them on the commit log mailing list easier. This also helps in case of debugging later on. Also if you have doubts about splitting or not splitting, do not hesitate to ask/discuss it on the developer mailing list.
Do not change behavior of the programs (renaming options etc) or public API or ABI without first discussing it on the ffmpeg-devel mailing list. Do not remove functionality from the code. Just improve!
Note: Redundant code can be removed.
Do not commit changes to the build system (Makefiles, configure script) which change behavior, defaults etc, without asking first. The same applies to compiler warning fixes, trivial looking fixes and to code maintained by other developers. We usually have a reason for doing things the way we do. Send your changes as patches to the ffmpeg-devel mailing list, and if the code maintainers say OK, you may commit. This does not apply to files you wrote and/or maintain.
We refuse source indentation and other cosmetic changes if they are mixed with functional changes, such commits will be rejected and removed. Every developer has his own indentation style, you should not change it. Of course if you (re)write something, you can use your own style, even though we would prefer if the indentation throughout FFmpeg was consistent (Many projects force a given indentation style - we do not.). If you really need to make indentation changes (try to avoid this), separate them strictly from real changes.
NOTE: If you had to put if(){ .. } over a large (> 5 lines) chunk of code, then either do NOT change the indentation of the inner part within (do not move it to the right)! or do so in a separate commit
Always fill out the commit log message. Describe in a few lines what you changed and why. You can refer to mailing list postings if you fix a particular bug. Comments such as "fixed!" or "Changed it." are unacceptable. Recommanded format: area changed: Short 1 line description
details describing what and why and giving references.
Make sure the author of the commit is set correctly. (see git commit –author) If you apply a patch, send an answer to ffmpeg-devel (or wherever you got the patch from) saying that you applied the patch.
When applying patches that have been discussed (at length) on the mailing list, reference the thread in the log message.
Do NOT commit to code actively maintained by others without permission. Send a patch to ffmpeg-devel instead. If no one answers within a reasonable timeframe (12h for build failures and security fixes, 3 days small changes, 1 week for big patches) then commit your patch if you think it is OK. Also note, the maintainer can simply ask for more time to review!
Subscribe to the ffmpeg-cvslog mailing list. The diffs of all commits are sent there and reviewed by all the other developers. Bugs and possible improvements or general questions regarding commits are discussed there. We expect you to react if problems with your code are uncovered.
Update the documentation if you change behavior or add features. If you are unsure how best to do this, send a patch to ffmpeg-devel, the documentation maintainer(s) will review and commit your stuff.
Try to keep important discussions and requests (also) on the public developer mailing list, so that all developers can benefit from them.
Never write to unallocated memory, never write over the end of arrays, always check values read from some untrusted source before using them as array index or other risky things.
Remember to check if you need to bump versions for the specific libav parts (libavutil, libavcodec, libavformat) you are changing. You need to change the version integer. Incrementing the first component means no backward compatibility to previous versions (e.g. removal of a function from the public API). Incrementing the second component means backward compatible change (e.g. addition of a function to the public API or extension of an existing data structure). Incrementing the third component means a noteworthy binary compatible change (e.g. encoder bug fix that matters for the decoder).
Compiler warnings indicate potential bugs or code with bad style. If a type of warning always points to correct and clean code, that warning should be disabled, not the code changed. Thus the remaining warnings can either be bugs or correct code. If it is a bug, the bug has to be fixed. If it is not, the code should be changed to not generate a warning unless that causes a slowdown or obfuscates the code.
If you add a new file, give it a proper license header. Do not copy and paste it from a random place, use an existing file as template.

We think our rules are not too hard. If you have comments, contact us.

Note, these rules are mostly borrowed from the MPlayer project.

1.5 Submitting patches

First, read the (see Coding Rules) above if you did not yet.

When you submit your patch, please use git format-patch or git send-email. We cannot read other diffs :-)

Also please do not submit a patch which contains several unrelated changes. Split it into separate, self-contained pieces. This does not mean splitting file by file. Instead, make the patch as small as possible while still keeping it as a logical unit that contains an individual change, even if it spans multiple files. This makes reviewing your patches much easier for us and greatly increases your chances of getting your patch applied.

Use the patcheck tool of FFmpeg to check your patch. The tool is located in the tools directory.

Run the regression tests before submitting a patch so that you can verify that there are no big problems.

Patches should be posted as base64 encoded attachments (or any other encoding which ensures that the patch will not be trashed during transmission) to the ffmpeg-devel mailing list, see http://lists.ffmpeg.org/mailman/listinfo/ffmpeg-devel

It also helps quite a bit if you tell us what the patch does (for example ’replaces lrint by lrintf’), and why (for example ’*BSD isn’t C99 compliant and has no lrint()’)

Also please if you send several patches, send each patch as a separate mail, do not attach several unrelated patches to the same mail.

Your patch will be reviewed on the mailing list. You will likely be asked to make some changes and are expected to send in an improved version that incorporates the requests from the review. This process may go through several iterations. Once your patch is deemed good enough, some developer will pick it up and commit it to the official FFmpeg tree.

Give us a few days to react. But if some time passes without reaction, send a reminder by email. Your patch should eventually be dealt with.

1.6 New codecs or formats checklist

Did you use av_cold for codec initialization and close functions?
Did you add a long_name under NULL_IF_CONFIG_SMALL to the AVCodec or AVInputFormat/AVOutputFormat struct?
Did you bump the minor version number (and reset the micro version number) in ‘avcodec.h’ or ‘avformat.h’?
Did you register it in ‘allcodecs.c’ or ‘allformats.c’?
Did you add the CodecID to ‘avcodec.h’?
If it has a fourcc, did you add it to ‘libavformat/riff.c’, even if it is only a decoder?
Did you add a rule to compile the appropriate files in the Makefile? Remember to do this even if you’re just adding a format to a file that is already being compiled by some other rule, like a raw demuxer.
Did you add an entry to the table of supported formats or codecs in ‘doc/general.texi’?
Did you add an entry in the Changelog?
If it depends on a parser or a library, did you add that dependency in configure?
Did you git add the appropriate files before committing?
Did you make sure it compiles standalone, i.e. with configure --disable-everything --enable-decoder=foo (or --enable-demuxer or whatever your component is)?

1.7 patch submission checklist

Does ’make fate’ pass with the patch applied?
Was the patch generated with git format-patch or send-email?
Did you sign off your patch? (git commit -s) See http://kerneltrap.org/files/Jeremy/DCO.txt for the meaning of sign off.
Did you provide a clear git commit log message?
Is the patch against latest FFmpeg git master branch?
Are you subscribed to ffmpeg-dev? (the list is subscribers only due to spam)
Have you checked that the changes are minimal, so that the same cannot be achieved with a smaller patch and/or simpler final code?
If the change is to speed critical code, did you benchmark it?
If you did any benchmarks, did you provide them in the mail?
Have you checked that the patch does not introduce buffer overflows or other security issues?
Did you test your decoder or demuxer against damaged data? If no, see tools/trasher and the noise bitstream filter. Your decoder or demuxer should not crash or end in a (near) infinite loop when fed damaged data.
Does the patch not mix functional and cosmetic changes?
Did you add tabs or trailing whitespace to the code? Both are forbidden.
Is the patch attached to the email you send?
Is the mime type of the patch correct? It should be text/x-diff or text/x-patch or at least text/plain and not application/octet-stream.
If the patch fixes a bug, did you provide a verbose analysis of the bug?
If the patch fixes a bug, did you provide enough information, including a sample, so the bug can be reproduced and the fix can be verified? Note please do not attach samples >100k to mails but rather provide a URL, you can upload to ftp://upload.ffmpeg.org
Did you provide a verbose summary about what the patch does change?
Did you provide a verbose explanation why it changes things like it does?
Did you provide a verbose summary of the user visible advantages and disadvantages if the patch is applied?
Did you provide an example so we can verify the new feature added by the patch easily?
If you added a new file, did you insert a license header? It should be taken from FFmpeg, not randomly copied and pasted from somewhere else.
You should maintain alphabetical order in alphabetically ordered lists as long as doing so does not break API/ABI compatibility.
Lines with similar content should be aligned vertically when doing so improves readability.
Consider to add a regression test for your code.
If you added YASM code please check that things still work with –disable-yasm

1.8 Patch review process

All patches posted to ffmpeg-devel will be reviewed, unless they contain a clear note that the patch is not for the git master branch. Reviews and comments will be posted as replies to the patch on the mailing list. The patch submitter then has to take care of every comment, that can be by resubmitting a changed patch or by discussion. Resubmitted patches will themselves be reviewed like any other patch. If at some point a patch passes review with no comments then it is approved, that can for simple and small patches happen immediately while large patches will generally have to be changed and reviewed many times before they are approved. After a patch is approved it will be committed to the repository.

We will review all submitted patches, but sometimes we are quite busy so especially for large patches this can take several weeks.

When resubmitting patches, please do not make any significant changes not related to the comments received during review. Such patches will be rejected. Instead, submit significant changes or new features as separate patches.

1.9 Regression tests

Before submitting a patch (or committing to the repository), you should at least test that you did not break anything.

The regression tests build a synthetic video stream and a synthetic audio stream. These are then encoded and decoded with all codecs or formats. The CRC (or MD5) of each generated file is recorded in a result file. A ’diff’ is launched to compare the reference results and the result file. The output is checked immediately after each test has run.

The regression tests then go on to test the FFserver code with a limited set of streams. It is important that this step runs correctly as well.

Run ’make test’ to test all the codecs and formats. Commands like ’make regtest-mpeg2’ can be used to run a single test. By default, make will abort if any test fails. To run all tests regardless, use make -k. To get a more verbose output, use ’make V=1 test’ or ’make V=2 test’.

Run ’make fulltest’ to test all the codecs, formats and FFserver.

[Of course, some patches may change the results of the regression tests. In this case, the reference results of the regression tests shall be modified accordingly].

This document was generated by Kyle Schwarz on June 10, 2011 using texi2html 1.82.