This is easier and more efficient to loop on, because only two
variables get modified (src and dest).

GLib's macros are implemented with inline assembly, and probably a lot
faster than our naive bit shifting implementation.

The other audio_format attributes are not used.

This function is too complex for being inlined everywhere.

Send another "magic" MIME type when the byte order must be reversed.
This also fixes byte order issues when big-endian was involved (input
data or host byte order) - that was completely broken.

This function is not critical for performance, and the inline
expansion looks too expensive.

For better optimization.

Compile the plugins only once with the same C flags.

In a C file, that is too late, because _mingw.h defaults to 0x0502,
and ours would be illegal redefintion.

Use local variable "is_float".

Cast to enum sample_format.  Without the cast, it's just a plain
integer, and gcc cannot know that a "case" statement is missing.

The protocol mandates that playlist URIs are absolute (i.e. full URI
relative to the music directory, not relative to the parent
directory).  This adds the parameter "directory" to the "playlist"
visitor method.

.. instead of failing playback completely.

Allow gcc to warn when a new format isn't supported.

Let the libsamplerate code initialize itself.

Load the samplerate_converter on MPD startup.  Fail if the converter
name is invalid.

Defaults to "no", which fixes the noise problems.

When we have an absolute path that's not inside the music directory,
allow loading it anyway, if we're in "secure" mode (i.e. the client is
connected via UNIX socket).

Right now, a playlist with absolute pathnames can only add songs that
are in the same the directory of the playlist or under it.

If uri is an absolute pathname and base_uri is set,
playlist_check_translate_song() will check that base_uri is a prefix
of uri, excluding every other song in the music directory outside
base_uri.

I think in this case base_uri should be completely ignored (and made
NULL) and uri should just be checked against music root directory.

Previously, the condition "defined(play_audio_format)" was used to see
if an output device has been opened, but if the device had failed on
startup, an assertion failure could occur.  This patch adds a separate
flag.

Common function that manages "player" attributes after
audio_output_all_open() has returned.

Now that the player thread can handle SEEK commands while not (yet)
playing, we can remove the "pc.state" check from pc_seek().

This checks both conditions: pending and running.  Fixes yet another
assertion failure!

.. and add a few code comments.

Don't read song.start_ms and song.end_ms, let the player thread manage
this logic instead.

Fixes one more assertion failure.