GitLab

Controls can be processed if the playlist is empty. They just do
nothing. Conversely, stop need not be processed if the playlist was
already killed.

Those two functions have no effects if the playlist is stopped.
Otherwise they force the playlist to playing ("running") or paused
state respectively.

As a reminder, the existing playlist_Play() forces the playlist into
running state (unless it is empty), and playlist_Stop() forces the
playlist into stopped state (unless the input thread refuses to die).
There are no functions to force the playlist to paused state.

The behaviour is unchanged:
 - if stopped: starts playing,
 - if playing: pauses playing,
 - if paused: resumes playing.

Fixes #8082

Re-license almost all of libVLC and libVLCcore to LGPLv2.1+

This move was authorized by the developers, either:
 - by e-mail,
 - by vote at the VideoLAN Dev Days 2011,
 - on the license website,
 - in a contract, oral or written.
No objection was raised, so far.

The developers agreeing are:
  Justus Piater
  Alexis Ballier
  Alexander Bethke
  Mohammed Adnène Trojette
  Alex Converse
  Alexey Sokolov
  Alexis de Lattre
  Andre Pang
  Anthony Loiseau
  Cyril Deguet
  André Weber
  Boris Dorès
  Brieuc Jeunhomme
  Benjamin Drung
  Hugo Beauzée-Luyssen
  Benoit Steiner
  Benjamin Pracht
  Bernie Purcell
  Przemyslaw Fiala
  Arnaud de Bossoreille de Ribou
  Brad Smith
  Nick Briggs
  Christopher Rath
  Christophe Courtaut
  Christopher Mueller
  Clement Chesnin
  Andres Krapf
  Damien Fouilleul
  David Flynn
  Sebastien Zwickert
  Antoine Cellerier
  Jérôme Decoodt
  Jérome Decoodt
  Dylan Yudaken
  Eduard Babayan
  Eugenio Jarosiewicz
  Elliot Murphy
  Eric Petit
  Erwan Tulou
  Etienne Membrives
  Ludovic Fauvet
  Fabio Ritrovato
  Tobias Güntner
  Jakub Wieczorek
  Frédéric Crozat
  Francois Cartegnie
  Laurent Aimar
  Florian G. Pflug
  Felix Paul Kühne
  Frank Enderle
  Rafaël Carré
  Simon Latapie
  Gildas Bazin
  Geoffroy Couprie
  Julien / Gellule
  Gildas Bazin
  Arnaud Schauly
  Toralf Niebuhr
  Vicente Jimenez Aguilar
  Derk-Jan Hartman
  Henri Fallon
  Ilkka Ollakka
  Olivier Teulière
  Rémi Duraffort
  Jakob Leben
  Jean-Baptiste Kempf
  Jean-Paul Saman
  Jean-Philippe Grimaldi
  Jean-François Massol
  Gaël Hendryckx
  Jakob Leben
  Jean-Marc Dressler
  Jai Menon
  Johan Bilien
  Johann Ransay
  Joris van Rooij
  JP Dinger
  Jean-Philippe André
  Adrien Grand
  Juha Jeronen
  Juho Vähä-Herttua
  Kaarlo Raiha
  Kaarlo Raiha
  Kamil Baldyga
  Keary Griffin
  Ken Self
  KO Myung-Hun
  Pierre Ynard
  Filippo Carone
  Loïc Minier
  Luca Barbato
  Lucas C. Villa Real
  Lukas Durfina
  Adrien Maglo
  Marc Ariberti
  Mark Lee
  Mark Moriarty
  Martin Storsjö
  Christophe Massiot
  Michel Kaempf
  Marian Ďurkovič
  Mirsal Ennaime
  Carlo Calabrò
  Damien Lucas
  Naohiro Koriyama
  Basos G
  Pierre Baillet
  Vincent Penquerc'h
  Olivier Aubert
  Pankaj Yadav
  Paul Corke
  Pierre d'Herbemont
  Philippe Morin
  Antoine Lejeune
  Michael Ploujnikov
  Jean-Marc Dressler
  Michael Hanselmann
  Rafaël Carré
  Ramiro Polla
  Rémi Denis-Courmont
  Renaud Dartus
  Richard Shepherd
  Faustino Osuna
  Arnaud Vallat
  Rob Jonson
  Robert Jedrzejczyk
  Steve Lhomme
  Rocky Bernstein
  Romain Goyet
  Rov Juvano
  Sam Hocevar
  Martin T. H. Sandsmark
  Sebastian Birk
  Sébastien Escudier
  Vincent Seguin
  Fabio Ritrovato
  Sigmund Augdal Helberg
  Casian Andrei
  Srikanth Raju
  Hannes Domani
  Stéphane Borel
  Stephan Krempel
  Stephan Assmus
  Tony Castley
  Pavlov Konstantin
  Eric Petit
  Tanguy Krotoff
  Dennis van Amerongen
  Michel Lespinasse
  Can Wu
  Xavier Marchesini
  Sébastien Toque
  Christophe Mutricy
  Yoann Peronneau
  Yohann Martineau
  Yuval Tze
  Scott Caudle
  Clément Stenac

It is possible, that some minor piece of code was badly tracked, for
some reasons (SVN, mainly) or that some small developers did not answer.
However, as an "œuvre collective", defined as in "CPI 113-2 alinéa 3",
and seeing "Cour. Cass. 17 Mai 1978", and seeing that the editor and
the very vast majority of developers have agreed (> 99.99% of the code,
> 99% of developers), we are fine here.

This fix a strange behavior if playlist_Pause is called without input.

pl_Hold() assumed the playlist was already created.
pl_Release() does not solve any problem. Playlist synchronization is
done with the playlist lock, while playlist deletion must anyway be done
after all interfaces, services and other threads are joined.

The previous fix made the fetcher and preparser persistent throughout
the lifetime of the playlist. So we don't need to protect the pointers
anymore. Also, they are internally thread-safe (with their own lock).

(This does NOT fix the RC pause deadlock)

stop the playlist just after removing the playing item (so that's ok to stop
the PL even when it's empty).

(cherry picked from commit be1f4305f1203ccf5acc402fd2cc80dde3f5b1cd)
Signed-off-by: Rémi Duraffort <ivoire@videolan.org>

No useless wake up and simplify.
Fix a few race conditions with deactivation.

(Moved around playlist code)

No functionnality changes.

Some interfaces may assert because of missing playlist lock.