Difference between revisions of "Build the Viewer on Linux"

From Second Life Wiki
Jump to navigation Jump to search
(→‎Compiling with CMake: Added notes to use distcc.)
m (Note out of date third party libraries)
 
(218 intermediate revisions by 19 users not shown)
Line 1: Line 1:
{{KBcaution|custom=Note|
Linden Lab no longer actively maintains Linux viewer third party dependencies. You may run into issues building the viewer on modern distros.}}
{{multi-lang}}
{{multi-lang}}
{{CompileNav}}
{{CompileNav}}


The following are instructions for building the Second Life viewer on linux. This process has been used on [http://www.debian.org/ debian] and debian based systems like [http://www.ubuntu.com/ ubuntu], and also on [http://www.fedoraproject.org/ Fedora]. For other platforms, see [[Get source and compile]].
The following are instructions for building the Second Life viewer on linux. This process has been used on [http://www.debian.org/ debian] and debian based systems like [http://www.ubuntu.com/ ubuntu], and also on [http://www.fedoraproject.org/ Fedora]. For other platforms, see [[Get source and compile]].
 
<onlyinclude>
== Required tools ==
== Required tools ==
There are a number of tools that need to be installed first.
There are a number of tools that need to be installed first.


* <b>python 2.4</b> [package: python]
* '''hg''' [package: mercurial]
** python 2.3 or before doesn't work in SL development.
* '''cmake''' [package: cmake]
** python 2.5 or later works, but it is not recommended, since it generates a lot of warning messages when running develop.py.
** Currently (snowglobe 1.3.2 and higher) cmake 2.6.2 is the minimum required version.
* <b>GCC 4.1</b> or later C++ compiler.
* <b>yacc</b> or compatible tool [suggested package: bison]
** In a past, <b>GCC 3.4</b> C/C++ compiler [debian/ubuntu: g++-3.4] [fedora: compat-gcc-34-c++] was recommended for SL development, but recent SL viewer requires gcc 4 series.  
* <b>lex</b> or compatible tool [suggested package: flex]
*** '''libboost-program-options-dev''' is needed on Ubuntu 8.04 to use cmake
* <b>python</b> [package: python]
* <b>yacc</b> or compatible tool [suggest packages: bison]
** python 2.4.3 is the minimum required version.
* <b>lex</b> or compatible tool [suggest packages: flex]
** All more recent 2.x versions should work, too. If you encounter problems or deprecation warnings, please [[issue tracker|report]] them.
** python 3.x hasn't been tested yet.
* <b>g++</b> [package: g++]
** Note (not relevant for standalone): gcc 4.4 (which is in recent Ubuntu and debian) won't work in versions prior to 2.0 (or Snowglobe 1.3) if you are not building ''standalone'' (see below), because it chokes on some parts of boost prior to 1.37 (http://svn.boost.org/trac/boost/ticket/2069). The solution is to install <b>GCC 4.3</b> and to run 'export CXX=/usr/bin/g++-4.3' or whatever your binary is before trying to compile.
:Another workaround is to use update-alternatives; for example, if you already have 4.4 installed, try this:
sudo apt-get install g++-4.3
sudo update-alternatives --remove-all gcc
sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-4.3 43  --slave /usr/bin/g++ g++ /usr/bin/g++-4.3  --slave /usr/bin/gcov gcov /usr/bin/gcov-4.3
sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-4.4 44  --slave /usr/bin/g++ g++ /usr/bin/g++-4.4  --slave /usr/bin/gcov gcov /usr/bin/gcov-4.4
:then choose 4.3:
sudo update-alternatives --config gcc
* Linden Lab has backported gcc-4.6.3 to Debian Squeeze; this port and its associated libraries are used when building the official Linux viewer.
:The sources and Debian packages for this backport are available from [[Debian GCC Backport]]
* '''libboost-program-options-dev''' was needed on Ubuntu 8.04 to use cmake, but it's '''not needed anymore''' to build a 1.23-render-pipeline on Ubuntu 9.04-beta (nor Snowglobe 2 on Ubuntu 9.10)
* '''make''' (package: make)
* '''bzip2''' (package: bzip2)


The build process may use the following optional tool:
The build process may use the following optional tools:


* <b>distcc</b> distributed compiler (useful if you have multiple PCs.)
* <b>distcc</b> distributed compiler (useful if you have multiple PCs.)
* <b>ccache</b> a fast compiler cache (speed up recompilation, a must if you intend to do development.)
{{KBhint|To boost your productivity as a developer, try the suggestions at [[Development Environment for Multiple Viewers]] about directory structure, multiple viewers and automation of repetitive tasks.}}


== Unpack the source tree ==
</onlyinclude>


You can get the source either from the source archives available at [[Source downloads]], or you can get it from [[Subversion]].
== What does 'Standalone' mean? ==


Using a source archive is the easier option for getting started, and is strongly recommended especially for the first-time builders.  However, some says that if you plan to do active development and want to make your own modifications, while merging Linden Lab's changes, Subversion will probably be much more convenient on the long term.
''The current sources no longer use the term "standalone"; it has been replaced by "usesystemlibs".''


=== Getting the source from the source archives ===
A ''standalone'' build of the viewer refers to building a viewer against the shared libraries that are installed on your system instead of using precompiled libraries provided by Linden Lab.
Get the source from [[Source downloads]]. You will need all 3 archives: source, artwork and libraries.


Choose a location and unpack the source tree and the art work.  They will be extracted into a
In order to build standalone, you will have to configure using <code>--standalone</code>, see [[#Configuring_your_tree|configuration]] below.
directory called <b>linden</b>.


<code>
The advantage of building standalone is that you might use less RAM: you'll be using the same shared libraries
% tar -xzf slviewer-src-<version>.tar.gz
that other applications use. However this is a rather minor advantage. The disadvantages of building standalone is a long list of potential problems: you will have to manually install all the development packages of many many libraries. In some cases there isn't a package for it from your distribution, so you have to especially get those sources separately and compile and install them in a way that the viewer can find them. Building standalone is not supported officially by Linden Lab and therefore not tested. Hence, it might not work at all (although the open source snowglobe developers usually make sure that it works for snowglobe). Nevertheless, every now and then it happens that the viewer won't build with some new released library (ie, boost), in which case you most likely will have to use the latest source code retrieved with subversion.
% tar -xzf slviewer-linux-libs-<version>.tar.gz
% unzip slviewer-artwork-<version>.zip
</code>


=== Getting the source from Subversion ===
Nevertheless, there might be reasons that you want to do a standalone build anyway. The most important one being that Linden Lab also doesn't support 64-bit: they do not provided shared libraries for 64-bit; so, if you want to build native 64 bit you must use standalone. Another reason might be that Linden Lab is using a few rather old library versions that are incompatible with newer versions. If you have any reason to use a newer version for some library, then chances are that things will break unless you build standalone. One reason would be if you want to debug a library and fix it because it is buggy.
Select a source [[Source branches|branch]] to check out.


Check it out with:
Conclusion: if you are building the viewer for the first time, and you are building for 32-bit, then you should ''not'' use standalone.
<code>
% svn co http://svn.secondlife.com/svn/linden/<branch> [directory]
</code>
If you don't specify the directory, it'll use the name of the branch as the default.


Now you will have a directory with the source code. This only includes the source code itself, without the libraries or artwork. You will have to get them from the [[Source downloads]] and unpack them into the tree.
== Getting the source ==


If the directory SVN created is called 'linden' then you can directly unzip and untar the archives on top of it. If it's called something else, create a symlink:
The source for the official Viewer's 2.x codebase resides in mercurial repositories hosted on http://bitbucket.org/lindenlab (a.k.a. http://bitbucket.org/lindenlab).
<code>
% ln -sf <branch> linden
</code>


Now unpack the libraries and artwork:
{{KBnote|custom=Old Versions|Find instructions on how to obtain the sources for official 1.x codebase as well as [[Snowglobe]] 1.x and 2.x on the [https://wiki.secondlife.com/w/index.php?title{{=}}Compiling_the_viewer_%28Linux%29&oldid{{=}}1126012#Getting_the_source archived version of this section].}}
<code>
% tar -xzf slviewer-<os>-libs-<version>.tar.gz
% unzip slviewer-artwork-<version>.zip
</code>


<br />
To check out the '''[[viewer-release (source branch)|lindenlab/viewer-release]]''' branch, do
hg clone http://bitbucket.org/lindenlab/viewer-release
This will create a new folder called <code>viewer-release</code> inside your current working directory, containing the source tree. (You can optionally specify a different target directory. See <code>hg help clone</code>.) Other long-lived branches of interest are '''[[viewer-beta (source branch)|lindenlab/viewer-beta]]''' and '''[[viewer-development (source branch)|lindenlab/viewer-development]]'''. Contributions should usually be based on '''viewer-development'''.


== Installing the required libraries (that Linden Lab can not or does not provide)==
== Installing the required libraries (that Linden Lab can not or does not provide)==
Even non-standalone still uses a few shared libraries from your system.
This paragraph deals with those libraries. For ''standalone'' you will
need a lot more (see later on).


=== Libraries and header files that usually come with a Linux distribution ===
=== Libraries and header files that usually come with a Linux distribution ===
Make sure the libraries and header files for the following packages are installed on your system:
Make sure the libraries and header files for the following packages are installed on your system:
; GL : [ubuntu: mesa-common-dev, fedora: mesa-libGL-devel]
{|border="1" cellpadding="5" cellspacing="0"
; GLU : [ubuntu: libglu1-mesa-dev, fedora: mesa-libGLU-devel]
|-
; glibc : [ubuntu: libc6-dev, fedora: glibc-devel]
!Library
; stdc++ library : [ubuntu: libstdc++6, fedora: libstdc++-devel]
!Debian/Ubuntu
; X11 : [ubuntu: libx11-dev, fedora: libX11-devel]
!Fedora/Red Hat
; zlib : [ubuntu: zlib1g-dev, fedora: zlib-devel]
|-
; openssl : [ubuntu: libssl-dev, redhat: openssl-devel]
!libc.so
|libc6-dev
|glibc-devel
|-
!libstdc++.so.6
|libstdc++6
|libstdc++-devel
|-
!libX11.so
|libx11-dev
|libX11-devel
|-
!libGL.so
|nvidia-glx-dev '''or''' libgl1-mesa-dev
|mesa-libGL-devel
|-
!libXrender.so
|libxrender-dev
|libXrender-devel
|-
|libidn.so
|libidn11-dev
|libidn11-devel
|}
 
==== shortcut commands for the above ====
 
Debian/Ubuntu:
 
sudo apt-get install libc6-dev libstdc++6 libx11-dev libxrender-dev libidn11-dev [libgl1-mesa-dev|nvidia-current-dev|...]
 
Fedora/Red Hat:
 
sudo yum install glibc-devel libstdc++-devel libX11-devel mesa-libGL-devel libXrender-devel libidn11-devel
 
=== Recommended libraries and headers ===
 
The following packages are required when building standalone, but even if you are building non-standalone you should probably install them.
 
{|border="1" cellpadding="5" cellspacing="0"
|-
!Library
!Debian/Ubuntu
!Fedora/Red Hat
|-
!libGLU.so
|libglu1-mesa-dev
|mesa-libGLU-devel
|-
!libz.so
|zlib1g-dev
|zlib-devel
|-
!libssl.so
|libssl-dev
|openssl-devel
|-
!libogg.so
|libogg-dev
|libogg-devel
|-
!libpng12.so
|libpng12-dev
|libpng-devel
|-
!libdbus-glib-1.so
|libdbus-glib-1-dev
|dbus-glib-devel
|-
!libatk-1.0.so<br/>libcairo.so<br/>libgdk-x11-2.0.so<br/>libgdk_pixbuf-2.0.so<br/>libgmodule-2.0.so<br/>libglib-2.0.so<br/>libgtk-x11-2.0.so<br/>libgthread-2.0.so<br/>libpango-1.0.so<br/>and many more
|libgtk2.0-dev
|atk-devel cairo-devel gtk2-devel glib2-devel pango-devel
|}
 
==== shortcut commands for the above (2) ====
 
Debian/Ubuntu:
 
<pre><nowiki>
sudo apt-get install libglu1-mesa-dev zlib1g-dev libssl-dev libogg-dev libpng12-dev libdbus-glib-1-dev libgtk2.0-dev
</nowiki></pre>
 
Fedora/Red Hat:
 
<pre><nowiki>
sudo yum install mesa-libGLU-devel zlib-devel openssl-devel libogg-devel libpng-devel dbus-glib-devel \
  atk-devel cairo-devel gtk2-devel glib2-devel pango-devel
</nowiki></pre>
 
== Necessary libraries when building standalone ==
 
=== Existing package names ===
 
If you want to build 'standalone', then the following packages are required in addition:
 
{|border="1" cellpadding="5" cellspacing="0"
|-
!Library
!Debian/Ubuntu
!Fedora/Red Hat
!openSUSE
|-
!libopenal.so
|libopenal-dev
|openal-soft-devel
|openal-soft-devel
|-
!libvorbis.so<br/>libvorbisenc.so<br/>libvorbisfile.so
|libvorbis-dev
|libvorbis-devel
|libvorbis-devel
|-
!libalut.so
|libalut-dev
|freealut-devel
|freealut-devel
|-
!libapr-1.so
|libapr1-dev
|apr-devel
|libapr1-devel
|-
!libaprutil-1.so
|libaprutil1-dev
|apr-util-devel
|libapr-util1-devel
|-
!libboost_program_options.so<br/>libboost_regex.so<br/>libboost_signals.so
|libboost-dev
|boost-devel
|boost-devel
|-
!libcares.so
|libc-ares-dev
|c-ares-devel
|libcares-devel
|-
!libxmlrpc-epi.so
|libxmlrpc-epi-dev
|?
|libxmlrpc-epi0-devel
|-
!libopenjpeg.so
|libopenjpeg-dev
|openjpeg-devel
|openjpeg-devel
|-
!libjpeg.so
|libjpeg62-dev (libjpeg-dev)
|libjpeg-devel
|libjpeg62-devel
|-
!libSDL.so
|libsdl1.2-dev
|SDL-devel
|libSDL-devel
|-
!libgstreamer-0.10.so
|libgstreamer0.10-dev
|gstreamer-devel
|gstreamer-0_10-devel
|-
!gstreamer-0.10 plugins
|libgstreamer-plugins-base0.10-dev
|gstreamer-plugins-base-devel
|gstreamer-0_10-plugins-base-devel
|-
!libgmock.so
|google-mock (Snowglobe 2 only)
|?
|?
|}
 
==== shortcut commands for the above (3) ====
 
Debian/Ubuntu:
 
<pre><nowiki>
sudo apt-get install libopenal-dev libvorbis-dev libalut-dev libapr1-dev libaprutil1-dev libboost-dev \
  libc-ares-dev libxmlrpc-epi-dev libopenjpeg-dev libjpeg62-dev libgtk2.0-dev libsdl1.2-dev \
  libgstreamer0.10-dev libgstreamer-plugins-base0.10-dev google-mock
</nowiki></pre>
 
Fedora/Red Hat:
 
<pre><nowiki>
sudo yum install openal-soft-devel libvorbis-devel freealut-devel apr-devel apr-util-devel boost-devel \
  c-ares-devel openjpeg-devel libjpeg-devel SDL-devel gstreamer-devel gstreamer-plugins-base-devel
</nowiki></pre>
 
openSUSE:
 
You will have to add the buildservice repository home:lemmy04:snowglobe (http://download.opensuse.org/repositories/home:/lemmy04:/snowglobe/) to your package sources.


=== FMOD (audio) ===
<pre><nowiki>
FMOD is an audio output library.  FMOD is required to make SL viewer to generates sounds.
sudo zypper install openal-soft-devel libvorbis-devel freealut-devel libapr1-devel libapr-util1-devel \
  boost-devel libcares-devel libxmlrpc-epi0-devel openjpeg-devel libjpeg62-devel libSDL-devel gstreamer-0_10-devel gstreamer-0_10-plugins-base-devel
</nowiki></pre>


* Fetch and unpack <b>FMOD 3.75</b> <http://www.fmod.org/>
=== More problematic libraries (standalone) ===
* Note: in trunk code (march 2009) there is no need to manually create the three linden/libraries/i686-linux/... directories


wget http://www.fmod.org/index.php/release/version/fmodapi375linux.tar.gz
Easy so far, now here's where the trouble starts.
tar -xzvf fmodapi375linux.tar.gz
Not all required packages are in debian, not to mention you might not even be ''using'' debian
mkdir -p linden/libraries/i686-linux/lib_release_client/
of course; but we (standalone builders) rely for the most part on the tremendous work
mkdir -p linden/libraries/i686-linux/lib_debug/
done by 64-bit pioneer [[User:Robin_Cornelius]] who happens to use debian. Although many
mkdir -p linden/libraries/i686-linux/lib_release/
required packages are now in debian, some still need to be downloaded from Robin's
cd fmodapi375linux/
private repository (assuming you're on debian). TODO: Add instructions for those not
cp api/inc/* ../linden/libraries/i686-linux/include/
using ubuntu/debian.
cp api/libfmod-3.75.so ../linden/libraries/i686-linux/lib_release_client/
<br/>
cp api/libfmod-3.75.so ../linden/libraries/i686-linux/lib_release/
cp api/libfmod-3.75.so ../linden/libraries/i686-linux/lib_debug/


* FMOD is available free of charge, but it is not itself an open-source.  We understand that some Linux users dislike to use software whose source is not open, and FMOD falls into this category.  <b>If</b> you wish to avoid FMOD, thus disabling audio, you may make these changes:
Edit your <code>/etc/apt/sources.list</code> and add the following lines:
** Comment-out the libfmod line in indra/newview/viewer_manifest.py


== Installing the required dependencies (prepackaged by Linden Lab)==
# Repository for SecondLife.
deb <nowiki>http://apt.byteme.org.uk</nowiki> squeeze main
deb-src <nowiki>http://apt.byteme.org.uk</nowiki> squeeze main


The Second Life Viewer has a number of compile/link dependencies on external libraries which
or
are needed - to help you, the source download page contains a link to a <i>slviewer-linux-libs</i> package which you unpack over the source tree to fill most of the dependencies (and thus avoid most of the fiddly work described on this page).
 
# Repository for SecondLife.
deb <nowiki>http://apt.byteme.org.uk</nowiki> lenny main
deb-src <nowiki>http://apt.byteme.org.uk</nowiki> lenny main
 
depending on whether you are using testing (squeeze) or stable (lenny).
Note that you won't be able to compile Snowglobe 2.x on lenny because the Qt library is too old.
 
Run <code>apt-get update</code> to get the Package files from byteme as usual.
You can ignore the warning about the missing public key, or do what is
described at the bottom of [http://omvviewer.byteme.org.uk/index.php/downloads/debian/ this page] under 'GPG Keys and Signed repository' (note the remark about not using root), thus:
 
% gpg --keyserver hkp://keyserver.ubuntu.com --recv-key 0x729A79A23B7EE764
% gpg --export 0x729A79A23B7EE764 | sudo apt-key add -
 
or for ubuntu users we use the openmetaverse repo, run this command to add the repo:
 
# Openmetaverse repo
sudo add-apt-repository ppa:openmetaverse/ppa
 
'''NOTE:''' If you are on Ubuntu 12.04 ("Precise") you will also need to edit the file /etc/apt/sources.list.d/openmetaverse-ppa-precise.list so that it refers to the "lucid" distro instead of "precise" because the maintainers haven't made packages for "precise" yet.
 
Change /etc/apt/source.list.d/openmetaverse-ppa-precise.list from:
 
deb http://ppa.launchpad.net/openmetaverse/ppa/ubuntu precise main
deb-src http://ppa.launchpad.net/openmetaverse/ppa/ubuntu precise main
 
To be:
 
deb http://ppa.launchpad.net/openmetaverse/ppa/ubuntu lucid main
deb-src http://ppa.launchpad.net/openmetaverse/ppa/ubuntu lucid main
 
and of course to add the keys ('''only on Debian''')
 
gpg --keyserver hkp://keyserver.ubuntu.com --recv-key 0x50A40F50
gpg --export 0x50A40F50 | sudo apt-key add -
 
and always remember to update after adding repos
 
apt-get update
 
Finally install these packages:
 
{|border="1" cellpadding="5" cellspacing="0"
|-
!Library
!Debian/Ubuntu
|-
!libcurl-cares.so
|libcurl4-cares-dev
|-
!llqtwebkit
|libllqtwebkit-dev (Snowglobe 1) or<br/>libllqtwebkit2-dev (Snowglobe 2)
|-
!JsonCpp
|libjsoncpp-dev
|-
|libndofdev
|libndofdev-dev (optional, needed if you have a Space Navigator joystick)
|-
!glh-linear
|glh-linear (Snowglobe 2 only)
|-
!boost-coroutine
|boost-coroutine (Snowglobe 2 only)
|}
<br/>
 
Shortcut commands for the above
 
sudo apt-get install libcurl4-cares-dev libllqtwebkit2-dev glh-linear boost-coroutine libndofdev-dev
 
If you are on lenny you will also (automatically) install the packages '''cmake''', '''libopenjpeg-dev''' and '''libopenjpeg2''' from byteme. This is because the packages in lenny contain bugs.
 
The libjsoncpp-dev installs a library called <code>/usr/lib/libjson.so</code>, but the viewer is looking for one called <code>libjson_linux-gcc-${_gcc_COMPILER_VERSION}_libmt.so</code> in '/usr/lib' and '/usr/local/lib'. Therefore execute the following command as root:
 
# This is not needed on Snowglobe 1.4.1 and higher.
% ln -s /usr/lib/libjson.so /usr/local/lib/libjson_linux-gcc-$(g++ -dumpversion)_libmt.so
 
In order to run the tests, you need a template library called 'tut'. Because this is a template library, we can just use the library archive from Linden Lab for any architecture. Therefore, to install tut, run:
 
% scripts/install.py tut
 
Until recently for some completely mysterious reason, the source code includes <code><tut/tut.hpp></code> while
configure checks for <code>tut.h</code> in <code>/usr/lib</code> and <code>/usr/local/lib</code> only (CXXFLAGS?
Never heard of that).
In order to fix this brokenness for non-Snowglobe viewers or based on older snowglobe versions (prior to 1.4, or 2.1),
execute the following (2.x viewers older than 2.1 and 1.x viewers older than 1.4 only):
 
# This is not needed on Snowglobe 1.4.1 and higher.
% sudo touch /usr/local/include/tut.h
 
which satisfies the configure check. The installed tut.hpp is still used anyway since <code>libraries/include</code> is in the include search path by default. Alternatively, edit <code>indra/cmake/Tut.cmake</code> and comment out these two lines:


If you download the libs to the top folder, where the <i>linden</i> folder is after getting and extracting the viewer source code tarball, the following command should unpack everything to the right spot.
<code>
<code>
tar xvfz slviewer-linux-libs-<version>.tar.gz
  #include(FindTut)
  #include_directories(${TUT_INCLUDE_DIR})
</code>
</code>


'''Note''': as of 1.21, very little of what once was included in the libraries tarball are still there, so this is a very small tarball.  The other libraries will be downloaded as part of the build process.
Oh WAIT! On 2.0 standalone <code>libraries/include</code> is NOT included anymore, and it seems impossible
to fix the cmake spaghetti so here's what you have to do on top of the above:


If you don't wish to use the precompiled libraries, see [[Building the viewer libraries (Linux)]].
sudo cp -r libraries/include/tut /usr/local/include


== Compiling with CMake ==
Sorry but there is no other way.
 
However, if you are using snowglobe 1.4 or a later 1.x, or snowglobe 2.1 or later, or a viewer based on those,
then the viewer actually looks for tut/tut.hpp. So, you can still install it in <code>/usr/local/include/tut/</code>,
but you can also install it in -say- /somewhere/include/tut/ and set the environment variable CMAKE_INCLUDE_PATH
(a colon separated list of paths) to include <code>/somewhere/include</code> before configuration. Note that due to
a cmake problem/bug it will find /somewhere/include but ''still'' use the include path of any other library,
including /usr/include or /usr/local/include, so make sure you remove any other tut installation from common
paths if you use the latter method.
 
=== Get voice chat working (standalone) ===
 
The voice binaries downloaded for non-standalone builds can be used in for standalone builds, too. Because they'll run as a separate process, this even works with 64-bit builds of the viewer (assuming your system is set up for also running 32-bit binaries). To download and unpack the binaries (which doesn't happen automatically when building standalone) run:
${SOURCE_DIR}/scripts/install.py slvoice
 
{{KBnote|In earlier revisions, the package was named <code>vivox</code> rather than <code>slvoice</code>.}}
 
After [[#Compiling|building]], you'll have to copy the binaries into the right place
(-- I don't think this is correct as this will also link the viewer against those libraries.
Instead you have to use a script for SLVoice that sets the right LD_LIBRARY_PATH and
then runs the real SLVoice --[[User:Aleric Inglewood|Aleric Inglewood]] 12:03, 13 July 2010 (UTC)):
cp ${SOURCE_DIR}/indra/newview/vivox-runtime/i686-linux/SLVoice ${BUILD_DIR}/newview/packaged/bin/
cp ${SOURCE_DIR}/indra/newview/vivox-runtime/i686-linux/*.so* ${BUILD_DIR}/newview/packaged/lib/
 
If you want, you can then remove the voice binaries again from the source tree with
${SOURCE_DIR}/scripts/install.py --uninstall vivox
 
=== Remaining issues (standalone) ===
 
In order to run the viewer you need to create a 'package' (a directory really, to which everything is copied that is needed to run it).
Unfortunately, if you are building standalone on a 32-bit machine, it ''still'' tries to happily copy the Linden Lab provided shared libraries over, which we never even downloaded mind you, causing the packaging to fail. Therefore, edit <code>indra/newview/viewer_manifest.py</code> and comment out the following lines in <code>class Linux_i686Manifest(LinuxManifest)</code>:
 
<pre><nowiki>
            #self.path("libapr-1.so.0")
            #self.path("libaprutil-1.so.0")
            #self.path("libdb-4.2.so")
            #self.path("libcrypto.so.0.9.7")
            #self.path("libexpat.so.1")
            #self.path("libssl.so.0.9.7")
            #self.path("libuuid.so.1")
            #self.path("libSDL-1.2.so.0")
            #self.path("libELFIO.so")
            #self.path("libopenjpeg.so.1.3.0", "libopenjpeg.so.1.3")
            #self.path("libalut.so")
            #self.path("libopenal.so", "libopenal.so.1")
            self.end_prefix("lib")
 
            # Vivox runtimes
            #if self.prefix(src="vivox-runtime/i686-linux", dst="bin"):
            #        self.path("SLVoice")
            #        self.end_prefix()
            #if self.prefix(src="vivox-runtime/i686-linux", dst="lib"):
            #        self.path("libortp.so")
            #        self.path("libvivoxsdk.so")
            #        self.end_prefix("lib")
</nowiki></pre>
 
Also, until [http://jira.secondlife.com/browse/VWR-9475 VWR-9475] ''really'' gets fixed (it was fixed before, but not really),
you need to install the 'SDL' precompiled library on standalone in order to get the needed cursor bitmaps (not necessary for Snowglobe 2.x):
 
% scripts/install.py SDL
 
However, if you are building standalone on 32-bit, that would also cause the wrong (non-system) headers and library to be included, so you will need to remove those again on 32-bit (from libraries/i686-linux/include/ and libraries/i686-linux/lib_release_client/).
 
== Installing the Non-Free Shared Libraries ==
'''Needs to be redone''' - [[User:Oz Linden|Oz Linden]]
 
== Compiling ==


This page describes how to build the Second Life viewer with [[CMake]]. CMake is a system for generating per-platform build files.  On Linux, it will generate your choice of Makefiles or KDevelop project files.
This page describes how to build the Second Life viewer with [[CMake]]. CMake is a system for generating per-platform build files.  On Linux, it will generate your choice of Makefiles or KDevelop project files.


'''NOTE:''' These instructions are for the viewer versions 1.21 and beyond.  For older viewers (1.20 and earlier) see [[Compiling the viewer with SCons (Linux)]]
'''NOTE:''' These instructions are for the viewers using cmake (versions 1.21 and beyond).  For older viewers (1.20 and earlier) see [[Compiling the viewer with SCons (Linux)]]


=== Configuring your tree ===
== Build the viewer with autobuild ==
Before you first run a build, you'll need to configure thingsThere's a <code>develop.py</code> script that will create a reasonably sane default configuration for you.
Configuring and building with autobuild works the same on all platformsFull instructions may be found at [[Build_Viewer_With_Autobuild]].


From the command line, '''cd into the indra subdirectory''' and run one of the following commands:
{{KBcaution| If you are using prepackaged 3rd party libraries from Linden (not a stand alone build) then you will need to build using gcc version 4.1; set CC and CXX accordingly. ''TODO: expand/update''}}
** make: <code>"./develop.py"</code>
** KDevelop: <code>"./develop.py -G KDevelop3"</code>


'''NOTE:''' The above commands will configure a "non-standalone" version of the source code tree. This means that the required third party library packages (as built by Linden Lab) will be downloaded during the CMake process.
=== Configuration notes ===
* To build for 'standalone' (see `[[#What_does_.27Standalone.27_mean.3F|What does 'standalone' mean?]]` above) choose an '''OpenSourceStandAlone''' option.
* If you want to use KDevelop add <code>-G KDevelop3</code>.
* Configuring a "non-standalone" version of the source code tree will cause the required (32-bit) third party library packages (as built by Linden Lab) to be downloaded during the CMake build process.
* '''OpenSource''' configurations do not install FMOD (which will be disabled).
* The environment variables CC and CXX are picked up automatically as usual. However, this is not the case of CXXFLAGS and LDFLAGS. You can pass the following to the configure commandline: <code>-DCMAKE_CXX_FLAGS:STRING="$CXXFLAGS" -DCMAKE_EXE_LINKER_FLAGS:STRING="$LDFLAGS"</code> to enable the use of these environment variables.
* You can make the build process more verbose (make it print the compiler commands it executes) by passing <code>-DCMAKE_VERBOSE_MAKEFILE:BOOL=ON</code> as well.
* As of version 2.0, passing <code>-DLL_TESTS:BOOL=FALSE</code> will stop running any tests (if any; depends on standalone and viewer version).
* If you want to restart the configuration process from scratch without any cached values affecting the configuration, erase the <code>CMakeCache.txt</code> in the build directory.
* Autobuild appears to bind the CXX and CC env. variables at configuration time.  So the following should work if done once:
**  CXX=g++-4.1 CC=gcc-4.1 autobuild configure -c Release


=== Starting the build ===
=== Where's the built viewer? ===


To start a build, do one of the following:
On Linux, your build will be at


# Run "./develop.py build"
  build-linux-ARCH/newview/packaged


-- or --
where "''ARCH''" should be something like "i686" or "x86_64" (depending on your platform).
{{KBcaution|Currently, the build dir will be <code>build-linux-'''i686'''</code>, even when a 64-bit build was requested by <code>autobuild configure [...] -- -DWORD_SIZE{{=}}64 [...]</code>. See {{jira|OPEN-106}}.}}


#  Find your build directory and change to it
=== Using ccache ===
#:  In the CMake world, we keep source and object files separate.  The <code>develop.py</code> script will create and populate a build directory for you.  On Linux, this will be named <code>viewer-linux-ARCH-BUILD</code> (where "ARCH" is "i686" or "x86_64", and "BUILD" varies on debug level)
#  Build with your preferred build tool
## Type <code>make</code>
##: -- or---
## Load it into KDevelop


=== Where's the built viewer? ===
It is very highly recommended that you use and install [http://ccache.samba.org/manual.html ccache] even before
the first compile; it will speed up subsequent compiles with a factor of 10 in case you need to redo the compile!
 
The easiest way to start using ccache is to make ''symbolic'' links to ccache
from <code>/usr/local/bin</code>. Make sure that <code>/usr/local/bin</code> comes first in your <code>PATH</code>,
before the real g++. Thus, as root and assuming you are using gcc/g++ (don't do this if you are using distcc, see below),
 
<code>
% apt-get install ccache
% cd /usr/local/bin
% ln -s ../../bin/ccache gcc
% ln -s ../../bin/ccache g++
</code>
 
And then to test, as normal user (make sure you are not in /usr/local/bin anymore),
 
<code>
% which g++
</code>


On Linux, your build will be here:
This should print <code>/usr/local/bin/g++</code>, if not fix you <code>PATH</code>.
<pre>
viewer-linux-ARCH-BUILD/newview/packaged
</pre>


...where "''ARCH''" is something like "i686" or "x86_64" (depending on your platform), and "''BUILD''" depends on which debug level you chose.
Note that by default ccache puts the cache in <code>$HOME/.ccache</code> and will grow till about 1 GB, so make sure you have that diskspace there. Alternatively you can set the environment variable <code>CCACHE_DIR</code> to change the location of the cache.


=== Using distcc ===
=== Using distcc ===


If you have multiple PCs, you can speed up the build process by using distcc.
If you have multiple PCs, you can speed up the build process by using distcc.
You probably should ''still'' use ccache which has to be run before distcc, therefore
both have to be specified in the <code>CXX</code> environment variable, by adding ccache to them.


To use distcc, you need to pass an environment variable <code>CXX</code> containing the distcc command prefix to <code>develop.py</code> when configuring, e.g.,  
To use distcc, you need to pass an environment variable <code>CXX</code> containing the distcc command prefix to <code>develop.py</code> when configuring, e.g.,  


  CXX="distcc g++" ./develop.py configure
  CXX="ccache distcc g++" ./develop.py configure


Or, if you are using a heterogeneous rows of PCs, you need to add a config prefix to <code>g++</code>, e.g. (in my case, as a Fedora user),
Or, if you are using a heterogeneous rows of PCs, you need to add a config prefix to <code>g++</code>, e.g. (in my case, as a Fedora user),


  CXX="distcc i386-redhat-linux-g++" ./develop.py configure
  CXX="ccache distcc i386-redhat-linux-g++" ./develop.py configure


Once you did so, the generated makefiles contain instructions to use distcc.  You don't need to specify anything special when invoking <code>develop.py</code> to build.  The number of concurrent jobs (i.e., <code>-j</code> option to <code>make</code> command) to run is automatically determined by <code>develop.py</code>.
Once you did so, the generated makefiles contain instructions to use distcc.  You don't need to specify anything special when invoking <code>develop.py</code> to build.  The number of concurrent jobs (i.e., <code>-j</code> option to <code>make</code> command) to run is automatically determined by <code>develop.py</code>.


If you invoke <code>make</code> command by yourself, don't forget appropriate number to <code>-j</code> option.
If you invoke <code>make</code> command by yourself, don't forget to add the <code>-j</code> option with an appropriate number.
 
= What to do if it doesn't work for you =
 
* Ask for help on [[IRC]] (irc.freenode.net #opensl)
* Find someone on the [[OpenSource-Dev|opensource-dev mailing list]]
* Fix it: [[Modifying CMake Files‎]] (and please, submit a patch!)
* If you are building on openSUSE 11.2 or newer, and your builds fail with lots of "undefined reference to ..." message: set the environment variable SUSE_ASNEEDED to 0 before building.
 
= Submitting Patches =
This is probably far down the road, but if you make changes to the source and want to submit them, see the page about [[Submitting patches|submitting patches]].


= Prebuilt libraries vs. standalone builds =
----


While many users will want to use the prebuilt libraries that we provide, we're also interested in making life as easy as possible for packagers who want to use their platform's native libraries.
{{KBcaution|
Everything below is probably outdated. It should be carefully sorted, cleaned up, updated and deleted (added 14/7/2010)
}}


If you run <code>ccmake</code>, you should see a <code>STANDALONE</code> option that determines whether the build will use your system's libraries or our prepackaged ones.  Flipping this to <code>ON</code> should be all you need to do to perform a packager-friendly build.
= Prebuilt libraries vs. standalone builds =


For standalone builds, we'd really like to beef up the checks for system libraries so that for example <code>cmake</code> will fail if a required library (such as OpenJPEG) isn't installed.  We welcome all patches that help out with this.
For standalone builds, we'd really like to beef up the checks for system libraries so that for example <code>cmake</code> will fail if a required library (such as OpenJPEG) isn't installed.  We welcome all patches that help out with this.
Line 186: Line 581:


* Preparing to run 'in-tree'
* Preparing to run 'in-tree'
** <i>ensure that you have indra/newview/app_settings/static_*.db2</i> - if not, you'll find it in the 'slviewer-artwork' download (a zip file).
** <i>ensure that you have indra/newview/app_settings/static_*.db2</i> - if not, you'll find it in the [[#slviewer-artwork-.3Cversion.3E.zip|slviewer-artwork]] download (a zip file).
** now, from the indra directory:
** now, from the indra directory:
<code>
<code>
Line 205: Line 600:
By default, the open-source Second Life Viewer uses the open-source OpenJPEG library to decode the (many) JPEG-2000 texture images it receives from the servers.  This isn't quite of comparable speed to the proprietary third-party library which the Linden Lab viewer builds have traditionally used, for which we are not permitted to redistribute the source.
By default, the open-source Second Life Viewer uses the open-source OpenJPEG library to decode the (many) JPEG-2000 texture images it receives from the servers.  This isn't quite of comparable speed to the proprietary third-party library which the Linden Lab viewer builds have traditionally used, for which we are not permitted to redistribute the source.


However, the <i>slviewer-linux-libs</i> package includes two pre-built libraries which facilitate the use of this slightly faster image decoding method: <b>libkdu_v42R.so</b> and <b>libllkdu.so</b>.  These are provided for your testing; again, we are not permitted to grant you the right to re-distribute these libraries to downstream users, but the viewer will still work (albeit slightly slower) without them.
However, the [[#slviewer-linux-libs-.3Cversion.3E.tar.gz|slviewer-linux-libs]] package includes two pre-built libraries which facilitate the use of this slightly faster image decoding method: <b>libkdu_v42R.so</b> and <b>libllkdu.so</b>.  These are provided for your testing; again, we are not permitted to grant you the right to re-distribute these libraries to downstream users, but the viewer will still work (albeit slightly slower) without them.


To <b>use these faster image-decoding libraries</b>, they simply need to be put into the right places relative to the viewer runtime directory - nothing needs to be reconfigured or recompiled.  If you're running the client from the source tree, the following will make the KDU libraries available:
To <b>use these faster image-decoding libraries</b>, they simply need to be put into the right places relative to the viewer runtime directory - nothing needs to be reconfigured or recompiled.  If you're running the client from the source tree, the following will make the KDU libraries available:
Line 247: Line 642:
Here are two scripts (one for v1.20 and older, and one for v1.21 and newer viewers) that basically do all what is described above, and more, and entitle you to compile a SL client very easily:
Here are two scripts (one for v1.20 and older, and one for v1.21 and newer viewers) that basically do all what is described above, and more, and entitle you to compile a SL client very easily:


* [[User:Henri Beauchamp/Automated Linux Build Script (1.20 and earlier)]]
* [[User:Henri Beauchamp/Automated Linux Build Script (1.20 and earlier)]]
* [[User:Henri Beauchamp/Building the viewer with CMake/cmake-SL script]]
* [[User:Henri Beauchamp/Building the viewer with CMake/cmake-SL script]]
 
* Please also see the (user contributed) instructions at [[User:Michelle2_Zenovka/cmake]]


=== FreeBSD ===
=== FreeBSD ===


A list of patches is given for [[Compiling the viewer (FreeBSD)]].
A list of patches is given for [[Compiling the viewer (FreeBSD)]]. They are obsolete, and no longer work.
 
= What to do if it doesn't work for you =
 
* Ask for help on [[IRC]] (irc.efnet.org #opensl)
* Find someone on the [[SLDev|SLDev mailing list]]
* Fix it: [[Modifying CMake Files‎]] (and please, submit a patch!)
 
Please also see the (user contributed) instructions at [[User:Michelle2_Zenovka/cmake]]
 
= Submitting Patches =
This is probably far down the road, but if you make changes to the source and want to submit them, see the page about [[Submitting patches|submitting patches]].


[[Category:Compiling viewer]]
[[Category:Compiling viewer]]
[[Category:Articles_in_need_of_updating]]

Latest revision as of 17:55, 5 February 2021

KBcaution.png Note Linden Lab no longer actively maintains Linux viewer third party dependencies. You may run into issues building the viewer on modern distros.

The following are instructions for building the Second Life viewer on linux. This process has been used on debian and debian based systems like ubuntu, and also on Fedora. For other platforms, see Get source and compile.

Required tools

There are a number of tools that need to be installed first.

  • hg [package: mercurial]
  • cmake [package: cmake]
    • Currently (snowglobe 1.3.2 and higher) cmake 2.6.2 is the minimum required version.
  • yacc or compatible tool [suggested package: bison]
  • lex or compatible tool [suggested package: flex]
  • python [package: python]
    • python 2.4.3 is the minimum required version.
    • All more recent 2.x versions should work, too. If you encounter problems or deprecation warnings, please report them.
    • python 3.x hasn't been tested yet.
  • g++ [package: g++]
    • Note (not relevant for standalone): gcc 4.4 (which is in recent Ubuntu and debian) won't work in versions prior to 2.0 (or Snowglobe 1.3) if you are not building standalone (see below), because it chokes on some parts of boost prior to 1.37 (http://svn.boost.org/trac/boost/ticket/2069). The solution is to install GCC 4.3 and to run 'export CXX=/usr/bin/g++-4.3' or whatever your binary is before trying to compile.
Another workaround is to use update-alternatives; for example, if you already have 4.4 installed, try this:
sudo apt-get install g++-4.3
sudo update-alternatives --remove-all gcc
sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-4.3 43  --slave /usr/bin/g++ g++ /usr/bin/g++-4.3  --slave /usr/bin/gcov gcov /usr/bin/gcov-4.3
sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-4.4 44  --slave /usr/bin/g++ g++ /usr/bin/g++-4.4  --slave /usr/bin/gcov gcov /usr/bin/gcov-4.4
then choose 4.3:
sudo update-alternatives --config gcc
  • Linden Lab has backported gcc-4.6.3 to Debian Squeeze; this port and its associated libraries are used when building the official Linux viewer.
The sources and Debian packages for this backport are available from Debian GCC Backport
  • libboost-program-options-dev was needed on Ubuntu 8.04 to use cmake, but it's not needed anymore to build a 1.23-render-pipeline on Ubuntu 9.04-beta (nor Snowglobe 2 on Ubuntu 9.10)
  • make (package: make)
  • bzip2 (package: bzip2)

The build process may use the following optional tools:

  • distcc distributed compiler (useful if you have multiple PCs.)
  • ccache a fast compiler cache (speed up recompilation, a must if you intend to do development.)
KBtip2.png Tip: To boost your productivity as a developer, try the suggestions at Development Environment for Multiple Viewers about directory structure, multiple viewers and automation of repetitive tasks.


What does 'Standalone' mean?

The current sources no longer use the term "standalone"; it has been replaced by "usesystemlibs".

A standalone build of the viewer refers to building a viewer against the shared libraries that are installed on your system instead of using precompiled libraries provided by Linden Lab.

In order to build standalone, you will have to configure using --standalone, see configuration below.

The advantage of building standalone is that you might use less RAM: you'll be using the same shared libraries that other applications use. However this is a rather minor advantage. The disadvantages of building standalone is a long list of potential problems: you will have to manually install all the development packages of many many libraries. In some cases there isn't a package for it from your distribution, so you have to especially get those sources separately and compile and install them in a way that the viewer can find them. Building standalone is not supported officially by Linden Lab and therefore not tested. Hence, it might not work at all (although the open source snowglobe developers usually make sure that it works for snowglobe). Nevertheless, every now and then it happens that the viewer won't build with some new released library (ie, boost), in which case you most likely will have to use the latest source code retrieved with subversion.

Nevertheless, there might be reasons that you want to do a standalone build anyway. The most important one being that Linden Lab also doesn't support 64-bit: they do not provided shared libraries for 64-bit; so, if you want to build native 64 bit you must use standalone. Another reason might be that Linden Lab is using a few rather old library versions that are incompatible with newer versions. If you have any reason to use a newer version for some library, then chances are that things will break unless you build standalone. One reason would be if you want to debug a library and fix it because it is buggy.

Conclusion: if you are building the viewer for the first time, and you are building for 32-bit, then you should not use standalone.

Getting the source

The source for the official Viewer's 2.x codebase resides in mercurial repositories hosted on http://bitbucket.org/lindenlab (a.k.a. http://bitbucket.org/lindenlab).

KBnote.png Old Versions Find instructions on how to obtain the sources for official 1.x codebase as well as Snowglobe 1.x and 2.x on the archived version of this section.

To check out the lindenlab/viewer-release branch, do

hg clone http://bitbucket.org/lindenlab/viewer-release

This will create a new folder called viewer-release inside your current working directory, containing the source tree. (You can optionally specify a different target directory. See hg help clone.) Other long-lived branches of interest are lindenlab/viewer-beta and lindenlab/viewer-development. Contributions should usually be based on viewer-development.

Installing the required libraries (that Linden Lab can not or does not provide)

Even non-standalone still uses a few shared libraries from your system. This paragraph deals with those libraries. For standalone you will need a lot more (see later on).

Libraries and header files that usually come with a Linux distribution

Make sure the libraries and header files for the following packages are installed on your system:

Library Debian/Ubuntu Fedora/Red Hat
libc.so libc6-dev glibc-devel
libstdc++.so.6 libstdc++6 libstdc++-devel
libX11.so libx11-dev libX11-devel
libGL.so nvidia-glx-dev or libgl1-mesa-dev mesa-libGL-devel
libXrender.so libxrender-dev libXrender-devel
libidn.so libidn11-dev libidn11-devel

shortcut commands for the above

Debian/Ubuntu:

sudo apt-get install libc6-dev libstdc++6 libx11-dev libxrender-dev libidn11-dev [libgl1-mesa-dev|nvidia-current-dev|...]

Fedora/Red Hat:

sudo yum install glibc-devel libstdc++-devel libX11-devel mesa-libGL-devel libXrender-devel libidn11-devel

Recommended libraries and headers

The following packages are required when building standalone, but even if you are building non-standalone you should probably install them.

Library Debian/Ubuntu Fedora/Red Hat
libGLU.so libglu1-mesa-dev mesa-libGLU-devel
libz.so zlib1g-dev zlib-devel
libssl.so libssl-dev openssl-devel
libogg.so libogg-dev libogg-devel
libpng12.so libpng12-dev libpng-devel
libdbus-glib-1.so libdbus-glib-1-dev dbus-glib-devel
libatk-1.0.so
libcairo.so
libgdk-x11-2.0.so
libgdk_pixbuf-2.0.so
libgmodule-2.0.so
libglib-2.0.so
libgtk-x11-2.0.so
libgthread-2.0.so
libpango-1.0.so
and many more
libgtk2.0-dev atk-devel cairo-devel gtk2-devel glib2-devel pango-devel

shortcut commands for the above (2)

Debian/Ubuntu:

sudo apt-get install libglu1-mesa-dev zlib1g-dev libssl-dev libogg-dev libpng12-dev libdbus-glib-1-dev libgtk2.0-dev

Fedora/Red Hat:

sudo yum install mesa-libGLU-devel zlib-devel openssl-devel libogg-devel libpng-devel dbus-glib-devel \
  atk-devel cairo-devel gtk2-devel glib2-devel pango-devel

Necessary libraries when building standalone

Existing package names

If you want to build 'standalone', then the following packages are required in addition:

Library Debian/Ubuntu Fedora/Red Hat openSUSE
libopenal.so libopenal-dev openal-soft-devel openal-soft-devel
libvorbis.so
libvorbisenc.so
libvorbisfile.so
libvorbis-dev libvorbis-devel libvorbis-devel
libalut.so libalut-dev freealut-devel freealut-devel
libapr-1.so libapr1-dev apr-devel libapr1-devel
libaprutil-1.so libaprutil1-dev apr-util-devel libapr-util1-devel
libboost_program_options.so
libboost_regex.so
libboost_signals.so
libboost-dev boost-devel boost-devel
libcares.so libc-ares-dev c-ares-devel libcares-devel
libxmlrpc-epi.so libxmlrpc-epi-dev ? libxmlrpc-epi0-devel
libopenjpeg.so libopenjpeg-dev openjpeg-devel openjpeg-devel
libjpeg.so libjpeg62-dev (libjpeg-dev) libjpeg-devel libjpeg62-devel
libSDL.so libsdl1.2-dev SDL-devel libSDL-devel
libgstreamer-0.10.so libgstreamer0.10-dev gstreamer-devel gstreamer-0_10-devel
gstreamer-0.10 plugins libgstreamer-plugins-base0.10-dev gstreamer-plugins-base-devel gstreamer-0_10-plugins-base-devel
libgmock.so google-mock (Snowglobe 2 only) ? ?

shortcut commands for the above (3)

Debian/Ubuntu:

sudo apt-get install libopenal-dev libvorbis-dev libalut-dev libapr1-dev libaprutil1-dev libboost-dev \
  libc-ares-dev libxmlrpc-epi-dev libopenjpeg-dev libjpeg62-dev libgtk2.0-dev libsdl1.2-dev \
  libgstreamer0.10-dev libgstreamer-plugins-base0.10-dev google-mock

Fedora/Red Hat:

sudo yum install openal-soft-devel libvorbis-devel freealut-devel apr-devel apr-util-devel boost-devel \
  c-ares-devel openjpeg-devel libjpeg-devel SDL-devel gstreamer-devel gstreamer-plugins-base-devel

openSUSE:

You will have to add the buildservice repository home:lemmy04:snowglobe (http://download.opensuse.org/repositories/home:/lemmy04:/snowglobe/) to your package sources.

sudo zypper install openal-soft-devel libvorbis-devel freealut-devel libapr1-devel libapr-util1-devel \
  boost-devel libcares-devel libxmlrpc-epi0-devel openjpeg-devel libjpeg62-devel libSDL-devel gstreamer-0_10-devel gstreamer-0_10-plugins-base-devel

More problematic libraries (standalone)

Easy so far, now here's where the trouble starts. Not all required packages are in debian, not to mention you might not even be using debian of course; but we (standalone builders) rely for the most part on the tremendous work done by 64-bit pioneer User:Robin_Cornelius who happens to use debian. Although many required packages are now in debian, some still need to be downloaded from Robin's private repository (assuming you're on debian). TODO: Add instructions for those not using ubuntu/debian.

Edit your /etc/apt/sources.list and add the following lines:

# Repository for SecondLife.
deb http://apt.byteme.org.uk squeeze main
deb-src http://apt.byteme.org.uk squeeze main

or

# Repository for SecondLife.
deb http://apt.byteme.org.uk lenny main
deb-src http://apt.byteme.org.uk lenny main

depending on whether you are using testing (squeeze) or stable (lenny). Note that you won't be able to compile Snowglobe 2.x on lenny because the Qt library is too old.

Run apt-get update to get the Package files from byteme as usual. You can ignore the warning about the missing public key, or do what is described at the bottom of this page under 'GPG Keys and Signed repository' (note the remark about not using root), thus:

% gpg --keyserver hkp://keyserver.ubuntu.com --recv-key 0x729A79A23B7EE764
% gpg --export 0x729A79A23B7EE764 | sudo apt-key add -

or for ubuntu users we use the openmetaverse repo, run this command to add the repo:

# Openmetaverse repo
sudo add-apt-repository ppa:openmetaverse/ppa

NOTE: If you are on Ubuntu 12.04 ("Precise") you will also need to edit the file /etc/apt/sources.list.d/openmetaverse-ppa-precise.list so that it refers to the "lucid" distro instead of "precise" because the maintainers haven't made packages for "precise" yet.

Change /etc/apt/source.list.d/openmetaverse-ppa-precise.list from:

deb http://ppa.launchpad.net/openmetaverse/ppa/ubuntu precise main
deb-src http://ppa.launchpad.net/openmetaverse/ppa/ubuntu precise main

To be:

deb http://ppa.launchpad.net/openmetaverse/ppa/ubuntu lucid main
deb-src http://ppa.launchpad.net/openmetaverse/ppa/ubuntu lucid main

and of course to add the keys (only on Debian)

gpg --keyserver hkp://keyserver.ubuntu.com --recv-key 0x50A40F50
gpg --export 0x50A40F50 | sudo apt-key add -

and always remember to update after adding repos

apt-get update

Finally install these packages:

Library Debian/Ubuntu
libcurl-cares.so libcurl4-cares-dev
llqtwebkit libllqtwebkit-dev (Snowglobe 1) or
libllqtwebkit2-dev (Snowglobe 2)
JsonCpp libjsoncpp-dev
libndofdev libndofdev-dev (optional, needed if you have a Space Navigator joystick)
glh-linear glh-linear (Snowglobe 2 only)
boost-coroutine boost-coroutine (Snowglobe 2 only)


Shortcut commands for the above

sudo apt-get install libcurl4-cares-dev libllqtwebkit2-dev glh-linear boost-coroutine libndofdev-dev

If you are on lenny you will also (automatically) install the packages cmake, libopenjpeg-dev and libopenjpeg2 from byteme. This is because the packages in lenny contain bugs.

The libjsoncpp-dev installs a library called /usr/lib/libjson.so, but the viewer is looking for one called libjson_linux-gcc-${_gcc_COMPILER_VERSION}_libmt.so in '/usr/lib' and '/usr/local/lib'. Therefore execute the following command as root:

# This is not needed on Snowglobe 1.4.1 and higher.
% ln -s /usr/lib/libjson.so /usr/local/lib/libjson_linux-gcc-$(g++ -dumpversion)_libmt.so

In order to run the tests, you need a template library called 'tut'. Because this is a template library, we can just use the library archive from Linden Lab for any architecture. Therefore, to install tut, run:

% scripts/install.py tut

Until recently for some completely mysterious reason, the source code includes <tut/tut.hpp> while configure checks for tut.h in /usr/lib and /usr/local/lib only (CXXFLAGS? Never heard of that). In order to fix this brokenness for non-Snowglobe viewers or based on older snowglobe versions (prior to 1.4, or 2.1), execute the following (2.x viewers older than 2.1 and 1.x viewers older than 1.4 only):

# This is not needed on Snowglobe 1.4.1 and higher.
% sudo touch /usr/local/include/tut.h

which satisfies the configure check. The installed tut.hpp is still used anyway since libraries/include is in the include search path by default. Alternatively, edit indra/cmake/Tut.cmake and comment out these two lines:

 #include(FindTut)
 #include_directories(${TUT_INCLUDE_DIR})

Oh WAIT! On 2.0 standalone libraries/include is NOT included anymore, and it seems impossible to fix the cmake spaghetti so here's what you have to do on top of the above:

sudo cp -r libraries/include/tut /usr/local/include

Sorry but there is no other way.

However, if you are using snowglobe 1.4 or a later 1.x, or snowglobe 2.1 or later, or a viewer based on those, then the viewer actually looks for tut/tut.hpp. So, you can still install it in /usr/local/include/tut/, but you can also install it in -say- /somewhere/include/tut/ and set the environment variable CMAKE_INCLUDE_PATH (a colon separated list of paths) to include /somewhere/include before configuration. Note that due to a cmake problem/bug it will find /somewhere/include but still use the include path of any other library, including /usr/include or /usr/local/include, so make sure you remove any other tut installation from common paths if you use the latter method.

Get voice chat working (standalone)

The voice binaries downloaded for non-standalone builds can be used in for standalone builds, too. Because they'll run as a separate process, this even works with 64-bit builds of the viewer (assuming your system is set up for also running 32-bit binaries). To download and unpack the binaries (which doesn't happen automatically when building standalone) run:

${SOURCE_DIR}/scripts/install.py slvoice
KBnote.png Note: In earlier revisions, the package was named vivox rather than slvoice.

After building, you'll have to copy the binaries into the right place (-- I don't think this is correct as this will also link the viewer against those libraries. Instead you have to use a script for SLVoice that sets the right LD_LIBRARY_PATH and then runs the real SLVoice --Aleric Inglewood 12:03, 13 July 2010 (UTC)):

cp ${SOURCE_DIR}/indra/newview/vivox-runtime/i686-linux/SLVoice ${BUILD_DIR}/newview/packaged/bin/
cp ${SOURCE_DIR}/indra/newview/vivox-runtime/i686-linux/*.so* ${BUILD_DIR}/newview/packaged/lib/

If you want, you can then remove the voice binaries again from the source tree with

${SOURCE_DIR}/scripts/install.py --uninstall vivox

Remaining issues (standalone)

In order to run the viewer you need to create a 'package' (a directory really, to which everything is copied that is needed to run it). Unfortunately, if you are building standalone on a 32-bit machine, it still tries to happily copy the Linden Lab provided shared libraries over, which we never even downloaded mind you, causing the packaging to fail. Therefore, edit indra/newview/viewer_manifest.py and comment out the following lines in class Linux_i686Manifest(LinuxManifest):

            #self.path("libapr-1.so.0")
            #self.path("libaprutil-1.so.0")
            #self.path("libdb-4.2.so")
            #self.path("libcrypto.so.0.9.7")
            #self.path("libexpat.so.1")
            #self.path("libssl.so.0.9.7")
            #self.path("libuuid.so.1")
            #self.path("libSDL-1.2.so.0")
            #self.path("libELFIO.so")
            #self.path("libopenjpeg.so.1.3.0", "libopenjpeg.so.1.3")
            #self.path("libalut.so")
            #self.path("libopenal.so", "libopenal.so.1")
            self.end_prefix("lib")

            # Vivox runtimes
            #if self.prefix(src="vivox-runtime/i686-linux", dst="bin"):
            #        self.path("SLVoice")
            #        self.end_prefix()
            #if self.prefix(src="vivox-runtime/i686-linux", dst="lib"):
            #        self.path("libortp.so")
            #        self.path("libvivoxsdk.so")
            #        self.end_prefix("lib")

Also, until VWR-9475 really gets fixed (it was fixed before, but not really), you need to install the 'SDL' precompiled library on standalone in order to get the needed cursor bitmaps (not necessary for Snowglobe 2.x):

% scripts/install.py SDL

However, if you are building standalone on 32-bit, that would also cause the wrong (non-system) headers and library to be included, so you will need to remove those again on 32-bit (from libraries/i686-linux/include/ and libraries/i686-linux/lib_release_client/).

Installing the Non-Free Shared Libraries

Needs to be redone - Oz Linden

Compiling

This page describes how to build the Second Life viewer with CMake. CMake is a system for generating per-platform build files. On Linux, it will generate your choice of Makefiles or KDevelop project files.

NOTE: These instructions are for the viewers using cmake (versions 1.21 and beyond). For older viewers (1.20 and earlier) see Compiling the viewer with SCons (Linux)

Build the viewer with autobuild

Configuring and building with autobuild works the same on all platforms. Full instructions may be found at Build_Viewer_With_Autobuild.

KBcaution.png Important: If you are using prepackaged 3rd party libraries from Linden (not a stand alone build) then you will need to build using gcc version 4.1; set CC and CXX accordingly. TODO: expand/update

Configuration notes

  • To build for 'standalone' (see `What does 'standalone' mean?` above) choose an OpenSourceStandAlone option.
  • If you want to use KDevelop add -G KDevelop3.
  • Configuring a "non-standalone" version of the source code tree will cause the required (32-bit) third party library packages (as built by Linden Lab) to be downloaded during the CMake build process.
  • OpenSource configurations do not install FMOD (which will be disabled).
  • The environment variables CC and CXX are picked up automatically as usual. However, this is not the case of CXXFLAGS and LDFLAGS. You can pass the following to the configure commandline: -DCMAKE_CXX_FLAGS:STRING="$CXXFLAGS" -DCMAKE_EXE_LINKER_FLAGS:STRING="$LDFLAGS" to enable the use of these environment variables.
  • You can make the build process more verbose (make it print the compiler commands it executes) by passing -DCMAKE_VERBOSE_MAKEFILE:BOOL=ON as well.
  • As of version 2.0, passing -DLL_TESTS:BOOL=FALSE will stop running any tests (if any; depends on standalone and viewer version).
  • If you want to restart the configuration process from scratch without any cached values affecting the configuration, erase the CMakeCache.txt in the build directory.
  • Autobuild appears to bind the CXX and CC env. variables at configuration time. So the following should work if done once:
    • CXX=g++-4.1 CC=gcc-4.1 autobuild configure -c Release

Where's the built viewer?

On Linux, your build will be at

build-linux-ARCH/newview/packaged

where "ARCH" should be something like "i686" or "x86_64" (depending on your platform).

KBcaution.png Important: Currently, the build dir will be build-linux-i686, even when a 64-bit build was requested by autobuild configure [...] -- -DWORD_SIZE=64 [...]. See OPEN-106.

Using ccache

It is very highly recommended that you use and install ccache even before the first compile; it will speed up subsequent compiles with a factor of 10 in case you need to redo the compile!

The easiest way to start using ccache is to make symbolic links to ccache from /usr/local/bin. Make sure that /usr/local/bin comes first in your PATH, before the real g++. Thus, as root and assuming you are using gcc/g++ (don't do this if you are using distcc, see below),

% apt-get install ccache
% cd /usr/local/bin
% ln -s ../../bin/ccache gcc
% ln -s ../../bin/ccache g++

And then to test, as normal user (make sure you are not in /usr/local/bin anymore),

% which g++

This should print /usr/local/bin/g++, if not fix you PATH.

Note that by default ccache puts the cache in $HOME/.ccache and will grow till about 1 GB, so make sure you have that diskspace there. Alternatively you can set the environment variable CCACHE_DIR to change the location of the cache.

Using distcc

If you have multiple PCs, you can speed up the build process by using distcc. You probably should still use ccache which has to be run before distcc, therefore both have to be specified in the CXX environment variable, by adding ccache to them.

To use distcc, you need to pass an environment variable CXX containing the distcc command prefix to develop.py when configuring, e.g.,

CXX="ccache distcc g++" ./develop.py configure

Or, if you are using a heterogeneous rows of PCs, you need to add a config prefix to g++, e.g. (in my case, as a Fedora user),

CXX="ccache distcc i386-redhat-linux-g++" ./develop.py configure

Once you did so, the generated makefiles contain instructions to use distcc. You don't need to specify anything special when invoking develop.py to build. The number of concurrent jobs (i.e., -j option to make command) to run is automatically determined by develop.py.

If you invoke make command by yourself, don't forget to add the -j option with an appropriate number.

What to do if it doesn't work for you

  • Ask for help on IRC (irc.freenode.net #opensl)
  • Find someone on the opensource-dev mailing list
  • Fix it: Modifying CMake Files‎ (and please, submit a patch!)
  • If you are building on openSUSE 11.2 or newer, and your builds fail with lots of "undefined reference to ..." message: set the environment variable SUSE_ASNEEDED to 0 before building.

Submitting Patches

This is probably far down the road, but if you make changes to the source and want to submit them, see the page about submitting patches.


KBcaution.png Important:

Everything below is probably outdated. It should be carefully sorted, cleaned up, updated and deleted (added 14/7/2010)

Prebuilt libraries vs. standalone builds

For standalone builds, we'd really like to beef up the checks for system libraries so that for example cmake will fail if a required library (such as OpenJPEG) isn't installed. We welcome all patches that help out with this.


Testing and packaging the client

Testing the result from inside the tree

You may find it simpler and less error-prone to follow the instructions in the Packaging the client section below to run the client under the same conditions as an end-user would.

  • 2008-05-29 (Ochi Wolfe): Compiling the 1.20.7 r88152 viewer, it seems like even when compiling as "release" the viewer is built ready-to-go inside the newview/packaged/ directory including the message_template.msg and message.xml in the right place. Try to cd to the newview/packaged/ directory and run SL from there with the ./secondlife command as you would normally do.

Otherwise:

  • Preparing to run 'in-tree'
    • ensure that you have indra/newview/app_settings/static_*.db2 - if not, you'll find it in the slviewer-artwork download (a zip file).
    • now, from the indra directory:

$ cp ../scripts/messages/message_template.msg newview/app_settings/
$ cp ../etc/message.xml newview/app_settings/

Important: Starting from version 1.18.0, copying message.xml is also required. Missing it will cause group IMs to fail to work, although the viewer will run fine otherwise.

  • Running it: The LD_LIBRARY_PATH stuff ensures that the binary looks for its libraries in the right places. From the indra directory:

$ ( cd newview && LD_LIBRARY_PATH="`pwd`"/../../libraries/i686-linux/lib_release_client:"`pwd`"/app_settings/mozilla-runtime-linux-i686:${LD_LIBRARY_PATH}:/usr/local/lib  ./secondlife-i686-bin )


The client seems kinda slow.

By default, the open-source Second Life Viewer uses the open-source OpenJPEG library to decode the (many) JPEG-2000 texture images it receives from the servers. This isn't quite of comparable speed to the proprietary third-party library which the Linden Lab viewer builds have traditionally used, for which we are not permitted to redistribute the source.

However, the slviewer-linux-libs package includes two pre-built libraries which facilitate the use of this slightly faster image decoding method: libkdu_v42R.so and libllkdu.so. These are provided for your testing; again, we are not permitted to grant you the right to re-distribute these libraries to downstream users, but the viewer will still work (albeit slightly slower) without them.

To use these faster image-decoding libraries, they simply need to be put into the right places relative to the viewer runtime directory - nothing needs to be reconfigured or recompiled. If you're running the client from the source tree, the following will make the KDU libraries available:

cp "$SLSRC/libraries/i686-linux/lib_release_client/libllkdu.so" "$SLSRC/indra/newview/libllkdu.so"
mkdir "$SLSRC/indra/lib"
cp "$SLSRC/libraries/i686-linux/lib_release_client/libkdu_v42R.so" "$SLSRC/indra/lib/libkdu_v42R.so"

The file indra/newview/viewer_manifest.py contains some commented-out entries describing where these libraries belong; if you uncomment the two lines corresponding to libllkdu and libkdu then they will be automatically copied into the right place in the runtime directory when you follow the 'Packaging the client' instructions below.

File Dialogs Don't Work on 64 bit system

If you run a 64 bit system, and your file dialogs don't work, or they worked before and stopped after you installed an update, it may be due to a mismatch between the headers used to compile the viewer and the library it's using. The log will contain something like this:

2007-06-21T01:28:35Z INFO: ll_try_gtk_init: Starting GTK Initialization.
2007-06-21T01:28:36Z INFO: ll_try_gtk_init: GTK Initialized.
2007-06-21T01:28:36Z INFO: ll_try_gtk_init: - Compiled against GTK version 2.10.11
2007-06-21T01:28:36Z INFO: ll_try_gtk_init: - Running against GTK version 2.10.6
2007-06-21T01:28:36Z WARNING: ll_try_gtk_init: - GTK COMPATIBILITY WARNING: Gtk+ version too old (micro mismatch)

What happens here is that your distribution includes 32 bit GTK libraries, but the package only includes the libraries themselves and not the headers. When building, the SL client will build against the headers included with the main 64 bit GTK package. This will work if the 64 bit version of the library is the same or older than the 32 bit one. However, if your 32 bit library is older, then the viewer will detect the mismatch (built with headers for a newer version of GTK than it's using) and turn GTK off.

Possible solutions:

  • Download the source for the version of the 32 bit GTK libraries your distribution comes with, and build your viewer against those headers.
  • Upgrade your 32 bit GTK package so that it's the same or newer as the 64 bit one.
  • Downgrade your 64 bit package (may not be a good idea).

Packaging the client

If you substitute 'BUILD=release' with 'BUILD=releasefordownload' in the 'Compiling' section above, then packaging the resulting code, libraries, data and documentation into a tarball for the end-user will be done automatically as the final stage of the build process; the pristine end-user client distribution has been assembled into the directory indra/newview/SecondLife_i686_1_X_Y_Z/ and has also been tarred into indra/newview/SecondLife_i686_1_X_Y_Z.tar.bz2

The file which controls what (and where) files go into the end-user runtime viewer directory is indra/newview/viewer_manifest.py


Resident contributed instructions

Automated libraries and headers adjustments, compilation and packaging

Here are two scripts (one for v1.20 and older, and one for v1.21 and newer viewers) that basically do all what is described above, and more, and entitle you to compile a SL client very easily:

FreeBSD

A list of patches is given for Compiling the viewer (FreeBSD). They are obsolete, and no longer work.