- command.
- much more convenient.
- searched by default.
- line options to most client programs.
-
-
-
Getting Started
-
- The following is a quick summary of how to get
PostgreSQL> up and
- running once installed. The main documentation contains more information.
-
-
-
- Create a user account for the
PostgreSQL>
- server. This is the user the server will run as. For production
- use you should create a separate, unprivileged account
- (postgres> is commonly used). If you do not have root
- access or just want to play around, your own user account is
- enough, but running the server as root is a security risk and
- will not work.
-
-adduser postgres>
-
-
-
-
-
- Create a database installation with the initdb>
- command. To run initdb> you must be logged in to your
-
PostgreSQL> server account. It will not work as
- root.
-
-root# mkdir /usr/local/pgsql/data>
-root# chown postgres /usr/local/pgsql/data>
-root# su - postgres>
-postgres$ /usr/local/pgsql/bin/initdb -D /usr/local/pgsql/data>
-
-
-
- The
- will be stored. You can use any path you want, it does not have
- to be under the installation directory. Just make sure that the
- server account can write to the directory (or create it, if it
- doesn't already exist) before starting initdb>, as
- illustrated here.
-
-
-
-
- The previous step should have told you how to start up the
- database server. Do so now. The command should look something
- like
-/usr/local/pgsql/bin/postmaster -D /usr/local/pgsql/data
-
- This will start the server in the foreground. To put the server
- in the background use something like
-nohup /usr/local/pgsql/bin/postmaster -D /usr/local/pgsql/data \
- </dev/null >>server.log 2>&1 </dev/null &
-
-
-
- To stop a server running in the background you can type
-kill `cat /usr/local/pgsql/data/postmaster.pid`
-
-
-
- In order to allow TCP/IP connections (rather than only Unix
- domain socket ones) you need to pass the
- postmaster>.
-
-
-
-
- Create a database:
-
-createdb testdb>
-
- Then enter
-
-psql testdb>
-
- to connect to that database. At the prompt you can enter SQL
- commands and start experimenting.
-
-
-
-
-
-
-
What Now?
-
-
-
- The
PostgreSQL> distribution contains a
- comprehensive documentation set, which you should read sometime.
- After installation, the documentation can be accessed by
- pointing your browser to
- /usr/local/pgsql/doc/html/index.html>, unless you
- changed the installation directories.
-
-
- The first few chapters of the main documentation are the Tutorial,
- which should be your first reading if you are completely new to
-
SQL> databases. If you are familiar with database
- concepts then you want to proceed with part on server
- administration, which contains information about how to set up
- the database server, database users, and authentication.
-
-
-
-
- Usually, you will want to modify your computer so that it will
- automatically start the database server whenever it boots. Some
- suggestions for this are in the documentation.
-
-
-
-
- Run the regression tests against the installed server (using
- gmake installcheck). If you didn't run the
- tests before installation, you should definitely do it now. This
- is also explained in the documentation.
-
-
-
-
- By default,
PostgreSQL> is configured to run on
- minimal hardware. This allows it to start up with almost any
- hardware configuration. The default configuration is, however,
- not designed for optimum performance. To achieve optimum
- performance, several server parameters must be adjusted, the two
- most common being shared_buffers and
- work_mem.
- Other parameters mentioned in the documentation also affect
- performance.
-
-
-
-
-
-]]>
-
-
-
-
Supported Platforms
-
-
PostgreSQL> has been verified by the developer
- community to work on the platforms listed below. A supported
- platform generally means that
PostgreSQL> builds and
- installs according to these instructions and that the regression
- tests pass. Build farm
entries refer to builds
- reported by the
PostgreSQL
- Build Farm.
-
-
-
- If you are having problems with the installation on a supported
- listed here.
-
-
-
-
-
-
- |
- Processor
- Version
- Reported
- Remarks
-
-
-
- |
- AIX>
- PowerPC>
- 8.0.0
- see also doc/FAQ_AIX
-
- |
- AIX>
- RS6000>
- 8.0.0
- see also doc/FAQ_AIX
-
- |
- BSD/OS>
- x86>
- 8.0.0
- 4.3.1
-
- |
- Debian GNU/Linux>
- Alpha>
- 7.4
-
-
- |
- Debian GNU/Linux>
- AMD64>
- 8.0.0
- Build farm panda, snapshot 2004-12-06 01:20:02
- sid, kernel 2.6
-
- |
- Debian GNU/Linux>
- arm41>
- 7.4
-
-
- |
- Debian GNU/Linux>
- Itanium>
- 7.4
-
-
- |
- Debian GNU/Linux>
- m68k>
- 8.0.0
- sid
-
- |
- Debian GNU/Linux>
- MIPS>
- 8.0.0
- Build farm lionfish, snapshot 2004-12-06 11:00:08
- 3.1 (sarge), kernel 2.4
-
- |
- Debian GNU/Linux>
- PA-RISC>
- 8.0.0
-
-
- |
- Debian GNU/Linux>
- PowerPC>
- 8.0.0
-
-
- |
- Debian GNU/Linux>
- S/390>
- 7.4
-
-
- |
- Debian GNU/Linux>
- Sparc>
- 8.0.0
- sid, 32-bit
-
- |
- Debian GNU/Linux>
- x86>
- 8.0.0
- 3.1 (sarge), kernel 2.6
-
- |
- Fedora>
- AMD64>
- 8.0.0
- FC3
-
- |
- Fedora>
- x86>
- 8.0.0
- Build farm dog, snapshot 2004-12-06 02:06:01
- FC1
-
- |
- FreeBSD>
- Alpha>
- 7.4
- 4.8
-
- |
- FreeBSD>
- x86>
- 8.0.0
- Build farm cockatoo, snapshot 2004-12-06 14:10:01 (4.10);
-
-
- |
- Gentoo Linux>
- x86>
- 8.0.0
-
-
- |
- HP-UX>
- PA-RISC>
- 7.4
-
-
- gcc> and cc>; see also doc/FAQ_HPUX
-
- |
- IRIX>
- MIPS>
- 7.4
- 6.5.20, cc only
-
- |
- Mac OS X>
- PowerPC>
- 8.0.0
- 10.3.5
-
- |
- Mandrakelinux>
- x86>
- 8.0.0
- Build farm shrew, snapshot 2004-12-06 02:02:01
- 10.0
-
- |
- NetBSD>
- arm32>
- 7.4
- 1.6ZE/acorn32
-
- |
- NetBSD>
- Sparc>
- 7.4.1
- 1.6.1, 32-bit
-
- |
- NetBSD>
- x86>
- 8.0.0
- Build farm canary, snapshot 2004-12-06 03:30:00
- 1.6
-
- |
- OpenBSD>
- Sparc>
- 7.4
- 3.4
-
- |
- OpenBSD>
- x86>
- 8.0.0
- Build farm emu, snapshot 2004-12-06 11:35:03
- 3.6
-
- |
- Red Hat Linux>
- AMD64>
- 8.0.0
- RHEL 3AS
-
- |
- Red Hat Linux>
- IA64>
- 8.0.0
- RHEL 3AS
-
- |
- Red Hat Linux>
- PowerPC>
- 8.0.0
- RHEL 3AS
-
- |
- Red Hat Linux>
- PowerPC 64>
- 8.0.0
- RHEL 3AS
-
- |
- Red Hat Linux>
- S/390>
- 8.0.0
- RHEL 3AS
-
- |
- Red Hat Linux>
- S/390x>
- 8.0.0
- RHEL 3AS
-
- |
- Red Hat Linux>
- x86>
- 8.0.0
- RHEL 3AS
-
- |
- Solaris>
- Sparc>
- 8.0.0
- Solaris 8; see also doc/FAQ_Solaris
-
- |
- Solaris>
- x86>
- 8.0.0
- Build farm kudu, snapshot 2004-12-10 02:30:04 (cc);
- dragonfly, snapshot 2004-12-09 04:30:00 (gcc)
- Solaris 9; see also doc/FAQ_Solaris
-
- |
- Tru64 UNIX>
- Alpha>
- 7.4
-
-
- |
- UnixWare>
- x86>
- 8.0.0
- cc, 7.1.4; see also doc/FAQ_SCO
-
- |
- Windows>
- x86>
- 8.0.0
- see doc/FAQ_MINGW
-
- |
- x86>
- 8.0.0
- Build farm gibbon, snapshot 2004-12-11 01:33:01
- see doc/FAQ_CYGWIN
-
-
-
-
-
-
-
Unsupported Platforms:
- The following platforms are either known not to work, or they used
- to work in a previous release and we did not receive explicit
- confirmation of a successful test with version &majorversion; at
- the time this list was compiled. We include these here to let you
- know that these platforms could> be supported if given
- some attention.
-
-
-
-
-
-
- |
- Processor
- Version
- Reported
- Remarks
-
-
-
-
- |
- BeOS>
- x86>
- 7.2
- 2001-11-29,
- needs updates to semaphore code
-
- |
- Linux>
- PlayStation 2>
- 7.4
- 2003-11-02,
-
- needs new config.guess,
-
-
-
- |
- NetBSD>
- Alpha>
- 7.2
- 2001-11-20,
- 1.5W
-
- |
- NetBSD>
- MIPS>
- 7.2.1
- 2002-06-13,
- 1.5.3
-
- |
- NetBSD>
- PowerPC>
- 7.2
- 2001-11-28,
- 1.5
-
- |
- NetBSD>
- VAX>
- 7.1
- 2001-03-30,
- 1.5
-
- |
- QNX 4 RTOS>
- x86>
- 7.2
- 2001-12-10,
-
- needs updates to semaphore code;
- see also doc/FAQ_QNX4
-
- |
- QNX RTOS v6>
- x86>
- 7.2
- patches available in archives, but too late for 7.2
-
- |
- SCO OpenServer>
- x86>
- 7.3.1
- 2002-12-11,
- 5.0.4, gcc>; see also doc/FAQ_SCO
-
- |
- SunOS 4>
- Sparc>
- 7.2
-
-
-
-
-
-
-
-
-
-
+\r
+\r
+\r
+
PostgreSQL>]]>\r
+ Installation Instructions\r
+\r
+ \r
+ \r
+\r
+ This \r
+ describes the installation of\r
+
PostgreSQL from the source code
\r
+ distribution.\r
+ \r
+\r
+ \r
+
Short Version\r
+\r
+\r
+./configure\r
+gmake\r
+su\r
+gmake install\r
+adduser postgres\r
+mkdir /usr/local/pgsql/data\r
+chown postgres /usr/local/pgsql/data\r
+su - postgres\r
+/usr/local/pgsql/bin/initdb -D /usr/local/pgsql/data\r
+/usr/local/pgsql/bin/postmaster -D /usr/local/pgsql/data >logfile 2>&1 &\r
+/usr/local/pgsql/bin/createdb test\r
+/usr/local/pgsql/bin/psql test\r
+\r
+ The long version is the rest of this\r
+ \r
+ \r
+ \r
+ \r
+\r
+\r
+ \r
+
Requirements\r
+\r
+ In general, a modern Unix-compatible platform should be able to run\r
+ The platforms that had received specific testing at the\r
+ time of release are listed in \r
+ below. In the doc> subdirectory of the distribution\r
+ there are several platform-specific
FAQ> documents you\r
+ might wish to consult if you are having trouble.\r
+ \r
+\r
+ The following software packages are required for building\r
+\r
+ \r
+ \r
+ \r
+ \r
+\r
+
GNU> make> is required; other\r
+
make> programs will not> work.\r
+
GNU> make> is often installed under\r
+ the name gmake; this document will always\r
+ refer to it by that name. (On some systems\r
+
GNU make> is the default tool with the name\r
+
make>.) To test for GNU\r
+\r
+gmake --version\r
+\r
+ It is recommended to use version 3.76.1 or later.\r
+ \r
+ \r
+\r
+ \r
+ You need an
ISO>/ANSI> C compiler. Recent\r
+ versions of
GCC> are recommendable, but\r
+
PostgreSQL> is known to build with a wide variety\r
+ of compilers from different vendors.\r
+ \r
+ \r
+\r
+ \r
+
gzip> is needed to unpack the distribution in the\r
+ first place.\r
+ past that hurdle.]]>\r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+\r
+ The
GNU> Readline> library (for\r
+ comfortable line editing and command history retrieval) will be\r
+ used by default. If you don't want to use it then you must\r
+ specify the option for\r
+
configure>. (On NetBSD,\r
+ the libedit library is\r
+
Readline-compatible and is used if
\r
+ libreadline is not found.)\r
+ \r
+ \r
+\r
+ \r
+ \r
+ on Windows\r
+ \r
+\r
+ To build on
NT>-based versions of\r
+
Windows> like Windows XP and 2003 see\r
+
doc/FAQ_MINGW>. For earlier Windows>\r
+ releases see doc/FAQ_CYGWIN>.\r
+\r
+ To build
Windows> client-only interfaces using\r
+ tools like
Visual C++> and Borland\r
+ C++> see\r
+ \r
+ Installation on Windows"]]> \r
+ linkend="install-win32">]]>.\r
+ \r
+ \r
+ \r
+ \r
+\r
+ The following packages are optional. They are not required in the\r
+ default configuration, but they are needed when certain build\r
+ options are enabled, as explained below.\r
+\r
+ \r
+ \r
+ To build the server programming language\r
+
PL/Perl you need a full
\r
+
Perl installation, including the
\r
+ libperl library and the header files.\r
+ Since
PL/Perl will be a shared
\r
+ libperl library must be a shared library\r
+ also on most platforms. This appears to be the default in\r
+ recent
Perl versions, but it was not
\r
+ in earlier versions, and in general it is the choice of whomever\r
+ installed Perl at your site.\r
+ \r
+\r
+ If you don't have the shared library but you need one, a message\r
+ like this will appear during the build to point out this fact:\r
+\r
+*** Cannot build PL/Perl because libperl is not a shared library.\r
+*** You might have to rebuild your Perl installation. Refer to\r
+*** the documentation for details.\r
+\r
+ (If you don't follow the on-screen output you will merely notice\r
+ that the
PL/Perl library object,
\r
+ plperl.so or similar, will not be\r
+ installed.) If you see this, you will have to rebuild and\r
+ install
Perl manually to be able to
\r
+ build
PL/Perl. During the
\r
+ configuration process for
Perl,
\r
+ request a shared library.\r
+ \r
+ \r
+\r
+ \r
+ To build the
PL/Python> server programming\r
+ language, you need a
Python\r
+ installation with the header files and the
distutils module.
\r
+ The
distutils module is included by default with
\r
+
Python 1.6 and later; users of
\r
+ earlier versions of
Python will need
\r
+ to install it.\r
+ \r
+\r
+ Since
PL/Python will be a shared
\r
+ library, the
libpython\r
+ libpython library must be a shared library\r
+ also on most platforms. This is not the case in a default\r
+
Python installation. If after
\r
+ building and installing you have a file called\r
+ plpython.so (possibly a different\r
+ extension), then everything went well. Otherwise you should\r
+ have seen a notice like this flying by:\r
+\r
+*** Cannot build PL/Python because libpython is not a shared library.\r
+*** You might have to rebuild your Python installation. Refer to\r
+*** the documentation for details.\r
+\r
+ That means you have to rebuild (part of) your\r
+
Python installation to supply this
\r
+ shared library.\r
+ \r
+\r
+ If you have problems, run
Python> 2.3 or later's\r
+ configure using the --enable-shared> flag. On some\r
+ operating systems you don't have to build a shared library, but\r
+ you will have to convince the
PostgreSQL> build\r
+ system of this. Consult the Makefile in\r
+ the src/pl/plpython directory for details.\r
+ \r
+ \r
+\r
+ \r
+ If you want to build the
PL/Tcl\r
+ procedural language, you of course need a Tcl installation.\r
+ \r
+ \r
+\r
+ \r
+ To enable Native Language Support (
NLS), that
\r
+ is, the ability to display a program's messages in a language\r
+ other than English, you need an implementation of the\r
+
Gettext> API. Some operating\r
+ systems have this built-in (e.g., \r
+ class="osname">Linux>, NetBSD>,\r
+ Solaris>), for other systems you\r
+ can download an add-on package from here:
\r
+ url="http://developer.postgresql.org/~petere/bsd-gettext/" >.\r
+ If you are using the
Gettext> implementation in\r
+ the
GNU C library then you will additionally
\r
+ need the
GNU Gettext package for some
\r
+ utility programs. For any of the other implementations you will\r
+ not need it.\r
+ \r
+ \r
+\r
+ \r
+
Kerberos>, OpenSSL>, or PAM>,\r
+ if you want to support authentication using these services.\r
+ \r
+ \r
+ \r
+ \r
+\r
+ If you are building from a
CVS tree instead of
\r
+ using a released source package, or if you want to do development,\r
+ you also need the following packages:\r
+\r
+ \r
+ \r
+ \r
+ \r
+ \r
+ \r
+ \r
+ \r
+\r
+ are needed to build a CVS checkout or if you changed the actual\r
+ scanner and parser definition files. If you need them, be sure\r
+ to get
Flex> 2.5.4 or later and\r
+
Bison> 1.875 or later. Other yacc>\r
+ programs can sometimes be used, but doing so requires extra\r
+ effort and is not recommended. Other
lex>\r
+ programs will definitely not work.\r
+ \r
+ \r
+ \r
+ \r
+\r
+ If you need to get a
GNU package, you can find
\r
+ it at your local
GNU mirror site (see
\r
+ url="http://www.gnu.org/order/ftp.html">>\r
+ url="ftp://ftp.gnu.org/gnu/">.\r
+ \r
+\r
+ Also check that you have sufficient disk space. You will need about\r
+ 65 MB for the source tree during compilation and about 15 MB for\r
+ the installation directory. An empty database cluster takes about\r
+ 25 MB, databases take about five times the amount of space that a\r
+ flat text file with the same data would take. If you are going to\r
+ run the regression tests you will temporarily need up to an extra\r
+ 90 MB. Use the df command to check for disk\r
+ space.\r
+ \r
+ \r
+\r
+\r
+ \r
+
Getting The Source\r
+\r
+ The
PostgreSQL> &version; sources can be obtained by\r
+ url="ftp://ftp.postgresql.org/pub/source/v&version;/postgresql-&version;.tar.gz">.\r
+ Use a mirror if possible. After you have obtained the file, unpack it:\r
+\r
+gunzip postgresql-&version;.tar.gz\r
+tar xf postgresql-&version;.tar\r
+\r
+ This will create a directory\r
+ postgresql-&version; under the current directory\r
+ with the
PostgreSQL> sources.\r
+ Change into that directory for the rest\r
+ of the installation procedure.\r
+ \r
+ \r
+]]>\r
+\r
+ \r
+
If You Are Upgrading\r
+\r
+ \r
+ \r
+\r
+ The internal data storage format changes with new releases of\r
+
PostgreSQL>. Therefore, if you are upgrading an\r
+ existing installation that does not have a version number\r
+ &majorversion;.x
, you must back up and restore your\r
+ data as shown here. These instructions assume that your existing\r
+ installation is under the /usr/local/pgsql> directory,\r
+ and that the data area is in /usr/local/pgsql/data>.\r
+ Substitute your paths appropriately.\r
+ \r
+\r
+ \r
+ Make sure that your database is not updated during or after the\r
+ backup. This does not affect the integrity of the backup, but the\r
+ changed data would of course not be included. If necessary, edit\r
+ the permissions in the file\r
+ /usr/local/pgsql/data/pg_hba.conf> (or equivalent) to\r
+ disallow access from everyone except you.\r
+ \r
+ \r
+\r
+ \r
+ \r
+ use during upgrade\r
+ \r
+\r
+ To back up your database installation, type:\r
+\r
+pg_dumpall > outputfile>\r
+\r
+ If you need to preserve OIDs (such as when using them as\r
+ foreign keys), then use the option when running\r
+ \r
+\r
+ save large objects. Check\r
+ \r
+ ]]>\r
+ if you need to do this.\r
+ \r
+\r
+ To make the backup, you can use the
pg_dumpall\r
+ command from the version you are currently running. For best\r
+ results, however, try to use the
pg_dumpall\r
+ command from
PostgreSQL &version;,
\r
+ since this version contains bug fixes and improvements over older\r
+ versions. While this advice might seem idiosyncratic since you\r
+ haven't installed the new version yet, it is advisable to follow\r
+ it if you plan to install the new version in parallel with the\r
+ old version. In that case you can complete the installation\r
+ normally and transfer the data later. This will also decrease\r
+ the downtime.\r
+ \r
+ \r
+\r
+ \r
+ If you are installing the new version at the same location as the\r
+ old one then shut down the old server, at the latest before you\r
+ install the new files:\r
+\r
+kill -INT `cat /usr/local/pgsql/data/postmaster.pid | sed 1q`>\r
+\r
+ Versions prior to 7.0 do not have this\r
+ postmaster.pid> file. If you are using such a version\r
+ you must find out the process ID of the server yourself, for\r
+ example by typing ps ax | grep postmaster>, and\r
+ supply it to the kill> command.\r
+ \r
+\r
+ On systems that have
PostgreSQL> started at boot time, there is\r
+ probably a start-up file that will accomplish the same thing. For\r
+ example, on a Red Hat Linux> system one might find that\r
+\r
+/etc/rc.d/init.d/postgresql stop\r
+\r
+ works. Another possibility is pg_ctl stop>.\r
+ \r
+ \r
+\r
+ \r
+ If you are installing in the same place as the old version then\r
+ it is also a good idea to move the old installation out of the\r
+ way, in case you have trouble and need to revert to it.\r
+ Use a command like this:\r
+\r
+mv /usr/local/pgsql /usr/local/pgsql.old>\r
+\r
+ \r
+ \r
+ \r
+\r
+ After you have installed
PostgreSQL> &version;, create a new database\r
+ directory and start the new server. Remember that you must execute\r
+ these commands while logged in to the special database user account\r
+ (which you already have if you are upgrading).\r
+/usr/local/pgsql/bin/initdb -D /usr/local/pgsql/data>\r
+/usr/local/pgsql/bin/postmaster -D /usr/local/pgsql/data>\r
+\r
+ Finally, restore your data with\r
+\r
+/usr/local/pgsql/bin/psql -d template1 -f outputfile>\r
+\r
+ using the
new> psql>.\r
+ \r
+\r
+ These topics are discussed at length in \r
+ documentation,]]> \r
+ linkend="migration">,]]> which you are encouraged to read in any\r
+ case.\r
+ \r
+ \r
+\r
+\r
+ \r
+
Installation Procedure\r
+\r
+\r
+ \r
+
Configuration>\r
+\r
+ \r
+ \r
+\r
+ The first step of the installation procedure is to configure the\r
+ source tree for your system and choose the options you would like.\r
+ This is done by running the configure> script. For a\r
+ default installation simply enter\r
+\r
+./configure\r
+\r
+ This script will run a number of tests to guess values for various\r
+ system dependent variables and detect some quirks of your\r
+ operating system, and finally will create several files in the\r
+ build tree to record what it found. (You can also run\r
+ configure in a directory outside the source\r
+ tree if you want to keep the build directory separate.)\r
+ \r
+\r
+ The default configuration will build the server and utilities, as\r
+ well as all client applications and interfaces that require only a\r
+ C compiler. All files will be installed under\r
+ /usr/local/pgsql> by default.\r
+ \r
+\r
+ You can customize the build and installation process by supplying one\r
+ or more of the following command line options to\r
+ configure:\r
+\r
+ \r
+ \r
+ \r
+ \r
+ Install all files under the directory PREFIX>\r
+ instead of /usr/local/pgsql. The actual\r
+ files will be installed into various subdirectories; no files\r
+ will ever be installed directly into the\r
+ PREFIX> directory.\r
+ \r
+\r
+ If you have special needs, you can also customize the\r
+ individual subdirectories with the following options. However,\r
+ if you leave these with their defaults, the installation will be\r
+ relocatable, meaning you can move the directory after\r
+ installation. (The man> and doc>\r
+ locations are not affected by this.)\r
+ \r
+\r
+ For relocatable installs, you might want to use \r
+ configure's --disable-rpath>\r
+ option. Also, you will need to tell the operating system how\r
+ to find the shared libraries.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ You can install architecture-dependent files under a\r
+ different prefix, EXEC-PREFIX>, than what\r
+ PREFIX> was set to. This can be useful to\r
+ share architecture-independent files between hosts. If you\r
+ omit this, then EXEC-PREFIX> is set equal to\r
+ PREFIX> and both architecture-dependent and\r
+ independent files will be installed under the same tree,\r
+ which is probably what you want.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Specifies the directory for executable programs. The default\r
+ is EXEC-PREFIX>/bin>, which\r
+ normally means /usr/local/pgsql/bin>.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Sets the directory for read-only data files used by the\r
+ installed programs. The default is\r
+ PREFIX>/share>. Note that this has\r
+ nothing to do with where your database files will be placed.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ The directory for various configuration files,\r
+ PREFIX>/etc> by default.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ The location to install libraries and dynamically loadable\r
+ modules. The default is\r
+ EXEC-PREFIX>/lib>.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ The directory for installing C and C++ header files. The\r
+ default is PREFIX>/include>.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ The man pages that come with
PostgreSQL> will be installed under\r
+ this directory, in their respective\r
+ manx>> subdirectories.\r
+ The default is PREFIX>/man>.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ \r
+ Documentation files, except man> pages, will be\r
+ installed into this directory. The default is\r
+ PREFIX>/doc>. If the option\r
+ is specified, the\r
+ documentation will not be installed by make\r
+ install. This is intended for packaging scripts\r
+ that have special methods for installing documentation.\r
+ \r
+ \r
+ \r
+ \r
+\r
+ \r
+ Care has been taken to make it possible to install\r
+
PostgreSQL> into shared installation locations\r
+ (such as /usr/local/include) without\r
+ interfering with the namespace of the rest of the system. First,\r
+ the string /postgresql
is\r
+ automatically appended to datadir,\r
+ sysconfdir, and docdir,\r
+ unless the fully expanded directory name already contains the\r
+ string postgres>
or\r
+ pgsql>
. For example, if you choose\r
+ /usr/local as prefix, the documentation will\r
+ be installed in /usr/local/doc/postgresql,\r
+ but if the prefix is /opt/postgres, then it\r
+ will be in /opt/postgres/doc. The public C\r
+ header files of the client interfaces are installed into\r
+ includedir and are namespace-clean. The\r
+ internal header files and the server header files are installed\r
+ into private directories under includedir. See\r
+ the documentation of each interface for information about how to\r
+ get at the its header files. Finally, a private subdirectory will\r
+ also be created, if appropriate, under libdir\r
+ for dynamically loadable modules.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ \r
+ DIRECTORIES> is a colon-separated list of\r
+ directories that will be added to the list the compiler\r
+ searches for header files. If you have optional packages\r
+ (such as GNU
Readline>) installed in a non-standard\r
+ location,\r
+ you have to use this option and probably also the corresponding\r
+
+ \r
+ Example: --with-includes=/opt/gnu/include:/usr/sup/include>.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ DIRECTORIES> is a colon-separated list of\r
+ directories to search for libraries. You will probably have\r
+ to use this option (and the corresponding\r
+
+ installed in non-standard locations.\r
+ \r
+ Example: --with-libraries=/opt/gnu/lib:/usr/sup/lib>.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Enables Native Language Support (
NLS),
\r
+ that is, the ability to display a program's messages in a\r
+ language other than English.\r
+ LANGUAGES is a space separated\r
+ list of codes of the languages that you want supported, for\r
+ example --enable-nls='de fr'>. (The intersection\r
+ between your list and the set of actually provided\r
+ translations will be computed automatically.) If you do not\r
+ specify a list, then all available translations are\r
+ installed.\r
+ \r
+\r
+ To use this option, you will need an implementation of the\r
+
Gettext> API; see above.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Set NUMBER> as the default port number for\r
+ server and clients. The default is 5432. The port can always\r
+ be changed later on, but if you specify it here then both\r
+ server and clients will have the same default compiled in,\r
+ which can be very convenient. Usually the only good reason\r
+ to select a non-default value is if you intend to run multiple\r
+
PostgreSQL> servers on the same machine.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Build the
PL/Perl> server-side language.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Build the
PL/Python> server-side language.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Build the
PL/Tcl> server-side language.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Tcl installs the file tclConfig.sh, which\r
+ contains configuration information needed to build modules\r
+ interfacing to Tcl. This file is normally found automatically\r
+ at a well-known location, but if you want to use a different\r
+ version of Tcl you can specify the directory in which to look\r
+ for it.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ \r
+ Build with support for Kerberos authentication. You can use\r
+ either Kerberos version 4 or 5, but not both. On many\r
+ systems, the Kerberos system is not installed in a location\r
+ that is searched by default (e.g., /usr/include>,\r
+ /usr/lib>), so you must use the options\r
+
+ addition to this option. configure> will check\r
+ for the required header files and libraries to make sure that\r
+ your Kerberos installation is sufficient before proceeding.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ The name of the Kerberos service principal.\r
+ postgres is the default. There's probably no\r
+ reason to change this.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ SSL\r
+ \r
+\r
+ \r
+ \r
+ Build with support for
SSL> (encrypted)\r
+ connections. This requires the
OpenSSL>\r
+ package to be installed. configure> will check\r
+ for the required header files and libraries to make sure that\r
+ your
OpenSSL> installation is sufficient\r
+ before proceeding.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ (Pluggable Authentication Modules) support.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Prevents the use of the
Readline> library. This disables\r
+ command-line editing and history in\r
+
psql, so it is not recommended.
\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Build with Rendezvous support.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Allow the build to succeed even if
PostgreSQL>\r
+ has no CPU spinlock support for the platform. The lack of\r
+ spinlock support will result in poor performance; therefore,\r
+ this option should only be used if the build aborts and\r
+ informs you that the platform lacks spinlock support. If this\r
+ option is required to build
PostgreSQL> on\r
+ your platform, please report the problem to the\r
+
PostgreSQL> developers.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Make the client libraries thread-safe. This allows\r
+ concurrent threads in
libpq and
\r
+
ECPG programs to safely control
\r
+ their private connection handles.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Prevents the use of the
Zlib> library. This disables\r
+ compression support in
pg_dump.
\r
+ This option is only intended for those rare systems where this\r
+ library is not available.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Compiles all programs and libraries with debugging symbols.\r
+ This means that you can run the programs through a debugger\r
+ to analyze problems. This enlarges the size of the installed\r
+ executables considerably, and on non-GCC compilers it usually\r
+ also disables compiler optimization, causing slowdowns. However,\r
+ having the symbols available is extremely helpful for dealing\r
+ with any problems that may arise. Currently, this option is\r
+ recommended for production installations only if you use GCC.\r
+ But you should always have it on if you are doing development work\r
+ or running a beta version.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Enables assertion> checks in the server, which test for\r
+ many can't happen> conditions. This is invaluable for\r
+ code development purposes, but the tests slow things down a little.\r
+ Also, having the tests turned on won't necessarily enhance the\r
+ stability of your server! The assertion checks are not categorized\r
+ for severity, and so what might be a relatively harmless bug will\r
+ still lead to server restarts if it triggers an assertion\r
+ failure. Currently, this option is not recommended for\r
+ production use, but you should have it on for development work\r
+ or when running a beta version.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+ \r
+ Enables automatic dependency tracking. With this option, the\r
+ makefiles are set up so that all affected object files will\r
+ be rebuilt when any header file is changed. This is useful\r
+ if you are doing development work, but is just wasted overhead\r
+ if you intend only to compile once and install. At present,\r
+ this option will work only if you use GCC.\r
+ \r
+ \r
+ \r
+\r
+ \r
+ \r
+\r
+ If you prefer a C compiler different from the one\r
+ configure picks then you can set the\r
+ environment variable CC> to the program of your choice.\r
+ By default, configure will pick\r
+ gcc unless this is inappropriate for the\r
+ platform. Similarly, you can override the default compiler flags\r
+ with the CFLAGS variable.\r
+ \r
+\r
+ You can specify environment variables on the\r
+ configure command line, for example:\r
+\r
+./configure CC=/opt/bin/gcc CFLAGS='-O2 -pipe'>\r
+\r
+ \r
+ \r
+\r
+ \r
+
Build\r
+\r
+ To start the build, type\r
+\r
+gmake\r
+\r
+ (Remember to use
GNU> make>.) The build\r
+ may take anywhere from 5 minutes to half an hour depending on your\r
+ hardware. The last line displayed should be\r
+\r
+All of PostgreSQL is successfully made. Ready to install.\r
+\r
+ \r
+ \r
+\r
+ \r
+
Regression Tests\r
+\r
+ \r
+ \r
+\r
+ If you want to test the newly built server before you install it,\r
+ you can run the regression tests at this point. The regression\r
+ tests are a test suite to verify that
PostgreSQL>\r
+ runs on your machine in the way the developers expected it\r
+ to. Type\r
+\r
+gmake check\r
+\r
+ (This won't work as root; do it as an unprivileged user.)\r
+ \r
+ src/test/regress/README> and the\r
+ documentation contain]]>\r
+ contains]]>\r
+ detailed information about interpreting the test results. You can\r
+ repeat this test at any later time by issuing the same command.\r
+ \r
+ \r
+\r
+ \r
+
Installing The Files\r
+\r
+ \r
+ If you are upgrading an existing system and are going to install\r
+ the new files over the old ones, then you should have backed up\r
+ your data and shut down the old server by now, as explained in\r
+ above.\r
+ \r
+ \r
+\r
+ To install
PostgreSQL> enter\r
+\r
+gmake install\r
+\r
+ This will install files into the directories that were specified\r
+ in . Make sure that you have appropriate\r
+ permissions to write into that area. Normally you need to do this\r
+ step as root. Alternatively, you could create the target\r
+ directories in advance and arrange for appropriate permissions to\r
+ be granted.\r
+ \r
+\r
+ You can use gmake install-strip instead of\r
+ gmake install to strip the executable files and\r
+ libraries as they are installed. This will save some space. If\r
+ you built with debugging support, stripping will effectively\r
+ remove the debugging support, so it should only be done if\r
+ debugging is no longer needed. install-strip\r
+ tries to do a reasonable job saving space, but it does not have\r
+ perfect knowledge of how to strip every unneeded byte from an\r
+ executable file, so if you want to save all the disk space you\r
+ possibly can, you will have to do manual work.\r
+ \r
+\r
+ The standard installation provides all the header files needed for client\r
+ application development as well as for any server-side program\r
+ development (such as custom functions or data types written in C).\r
+ \r
+\r
+ \r
+
Client-only installation:\r
+ If you want to install only the client applications and\r
+ interface libraries, then you can use these commands:\r
+\r
+gmake -C src/bin install>\r
+gmake -C src/include install>\r
+gmake -C src/interfaces install>\r
+gmake -C doc install>\r
+\r
+ \r
+ \r
+ \r
+ \r
+\r
+ \r
+
Registering eventlog> on \r
+ class="osname">Windows>:\r
+ To register a
Windows> eventlog>\r
+ library with the operating system, issue this command after installation:\r
+\r
+regsvr32 pgsql_library_directory>/pgevent.dll>\r
+\r
+ This creates registry entries used by the event viewer.\r
+ \r
+ \r
+\r
+ \r
+
Uninstallation:\r
+ To undo the installation use the command gmake\r
+ uninstall>. However, this will not remove any created directories.\r
+ \r
+ \r
+\r
+ \r
+
Cleaning:\r
+\r
+ After the installation you can make room by removing the built\r
+ files from the source tree with the command gmake\r
+ clean>. This will preserve the files made by the configure\r
+ program, so that you can rebuild everything with gmake>\r
+ later on. To reset the source tree to the state in which it was\r
+ distributed, use gmake distclean>. If you are going to\r
+ build for several platforms from the same source tree you must do\r
+ this and re-configure for each build.\r
+ \r
+ \r
+\r
+ If you perform a build and then discover that your configure>\r
+ options were wrong, or if you change anything that configure>\r
+ investigates (for example, software upgrades), then it's a good\r
+ idea to do gmake distclean> before reconfiguring and\r
+ rebuilding. Without this, your changes in configuration choices\r
+ may not propagate everywhere they need to.\r
+ \r
+ \r
+\r
+ \r
+
Post-Installation Setup\r
+\r
+ \r
+
Shared Libraries\r
+\r
+ \r
+ \r
+\r
+ On some systems that have shared libraries (which most systems do)\r
+ you need to tell your system how to find the newly installed\r
+ shared libraries. The systems on which this is\r
+ not necessary include \r
+ class="osname">BSD/OS>, FreeBSD>,\r
+ HP-UX>, \r
+ class="osname">IRIX>, Linux>,\r
+ NetBSD>, \r
+ class="osname">OpenBSD>, Tru64\r
+ UNIX> (formerly Digital UNIX>), and\r
+ Solaris>.\r
+ \r
+\r
+ The method to set the shared library search path varies between\r
+ platforms, but the most widely usable method is to set the\r
+ environment variable LD_LIBRARY_PATH> like so: In Bourne\r
+ shells (sh>, ksh>, bash>, zsh>)\r
+LD_LIBRARY_PATH=/usr/local/pgsql/lib\r
+export LD_LIBRARY_PATH\r
+\r
+ or in csh> or tcsh>\r
+setenv LD_LIBRARY_PATH /usr/local/pgsql/lib\r
+\r
+ Replace /usr/local/pgsql/lib> with whatever you set\r
+
+ You should put these commands into a shell start-up file such as\r
+ /etc/profile> or ~/.bash_profile>. Some\r
+ good information about the caveats associated with this method can\r
+ url="http://www.visi.com/~barr/ldpath.html">.\r
+ \r
+\r
+ On some systems it might be preferable to set the environment\r
+ variable LD_RUN_PATH before\r
+ building.\r
+ \r
+\r
+ On Cygwin, put the library\r
+ directory in the PATH or move the\r
+ .dll files into the bin\r
+ directory.\r
+ \r
+\r
+ If in doubt, refer to the manual pages of your system (perhaps\r
+ ld.so or rld). If you later\r
+ on get a message like\r
+\r
+psql: error in loading shared libraries\r
+libpq.so.2.1: cannot open shared object file: No such file or directory\r
+\r
+ then this step was necessary. Simply take care of it then.\r
+ \r
+\r
+ \r
+ \r
+ If you are on BSD/OS>, \r
+ class="osname">Linux>, or SunOS 4>\r
+ and you have root access you can run\r
+/sbin/ldconfig /usr/local/pgsql/lib\r
+\r
+ (or equivalent directory) after installation to enable the\r
+ run-time linker to find the shared libraries faster. Refer to the\r
+ manual page of ldconfig> for more information. On\r
+ FreeBSD>, \r
+ class="osname">NetBSD>, and \r
+ class="osname">OpenBSD> the command is\r
+/sbin/ldconfig -m /usr/local/pgsql/lib\r
+\r
+ instead. Other systems are not known to have an equivalent\r
+ command.\r
+ \r
+ \r
+\r
+ \r
+
Environment Variables\r
+\r
+ \r
+ \r
+\r
+ If you installed into /usr/local/pgsql> or some other\r
+ location that is not searched for programs by default, you should\r
+ add /usr/local/pgsql/bin> (or whatever you set\r
+
+ into your PATH>. Strictly speaking, this is not\r
+ necessary, but it will make the use of
PostgreSQL>\r
+ much more convenient.\r
+ \r
+\r
+ To do this, add the following to your shell start-up file, such as\r
+ ~/.bash_profile> (or /etc/profile>, if you\r
+ want it to affect every user):\r
+PATH=/usr/local/pgsql/bin:$PATH\r
+export PATH\r
+\r
+ If you are using csh> or tcsh>, then use this command:\r
+set path = ( /usr/local/pgsql/bin $path )\r
+\r
+ \r
+\r
+ \r
+ \r
+ To enable your system to find the
man>\r
+ documentation, you need to add lines like the following to a\r
+ shell start-up file unless you installed into a location that is\r
+ searched by default.\r
+MANPATH=/usr/local/pgsql/man:$MANPATH\r
+export MANPATH\r
+\r
+ \r
+\r
+ The environment variables PGHOST> and PGPORT>\r
+ specify to client applications the host and port of the database\r
+ server, overriding the compiled-in defaults. If you are going to\r
+ run client applications remotely then it is convenient if every\r
+ user that plans to use the database sets PGHOST>. This\r
+ is not required, however: the settings can be communicated via command\r
+ line options to most client programs.\r
+ \r
+ \r
+ \r
+\r
+\r
+\r
+ \r
+
Getting Started\r
+\r
+ The following is a quick summary of how to get
PostgreSQL> up and\r
+ running once installed. The main documentation contains more information.\r
+ \r
+\r
+ \r
+ Create a user account for the
PostgreSQL>\r
+ server. This is the user the server will run as. For production\r
+ use you should create a separate, unprivileged account\r
+ (postgres> is commonly used). If you do not have root\r
+ access or just want to play around, your own user account is\r
+ enough, but running the server as root is a security risk and\r
+ will not work.\r
+\r
+adduser postgres>\r
+\r
+ \r
+ \r
+\r
+ \r
+ Create a database installation with the initdb>\r
+ command. To run initdb> you must be logged in to your\r
+
PostgreSQL> server account. It will not work as\r
+ root.\r
+\r
+root# mkdir /usr/local/pgsql/data>\r
+root# chown postgres /usr/local/pgsql/data>\r
+root# su - postgres>\r
+postgres$ /usr/local/pgsql/bin/initdb -D /usr/local/pgsql/data>\r
+\r
+ \r
+\r
+ The
+ will be stored. You can use any path you want, it does not have\r
+ to be under the installation directory. Just make sure that the\r
+ server account can write to the directory (or create it, if it\r
+ doesn't already exist) before starting initdb>, as\r
+ illustrated here.\r
+ \r
+ \r
+\r
+ \r
+ The previous step should have told you how to start up the\r
+ database server. Do so now. The command should look something\r
+ like\r
+/usr/local/pgsql/bin/postmaster -D /usr/local/pgsql/data\r
+\r
+ This will start the server in the foreground. To put the server\r
+ in the background use something like\r
+nohup /usr/local/pgsql/bin/postmaster -D /usr/local/pgsql/data \\r
+ </dev/null >>server.log 2>&1 </dev/null &\r
+\r
+ \r
+\r
+ To stop a server running in the background you can type\r
+kill `cat /usr/local/pgsql/data/postmaster.pid`\r
+\r
+ \r
+\r
+ In order to allow TCP/IP connections (rather than only Unix\r
+ domain socket ones) you need to pass the
+ postmaster>.\r
+ \r
+ \r
+\r
+ \r
+ Create a database:\r
+\r
+createdb testdb>\r
+\r
+ Then enter\r
+\r
+psql testdb>\r
+\r
+ to connect to that database. At the prompt you can enter SQL\r
+ commands and start experimenting.\r
+ \r
+ \r
+ \r
+ \r
+\r
+ \r
+
What Now?\r
+\r
+ \r
+ \r
+ The
PostgreSQL> distribution contains a\r
+ comprehensive documentation set, which you should read sometime.\r
+ After installation, the documentation can be accessed by\r
+ pointing your browser to\r
+ /usr/local/pgsql/doc/html/index.html>, unless you\r
+ changed the installation directories.\r
+ \r
+\r
+ The first few chapters of the main documentation are the Tutorial,\r
+ which should be your first reading if you are completely new to\r
+
SQL> databases. If you are familiar with database\r
+ concepts then you want to proceed with part on server\r
+ administration, which contains information about how to set up\r
+ the database server, database users, and authentication.\r
+ \r
+ \r
+\r
+ \r
+ Usually, you will want to modify your computer so that it will\r
+ automatically start the database server whenever it boots. Some\r
+ suggestions for this are in the documentation.\r
+ \r
+ \r
+\r
+ \r
+ Run the regression tests against the installed server (using\r
+ gmake installcheck). If you didn't run the\r
+ tests before installation, you should definitely do it now. This\r
+ is also explained in the documentation.\r
+ \r
+ \r
+\r
+ \r
+ By default,
PostgreSQL> is configured to run on\r
+ minimal hardware. This allows it to start up with almost any\r
+ hardware configuration. The default configuration is, however,\r
+ not designed for optimum performance. To achieve optimum\r
+ performance, several server parameters must be adjusted, the two\r
+ most common being shared_buffers and\r
+ work_mem.\r
+ Other parameters mentioned in the documentation also affect\r
+ performance.\r
+ \r
+ \r
+ \r
+ \r
+ \r
+]]>\r
+\r
+\r
+ \r
+
Supported Platforms\r
+\r
+
PostgreSQL> has been verified by the developer\r
+ community to work on the platforms listed below. A supported\r
+ platform generally means that
PostgreSQL> builds and\r
+ installs according to these instructions and that the regression\r
+ tests pass. Build farm
entries refer to builds\r
+ reported by the
PostgreSQL\r
+ Build Farm.\r
+ \r
+\r
+ \r
+ If you are having problems with the installation on a supported\r
+ listed here.\r
+ \r
+ \r
+\r
+ \r
+ \r
+
\r
+ \r
+ Processor\r
+ Version\r
+ Reported\r
+ Remarks\r
+ \r
+
\r
+
\r
+ \r
+ AIX>\r
+ PowerPC>\r
+ 8.0.0\r
+ see also doc/FAQ_AIX\r
+ \r
+ \r
+ AIX>\r
+ RS6000>\r
+ 8.0.0\r
+ see also doc/FAQ_AIX\r
+ \r
+ \r
+ BSD/OS>\r
+ x86>\r
+ 8.0.0\r
+ 4.3.1\r
+ \r
+ \r
+ Debian GNU/Linux>\r
+ Alpha>\r
+ 7.4\r
+ \r
+ \r
+ \r
+ Debian GNU/Linux>\r
+ AMD64>\r
+ 8.0.0\r
+ Build farm panda, snapshot 2004-12-06 01:20:02\r
+ sid, kernel 2.6\r
+ \r
+ \r
+ Debian GNU/Linux>\r
+ arm41>\r
+ 7.4\r
+ \r
+ \r
+ \r
+ Debian GNU/Linux>\r
+ Itanium>\r
+ 7.4\r
+ \r
+ \r
+ \r
+ Debian GNU/Linux>\r
+ m68k>\r
+ 8.0.0\r
+ sid\r
+ \r
+ \r
+ Debian GNU/Linux>\r
+ MIPS>\r
+ 8.0.0\r
+ Build farm lionfish, snapshot 2004-12-06 11:00:08\r
+ 3.1 (sarge), kernel 2.4\r
+ \r
+ \r
+ Debian GNU/Linux>\r
+ PA-RISC>\r
+ 8.0.0\r
+ \r
+ \r
+ \r
+ Debian GNU/Linux>\r
+ PowerPC>\r
+ 8.0.0\r
+ \r
+ \r
+ \r
+ Debian GNU/Linux>\r
+ S/390>\r
+ 7.4\r
+ \r
+ \r
+ \r
+ Debian GNU/Linux>\r
+ Sparc>\r
+ 8.0.0\r
+ sid, 32-bit\r
+ \r
+ \r
+ Debian GNU/Linux>\r
+ x86>\r
+ 8.0.0\r
+ 3.1 (sarge), kernel 2.6\r
+ \r
+ \r
+ Fedora>\r
+ AMD64>\r
+ 8.0.0\r
+ FC3\r
+ \r
+ \r
+ Fedora>\r
+ x86>\r
+ 8.0.0\r
+ Build farm dog, snapshot 2004-12-06 02:06:01\r
+ FC1\r
+ \r
+ \r
+ FreeBSD>\r
+ Alpha>\r
+ 7.4\r
+ 4.8\r
+ \r
+ \r
+ FreeBSD>\r
+ x86>\r
+ 8.0.0\r
+ Build farm cockatoo, snapshot 2004-12-06 14:10:01 (4.10);\r
+ \r
+ \r
+ \r
+ Gentoo Linux>\r
+ x86>\r
+ 8.0.0\r
+ \r
+ \r
+ \r
+ HP-UX>\r
+ PA-RISC>\r
+ 7.4\r
+ \r
+ \r
+ gcc> and cc>; see also doc/FAQ_HPUX\r
+ \r
+ \r
+ IRIX>\r
+ MIPS>\r
+ 7.4\r
+ 6.5.20, cc only\r
+ \r
+ \r
+ Mac OS X>\r
+ PowerPC>\r
+ 8.0.0\r
+ 10.3.5\r
+ \r
+ \r
+ Mandrakelinux>\r
+ x86>\r
+ 8.0.0\r
+ Build farm shrew, snapshot 2004-12-06 02:02:01\r
+ 10.0\r
+ \r
+ \r
+ NetBSD>\r
+ arm32>\r
+ 7.4\r
+ 1.6ZE/acorn32\r
+ \r
+ \r
+ NetBSD>\r
+ Sparc>\r
+ 7.4.1\r
+ 1.6.1, 32-bit\r
+ \r
+ \r
+ NetBSD>\r
+ x86>\r
+ 8.0.0\r
+ Build farm canary, snapshot 2004-12-06 03:30:00\r
+ 1.6\r
+ \r
+ \r
+ OpenBSD>\r
+ Sparc>\r
+ 7.4\r
+ 3.4\r
+ \r
+ \r
+ OpenBSD>\r
+ x86>\r
+ 8.0.0\r
+ Build farm emu, snapshot 2004-12-06 11:35:03\r
+ 3.6\r
+ \r
+ \r
+ Red Hat Linux>\r
+ AMD64>\r
+ 8.0.0\r
+ RHEL 3AS\r
+ \r
+ \r
+ Red Hat Linux>\r
+ IA64>\r
+ 8.0.0\r
+ RHEL 3AS\r
+ \r
+ \r
+ Red Hat Linux>\r
+ PowerPC>\r
+ 8.0.0\r
+ RHEL 3AS\r
+ \r
+ \r
+ Red Hat Linux>\r
+ PowerPC 64>\r
+ 8.0.0\r
+ RHEL 3AS\r
+ \r
+ \r
+ Red Hat Linux>\r
+ S/390>\r
+ 8.0.0\r
+ RHEL 3AS\r
+ \r
+ \r
+ Red Hat Linux>\r
+ S/390x>\r
+ 8.0.0\r
+ RHEL 3AS\r
+ \r
+ \r
+ Red Hat Linux>\r
+ x86>\r
+ 8.0.0\r
+ RHEL 3AS\r
+ \r
+ \r
+ Solaris>\r
+ Sparc>\r
+ 8.0.0\r
+ Solaris 8; see also doc/FAQ_Solaris\r
+ \r
+ \r
+ Solaris>\r
+ x86>\r
+ 8.0.0\r
+ Build farm kudu, snapshot 2004-12-10 02:30:04 (cc);\r
+ dragonfly, snapshot 2004-12-09 04:30:00 (gcc)\r
+ Solaris 9; see also doc/FAQ_Solaris\r
+ \r
+ \r
+ Tru64 UNIX>\r
+ Alpha>\r
+ 7.4\r
+ \r
+ \r
+ \r
+ UnixWare>\r
+ x86>\r
+ 8.0.0\r
+ cc, 7.1.4; see also doc/FAQ_SCO\r
+ \r
+ \r
+ Windows>\r
+ x86>\r
+ 8.0.0\r
+ see doc/FAQ_MINGW\r
+ \r
+ \r
+ x86>\r
+ 8.0.0\r
+ Build farm gibbon, snapshot 2004-12-11 01:33:01\r
+ see doc/FAQ_CYGWIN\r
+ \r
+
\r
+ \r
+ \r
+\r
+ \r
+
Unsupported Platforms:\r
+ The following platforms are either known not to work, or they used\r
+ to work in a previous release and we did not receive explicit\r
+ confirmation of a successful test with version &majorversion; at\r
+ the time this list was compiled. We include these here to let you\r
+ know that these platforms could> be supported if given\r
+ some attention.\r
+ \r
+ \r
+\r
+ \r
+ \r
+
\r
+ \r
+ Processor\r
+ Version\r
+ Reported\r
+ Remarks\r
+ \r
+
\r
+\r
+
\r
+ \r
+ BeOS>\r
+ x86>\r
+ 7.2\r
+ 2001-11-29,\r
+ needs updates to semaphore code\r
+ \r
+ \r
+ Linux>\r
+ PlayStation 2>\r
+ 7.4\r
+ 2003-11-02,\r
+ \r
+ needs new config.guess,\r
+ \r
+ \r
+ \r
+ \r
+ NetBSD>\r
+ Alpha>\r
+ 7.2\r
+ 2001-11-20,\r
+ 1.5W\r
+ \r
+ \r
+ NetBSD>\r
+ MIPS>\r
+ 7.2.1\r
+ 2002-06-13,\r
+ 1.5.3\r
+ \r
+ \r
+ NetBSD>\r
+ PowerPC>\r
+ 7.2\r
+ 2001-11-28,\r
+ 1.5\r
+ \r
+ \r
+ NetBSD>\r
+ VAX>\r
+ 7.1\r
+ 2001-03-30,\r
+ 1.5\r
+ \r
+ \r
+ QNX 4 RTOS>\r
+ x86>\r
+ 7.2\r
+ 2001-12-10,\r
+ \r
+ needs updates to semaphore code;\r
+ see also doc/FAQ_QNX4\r
+ \r
+ \r
+ QNX RTOS v6>\r
+ x86>\r
+ 7.2\r
+ patches available in archives, but too late for 7.2\r
+ \r
+ \r
+ SCO OpenServer>\r
+ x86>\r
+ 7.3.1\r
+ 2002-12-11, \r
+ 5.0.4, gcc>; see also doc/FAQ_SCO\r
+ \r
+ \r
+ SunOS 4>\r
+ Sparc>\r
+ 7.2\r
+ \r
+ \r
+
\r
+ \r
+ \r
+ \r
+\r
+\r
+\r
+\r