Google Git
Sign in
chromium / external / github.com / python / cpython / refs/tags/v2.5.1 / . / README
blob: de09bb1b6082395bd667738b60b1de9cd91ecf9b [file] [log] [blame]
Anthony Baxter9515c962007-04-05 04:52:06[diff] [blame]1This is Python version 2.5.1
2============================
Guido van Rossum91447632000-04-11 17:11:09[diff] [blame]3
Anthony Baxtera7118d32007-01-06 04:50:44[diff] [blame]4Copyright (c) 2001, 2002, 2003, 2004, 2005, 2006, 2007 Python Software
5Foundation.
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]6All rights reserved.
7
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]8Copyright (c) 2000 BeOpen.com.
9All rights reserved.
10
Guido van Rossum23f7aed2001-04-12 20:53:31[diff] [blame]11Copyright (c) 1995-2001 Corporation for National Research Initiatives.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]12All rights reserved.
13
Guido van Rossum6ebd2992000-09-03 04:47:47[diff] [blame]14Copyright (c) 1991-1995 Stichting Mathematisch Centrum.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]15All rights reserved.
16
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]17
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]18License information
19-------------------
20
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]21See the file "LICENSE" for information on the history of this
22software, terms & conditions for usage, and a DISCLAIMER OF ALL
23WARRANTIES.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]24
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]25This Python distribution contains no GNU General Public Licensed
26(GPLed) code so it may be used in proprietary projects just like prior
27Python distributions. There are interfaces to some GNU code but these
28are entirely optional.
29
30All trademarks referenced herein are property of their respective
31holders.
Guido van Rossum79808261997-12-11 18:01:47[diff] [blame]32
33
Guido van Rossum8d7d4ed1996-07-30 21:41:07[diff] [blame]34What's new in this release?
35---------------------------
Guido van Rossum3ff96dd1996-07-30 18:05:04[diff] [blame]36
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]37See the file "Misc/NEWS".
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]38
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]39
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]40If you don't read instructions
41------------------------------
42
43Congratulations on getting this far. :-)
44
45To start building right away (on UNIX): type "./configure" in the
Guido van Rossum2dcec0c2002-02-04 01:59:23[diff] [blame]46current directory and when it finishes, type "make". This creates an
47executable "./python"; to install in /usr/local, first do "su root"
48and then "make install".
49
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]50The section `Build instructions' below is still recommended reading.
Guido van Rossum3ff96dd1996-07-30 18:05:04[diff] [blame]51
52
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]53What is Python anyway?
54----------------------
55
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]56Python is an interpreted, interactive object-oriented programming
57language suitable (amongst other uses) for distributed application
58development, scripting, numeric computing and system testing. Python
59is often compared to Tcl, Perl, Java, JavaScript, Visual Basic or
60Scheme. To find out more about what Python can do for you, point your
61browser to http://www.python.org/.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]62
63
Guido van Rossumf501b4e1996-10-25 14:32:48[diff] [blame]64How do I learn Python?
65----------------------
66
Fred Drake0e6444c1999-05-17 19:35:01[diff] [blame]67The official tutorial is still a good place to start; see
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]68http://docs.python.org/ for online and downloadable versions, as well
69as a list of other introductions, and reference documentation.
Guido van Rossumf501b4e1996-10-25 14:32:48[diff] [blame]70
Guido van Rossumd0a42e22000-03-31 20:16:45[diff] [blame]71There's a quickly growing set of books on Python. See
Michael W. Hudsonfad46402005-06-16 15:51:20[diff] [blame]72http://wiki.python.org/moin/PythonBooks for a list.
Guido van Rossum8d90f9d1997-05-22 20:13:25[diff] [blame]73
Guido van Rossumf501b4e1996-10-25 14:32:48[diff] [blame]74
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]75Documentation
76-------------
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]77
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]78All documentation is provided online in a variety of formats. In
79order of importance for new users: Tutorial, Library Reference,
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]80Language Reference, Extending & Embedding, and the Python/C API. The
81Library Reference is especially of immense value since much of
82Python's power is described there, including the built-in data types
83and functions!
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]84
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]85All documentation is also available online at the Python web site
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]86(http://docs.python.org/, see below). It is available online for
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]87occasional reference, or can be downloaded in many formats for faster
Fred Drake6caae142000-10-25 17:51:02[diff] [blame]88access. The documentation is available in HTML, PostScript, PDF, and
89LaTeX formats; the LaTeX version is primarily for documentation
90authors, translators, and people with special formatting requirements.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]91
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]92Unfortunately, new-style classes (new in Python 2.2) have not yet been
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]93integrated into Python's standard documentation. A collection of
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]94pointers to what has been written is at:
Barry Warsaw06557452001-12-14 20:47:12[diff] [blame]95
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]96 http://www.python.org/doc/newstyle.html
Barry Warsaw06557452001-12-14 20:47:12[diff] [blame]97
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]98
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]99Web sites
100---------
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]101
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]102New Python releases and related technologies are published at
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]103http://www.python.org/. Come visit us!
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]104
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]105There's also a Python community web site at
106http://starship.python.net/.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]107
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]108
109Newsgroups and Mailing Lists
110----------------------------
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]111
112Read comp.lang.python, a high-volume discussion newsgroup about
113Python, or comp.lang.python.announce, a low-volume moderated newsgroup
114for Python-related announcements. These are also accessible as
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]115mailing lists: see http://www.python.org/community/lists.html for an
116overview of these and many other Python-related mailing lists.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]117
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]118Archives are accessible via the Google Groups Usenet archive; see
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]119http://groups.google.com/. The mailing lists are also archived, see
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]120http://www.python.org/community/lists.html for details.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]121
122
123Bug reports
124-----------
125
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]126To report or search for bugs, please use the Python Bug
127Tracker at http://sourceforge.net/bugs/?group_id=5470.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]128
129
130Patches and contributions
131-------------------------
132
Fred Drake6caae142000-10-25 17:51:02[diff] [blame]133To submit a patch or other contribution, please use the Python Patch
134Manager at http://sourceforge.net/patch/?group_id=5470. Guidelines
135for patch submission may be found at http://www.python.org/patches/.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]136
137If you have a proposal to change Python, it's best to submit a Python
138Enhancement Proposal (PEP) first. All current PEPs, as well as
Walter Dörwald7dc65d62003-02-02 23:37:05[diff] [blame]139guidelines for submitting a new PEP, are listed at
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]140http://www.python.org/peps/.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]141
142
143Questions
144---------
145
146For help, if you can't find it in the manuals or on the web site, it's
147best to post to the comp.lang.python or the Python mailing list (see
148above). If you specifically don't want to involve the newsgroup or
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]149mailing list, send questions to help@python.org (a group of volunteers
150who answer questions as they can). The newsgroup is the most
151efficient way to ask public questions.
Guido van Rossum627b2d71993-12-24 10:39:16[diff] [blame]152
Guido van Rossum901454e2000-06-29 22:28:44[diff] [blame]153
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]154Build instructions
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]155==================
Guido van Rossum627b2d71993-12-24 10:39:16[diff] [blame]156
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]157Before you can build Python, you must first configure it.
158Fortunately, the configuration and build process has been automated
159for Unix and Linux installations, so all you usually have to do is
160type a few commands and sit back. There are some platforms where
161things are not quite as smooth; see the platform specific notes below.
162If you want to build for multiple platforms sharing the same source
163tree, see the section on VPATH below.
Guido van Rossum627b2d71993-12-24 10:39:16[diff] [blame]164
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]165Start by running the script "./configure", which determines your
166system configuration and creates the Makefile. (It takes a minute or
167two -- please be patient!) You may want to pass options to the
168configure script -- see the section below on configuration options and
169variables. When it's done, you are ready to run make.
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]170
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]171To build Python, you normally type "make" in the toplevel directory.
172If you have changed the configuration, the Makefile may have to be
173rebuilt. In this case you may have to run make again to correctly
174build your desired target. The interpreter executable is built in the
175top level directory.
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]176
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]177Once you have built a Python interpreter, see the subsections below on
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]178testing and installation. If you run into trouble, see the next
179section.
180
181Previous versions of Python used a manual configuration process that
182involved editing the file Modules/Setup. While this file still exists
183and manual configuration is still supported, it is rarely needed any
184more: almost all modules are automatically built as appropriate under
185guidance of the setup.py script, which is run by Make after the
186interpreter has been built.
Guido van Rossum627b2d71993-12-24 10:39:16[diff] [blame]187
Guido van Rossum0a516c91994-09-12 10:58:40[diff] [blame]188
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]189Troubleshooting
190---------------
Guido van Rossum627b2d71993-12-24 10:39:16[diff] [blame]191
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]192See also the platform specific notes in the next section.
193
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]194If you run into other trouble, see the FAQ
195(http://www.python.org/doc/faq) for hints on what can go wrong, and
196how to fix it.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]197
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]198If you rerun the configure script with different options, remove all
199object files by running "make clean" before rebuilding. Believe it or
200not, "make clean" sometimes helps to clean up other inexplicable
201problems as well. Try it before sending in a bug report!
202
Guido van Rossumd0fe8451996-08-26 03:02:37[diff] [blame]203If the configure script fails or doesn't seem to find things that
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]204should be there, inspect the config.log file.
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]205
206If you get a warning for every file about the -Olimit option being no
207longer supported, you can ignore it. There's no foolproof way to know
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]208whether this option is needed; all we can do is test whether it is
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]209accepted without error. On some systems, e.g. older SGI compilers, it
210is essential for performance (specifically when compiling ceval.c,
211which has more basic blocks than the default limit of 1000). If the
212warning bothers you, edit the Makefile to remove "-Olimit 1500" from
213the OPT variable.
Guido van Rossumd0fe8451996-08-26 03:02:37[diff] [blame]214
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]215If you get failures in test_long, or sys.maxint gets set to -1, you
216are probably experiencing compiler bugs, usually related to
Martin v. Löwis9cea5f52002-11-15 07:28:05[diff] [blame]217optimization. This is a common problem with some versions of gcc, and
218some vendor-supplied compilers, which can sometimes be worked around
219by turning off optimization. Consider switching to stable versions
Anthony Baxtere9d719b2004-11-30 01:49:18[diff] [blame]220(gcc 2.95.2, gcc 3.x, or contact your vendor.)
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]221
222From Python 2.0 onward, all Python C code is ANSI C. Compiling using
223old K&R-C-only compilers is no longer possible. ANSI C compilers are
224available for all modern systems, either in the form of updated
Martin v. Löwis9cea5f52002-11-15 07:28:05[diff] [blame]225compilers from the vendor, or one of the free compilers (gcc).
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]226
Georg Brandlfb4da1d2006-11-15 17:42:07[diff] [blame]227If "make install" fails mysteriously during the "compiling the library"
228step, make sure that you don't have any of the PYTHONPATH or PYTHONHOME
229environment variables set, as they may interfere with the newly built
230executable which is compiling the library.
231
Martin v. Löwis779ffc02002-12-02 22:17:01[diff] [blame]232Unsupported systems
233-------------------
234
Martin v. Löwis8cab8b02006-07-24 05:05:22[diff] [blame]235A number of features are not supported in Python 2.5 anymore. Some
236support code is still present, but will be removed in Python 2.6.
Martin v. Löwis779ffc02002-12-02 22:17:01[diff] [blame]237If you still need to use current Python versions on these systems,
238please send a message to python-dev@python.org indicating that you
Martin v. Löwis8cab8b02006-07-24 05:05:22[diff] [blame]239volunteer to support this system. For a more detailed discussion
240regarding no-longer-supported and resupporting platforms, as well
241as a list of platforms that became or will be unsupported, see PEP 11.
Martin v. Löwis779ffc02002-12-02 22:17:01[diff] [blame]242
243More specifically, the following systems are not supported any
244longer:
245- SunOS 4
246- DYNIX
247- dgux
248- Minix
Martin v. Löwis8cab8b02006-07-24 05:05:22[diff] [blame]249- NeXT
Martin v. Löwis779ffc02002-12-02 22:17:01[diff] [blame]250- Irix 4 and --with-sgi-dl
251- Linux 1
252- Systems defining __d6_pthread_create (configure.in)
253- Systems defining PY_PTHREAD_D4, PY_PTHREAD_D6,
254 or PY_PTHREAD_D7 in thread_pthread.h
255- Systems using --with-dl-dld
Martin v. Löwis7e4cfcb2002-12-19 16:21:49[diff] [blame]256- Systems using --without-universal-newlines
Martin v. Löwis8cab8b02006-07-24 05:05:22[diff] [blame]257- MacOS 9
258
259The following systems are still supported in Python 2.5, but
260support will be dropped in 2.6:
261- Systems using --with-wctype-functions
262- Win9x, WinME
263
264Warning on install in Windows 98 and Windows Me
265-----------------------------------------------
266
267Following Microsoft's closing of Extended Support for
268Windows 98/ME (July 11, 2006), Python 2.6 will stop
269supporting these platforms. Python development and
270maintainability becomes easier (and more reliable) when
271platform specific code targeting OSes with few users
272and no dedicated expert developers is taken out. The
273vendor also warns that the OS versions listed above
274"can expose customers to security risks" and recommends
275upgrade.
Martin v. Löwis779ffc02002-12-02 22:17:01[diff] [blame]276
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]277Platform specific notes
278-----------------------
279
Guido van Rossum0447a321995-10-08 01:22:33[diff] [blame]280(Some of these may no longer apply. If you find you can build Python
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]281on these platforms without the special directions mentioned here,
282submit a documentation bug report to SourceForge (see Bug Reports
283above) so we can remove them!)
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]284
Martin v. Löwiscf783eb2006-09-18 06:23:06[diff] [blame]285GCC 4.1,
Martin v. Löwis885d8e72006-09-18 06:00:52[diff] [blame]286GCC 4.2: There is a known incompatibility between Python and GCC,
Anthony Baxter347f3cc2006-09-18 06:45:59[diff] [blame]287 where GCC 4.1 and later uses an interpretation of C
288 different to earlier GCC releases in an area where the C
289 specification has undefined behaviour (namely, integer arithmetic
290 involving -sys.maxint-1).
291
Martin v. Löwiscf783eb2006-09-18 06:23:06[diff] [blame]292 As a consequence, compiling Python with GCC 4.1/4.2 is not
Anthony Baxter347f3cc2006-09-18 06:45:59[diff] [blame]293 recommended. It is likely that this problem will be resolved
Martin v. Löwiscf783eb2006-09-18 06:23:06[diff] [blame]294 in future Python releases. As a work-around, it seems that
Anthony Baxter347f3cc2006-09-18 06:45:59[diff] [blame]295 adding -fwrapv to the compiler options restores the earlier
Martin v. Löwiscf783eb2006-09-18 06:23:06[diff] [blame]296 GCC behaviour.
Martin v. Löwis885d8e72006-09-18 06:00:52[diff] [blame]297
Skip Montanaro6dc4a8e2003-05-02 15:27:59[diff] [blame]298Unix platforms: If your vendor still ships (and you still use) Berkeley DB
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]299 1.85 you will need to edit Modules/Setup to build the bsddb185
300 module and add a line to sitecustomize.py which makes it the
301 default. In Modules/Setup a line like
Skip Montanaro6dc4a8e2003-05-02 15:27:59[diff] [blame]302
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]303 bsddb185 bsddbmodule.c
Skip Montanaro6dc4a8e2003-05-02 15:27:59[diff] [blame]304
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]305 should work. (You may need to add -I, -L or -l flags to direct the
306 compiler and linker to your include files and libraries.)
Skip Montanaro6dc4a8e2003-05-02 15:27:59[diff] [blame]307
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]308XXX I think this next bit is out of date:
Skip Montanaro6dc4a8e2003-05-02 15:27:59[diff] [blame]309
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]31064-bit platforms: The modules audioop, imageop and rgbimg don't work.
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]311 The setup.py script disables them on 64-bit installations.
312 Don't try to enable them in the Modules/Setup file. They
313 contain code that is quite wordsize sensitive. (If you have a
314 fix, let us know!)
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]315
Guido van Rossum4462e931997-01-22 21:00:32[diff] [blame]316Solaris: When using Sun's C compiler with threads, at least on Solaris
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]317 2.5.1, you need to add the "-mt" compiler option (the simplest
318 way is probably to specify the compiler with this option as
319 the "CC" environment variable when running the configure
320 script).
Guido van Rossum4462e931997-01-22 21:00:32[diff] [blame]321
Guido van Rossum42992312002-09-23 14:02:18[diff] [blame]322 When using GCC on Solaris, beware of binutils 2.13 or GCC
323 versions built using it. This mistakenly enables the
324 -zcombreloc option which creates broken shared libraries on
325 Solaris. binutils 2.12 works, and the binutils maintainers
Skip Montanaro4de9cba2003-01-03 16:26:23[diff] [blame]326 are aware of the problem. Binutils 2.13.1 only partially
327 fixed things. It appears that 2.13.2 solves the problem
328 completely. This problem is known to occur with Solaris 2.7
329 and 2.8, but may also affect earlier and later versions of the
330 OS.
Guido van Rossum42992312002-09-23 14:02:18[diff] [blame]331
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]332 When the dynamic loader complains about errors finding shared
333 libraries, such as
Martin v. Löwis9cea5f52002-11-15 07:28:05[diff] [blame]334
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]335 ld.so.1: ./python: fatal: libstdc++.so.5: open failed:
336 No such file or directory
Martin v. Löwis9cea5f52002-11-15 07:28:05[diff] [blame]337
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]338 you need to first make sure that the library is available on
339 your system. Then, you need to instruct the dynamic loader how
340 to find it. You can choose any of the following strategies:
Martin v. Löwis9cea5f52002-11-15 07:28:05[diff] [blame]341
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]342 1. When compiling Python, set LD_RUN_PATH to the directories
343 containing missing libraries.
344 2. When running Python, set LD_LIBRARY_PATH to these directories.
345 3. Use crle(8) to extend the search path of the loader.
346 4. Modify the installed GCC specs file, adding -R options into the
347 *link: section.
Martin v. Löwis9cea5f52002-11-15 07:28:05[diff] [blame]348
Skip Montanaro9f6606a2005-11-17 18:19:39[diff] [blame]349 The complex object fails to compile on Solaris 10 with gcc 3.4 (at
350 least up to 3.4.3). To work around it, define Py_HUGE_VAL as
351 HUGE_VAL(), e.g.:
352
353 make CPPFLAGS='-D"Py_HUGE_VAL=HUGE_VAL()" -I. -I$(srcdir)/Include'
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]354 ./python setup.py CPPFLAGS='-D"Py_HUGE_VAL=HUGE_VAL()"'
Skip Montanaro9f6606a2005-11-17 18:19:39[diff] [blame]355
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]356Linux: A problem with threads and fork() was tracked down to a bug in
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]357 the pthreads code in glibc version 2.0.5; glibc version 2.0.7
358 solves the problem. This causes the popen2 test to fail;
359 problem and solution reported by Pablo Bleyer.
Guido van Rossum4462e931997-01-22 21:00:32[diff] [blame]360
Martin v. Löwis3efa04a2003-06-14 05:29:27[diff] [blame]361Red Hat Linux: Red Hat 9 built Python2.2 in UCS-4 mode and hacked
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]362 Tcl to support it. To compile Python2.3 with Tkinter, you will
363 need to pass --enable-unicode=ucs4 flag to ./configure.
Martin v. Löwis3efa04a2003-06-14 05:29:27[diff] [blame]364
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]365 There's an executable /usr/bin/python which is Python
366 1.5.2 on most older Red Hat installations; several key Red Hat tools
367 require this version. Python 2.1.x may be installed as
368 /usr/bin/python2. The Makefile installs Python as
369 /usr/local/bin/python, which may or may not take precedence
370 over /usr/bin/python, depending on how you have set up $PATH.
Guido van Rossum2dcec0c2002-02-04 01:59:23[diff] [blame]371
Guido van Rossum59230501999-06-21 20:51:46[diff] [blame]372FreeBSD 3.x and probably platforms with NCurses that use libmytinfo or
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]373 similar: When using cursesmodule, the linking is not done in
374 the correct order with the defaults. Remove "-ltermcap" from
375 the readline entry in Setup, and use as curses entry: "curses
376 cursesmodule.c -lmytinfo -lncurses -ltermcap" - "mytinfo" (so
377 called on FreeBSD) should be the name of the auxiliary library
378 required on your platform. Normally, it would be linked
379 automatically, but not necessarily in the correct order.
Guido van Rossum59230501999-06-21 20:51:46[diff] [blame]380
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]381BSDI: BSDI versions before 4.1 have known problems with threads,
382 which can cause strange errors in a number of modules (for
383 instance, the 'test_signal' test script will hang forever.)
384 Turning off threads (with --with-threads=no) or upgrading to
385 BSDI 4.1 solves this problem.
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]386
387DEC Unix: Run configure with --with-dec-threads, or with
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]388 --with-threads=no if no threads are desired (threads are on by
389 default). When using GCC, it is possible to get an internal
390 compiler error if optimization is used. This was reported for
391 GCC 2.7.2.3 on selectmodule.c. Manually compile the affected
392 file without optimization to solve the problem.
Guido van Rossum8eca2c21996-02-14 18:37:46[diff] [blame]393
Guido van Rossume1cdb161999-04-10 16:04:07[diff] [blame]394DEC Ultrix: compile with GCC to avoid bugs in the native compiler,
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]395 and pass SHELL=/bin/sh5 to Make when installing.
Guido van Rossume1cdb161999-04-10 16:04:07[diff] [blame]396
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]397AIX: A complete overhaul of the shared library support is now in
398 place. See Misc/AIX-NOTES for some notes on how it's done.
399 (The optimizer bug reported at this place in previous releases
400 has been worked around by a minimal code change.) If you get
401 errors about pthread_* functions, during compile or during
402 testing, try setting CC to a thread-safe (reentrant) compiler,
403 like "cc_r". For full C++ module support, set CC="xlC_r" (or
404 CC="xlC" without thread support).
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]405
Guido van Rossuma2534c82005-09-20 19:06:23[diff] [blame]406AIX 5.3: To build a 64-bit version with IBM's compiler, I used the
407 following:
408
409 export PATH=/usr/bin:/usr/vacpp/bin
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]410 ./configure --with-gcc="xlc_r -q64" --with-cxx="xlC_r -q64" \
Guido van Rossuma2534c82005-09-20 19:06:23[diff] [blame]411 --disable-ipv6 AR="ar -X64"
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]412 make
Guido van Rossuma2534c82005-09-20 19:06:23[diff] [blame]413
Anthony Baxter7c497472003-10-01 07:24:54[diff] [blame]414HP-UX: When using threading, you may have to add -D_REENTRANT to the
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]415 OPT variable in the top-level Makefile; reported by Pat Knight,
416 this seems to make a difference (at least for HP-UX 10.20)
417 even though pyconfig.h defines it. This seems unnecessary when
418 using HP/UX 11 and later - threading seems to work "out of the
419 box".
Anthony Baxter7c497472003-10-01 07:24:54[diff] [blame]420
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]421HP-UX ia64: When building on the ia64 (Itanium) platform using HP's
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]422 compiler, some experience has shown that the compiler's
423 optimiser produces a completely broken version of python
424 (see http://www.python.org/sf/814976). To work around this,
425 edit the Makefile and remove -O from the OPT line.
Guido van Rossum539c6622005-09-14 17:49:54[diff] [blame]426
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]427 To build a 64-bit executable on an Itanium 2 system using HP's
428 compiler, use these environment variables:
Guido van Rossum539c6622005-09-14 17:49:54[diff] [blame]429
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]430 CC=cc
431 CXX=aCC
432 BASECFLAGS="+DD64"
433 LDFLAGS="+DD64 -lxnet"
Guido van Rossum539c6622005-09-14 17:49:54[diff] [blame]434
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]435 and call configure as:
Guido van Rossum539c6622005-09-14 17:49:54[diff] [blame]436
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]437 ./configure --without-gcc
Guido van Rossum539c6622005-09-14 17:49:54[diff] [blame]438
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]439 then *unset* the environment variables again before running
440 make. (At least one of these flags causes the build to fail
441 if it remains set.) You still have to edit the Makefile and
442 remove -O from the OPT line.
Guido van Rossum2094e041998-05-14 15:47:35[diff] [blame]443
Guido van Rossum4e1dd7d2002-04-23 13:06:07[diff] [blame]444HP PA-RISC 2.0: A recent bug report (http://www.python.org/sf/546117)
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]445 suggests that the C compiler in this 64-bit system has bugs
446 in the optimizer that break Python. Compiling without
447 optimization solves the problems.
Guido van Rossum4e1dd7d2002-04-23 13:06:07[diff] [blame]448
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]449SCO: The following apply to SCO 3 only; Python builds out of the box
450 on SCO 5 (or so we've heard).
Guido van Rossuma16e2751998-04-13 20:14:05[diff] [blame]451
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]452 1) Everything works much better if you add -U__STDC__ to the
453 defs. This is because all the SCO header files are broken.
454 Anything that isn't mentioned in the C standard is
455 conditionally excluded when __STDC__ is defined.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]456
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]457 2) Due to the U.S. export restrictions, SCO broke the crypt
458 stuff out into a separate library, libcrypt_i.a so the LIBS
459 needed be set to:
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]460
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]461 LIBS=' -lsocket -lcrypt_i'
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]462
Martin v. Löwis387c5472001-09-06 08:16:17[diff] [blame]463UnixWare: There are known bugs in the math library of the system, as well as
464 problems in the handling of threads (calling fork in one
465 thread may interrupt system calls in others). Therefore, test_math and
466 tests involving threads will fail until those problems are fixed.
467
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]468QNX: Chris Herborth (chrish@qnx.com) writes:
469 configure works best if you use GNU bash; a port is available on
470 ftp.qnx.com in /usr/free. I used the following process to build,
471 test and install Python 1.5.x under QNX:
Guido van Rossumb7f454d1997-12-02 19:44:31[diff] [blame]472
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]473 1) CONFIG_SHELL=/usr/local/bin/bash CC=cc RANLIB=: \
474 ./configure --verbose --without-gcc --with-libm=""
Guido van Rossumb7f454d1997-12-02 19:44:31[diff] [blame]475
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]476 2) edit Modules/Setup to activate everything that makes sense for
477 your system... tested here at QNX with the following modules:
Guido van Rossumb7f454d1997-12-02 19:44:31[diff] [blame]478
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]479 array, audioop, binascii, cPickle, cStringIO, cmath,
480 crypt, curses, errno, fcntl, gdbm, grp, imageop,
481 _locale, math, md5, new, operator, parser, pcre,
482 posix, pwd, readline, regex, reop, rgbimg, rotor,
483 select, signal, socket, soundex, strop, struct,
484 syslog, termios, time, timing, zlib, audioop, imageop, rgbimg
Guido van Rossumb7f454d1997-12-02 19:44:31[diff] [blame]485
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]486 3) make SHELL=/usr/local/bin/bash
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]487
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]488 or, if you feel the need for speed:
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]489
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]490 make SHELL=/usr/local/bin/bash OPT="-5 -Oil+nrt"
Guido van Rossumb7f454d1997-12-02 19:44:31[diff] [blame]491
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]492 4) make SHELL=/usr/local/bin/bash test
Guido van Rossumec95c7b1998-08-04 17:59:56[diff] [blame]493
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]494 Using GNU readline 2.2 seems to behave strangely, but I
495 think that's a problem with my readline 2.2 port. :-\
Guido van Rossumec95c7b1998-08-04 17:59:56[diff] [blame]496
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]497 5) make SHELL=/usr/local/bin/bash install
Guido van Rossumb7f454d1997-12-02 19:44:31[diff] [blame]498
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]499 If you get SIGSEGVs while running Python (I haven't yet, but
500 I've only run small programs and the test cases), you're
501 probably running out of stack; the default 32k could be a
502 little tight. To increase the stack size, edit the Makefile
503 to read: LDFLAGS = -N 48k
Guido van Rossum40d63581997-08-14 19:45:30[diff] [blame]504
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]505BeOS: See Misc/BeOS-NOTES for notes about compiling/installing
506 Python on BeOS R3 or later. Note that only the PowerPC
507 platform is supported for R3; both PowerPC and x86 are
508 supported for R4.
Guido van Rossumec95c7b1998-08-04 17:59:56[diff] [blame]509
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]510Cray T3E: Mark Hadfield (m.hadfield@niwa.co.nz) writes:
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]511 Python can be built satisfactorily on a Cray T3E but based on
512 my experience with the NIWA T3E (2002-05-22, version 2.2.1)
513 there are a few bugs and gotchas. For more information see a
514 thread on comp.lang.python in May 2002 entitled "Building
515 Python on Cray T3E".
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]516
517 1) Use Cray's cc and not gcc. The latter was reported not to
518 work by Konrad Hinsen. It may work now, but it may not.
519
520 2) To set sys.platform to something sensible, pass the
521 following environment variable to the configure script:
522
523 MACHDEP=unicosmk
524
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]525 2) Run configure with option "--enable-unicode=ucs4".
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]526
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]527 3) The Cray T3E does not support dynamic linking, so extension
528 modules have to be built by adding (or uncommenting) lines
529 in Modules/Setup. The minimum set of modules is
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]530
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]531 posix, new, _sre, unicodedata
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]532
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]533 On NIWA's vanilla T3E system the following have also been
534 included successfully:
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]535
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]536 _codecs, _locale, _socket, _symtable, _testcapi, _weakref
537 array, binascii, cmath, cPickle, crypt, cStringIO, dbm
538 errno, fcntl, grp, math, md5, operator, parser, pcre, pwd
539 regex, rotor, select, struct, strop, syslog, termios
540 time, timing, xreadlines
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]541
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]542 4) Once the python executable and library have been built, make
543 will execute setup.py, which will attempt to build remaining
544 extensions and link them dynamically. Each of these attempts
545 will fail but should not halt the make process. This is
546 normal.
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]547
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]548 5) Running "make test" uses a lot of resources and causes
549 problems on our system. You might want to try running tests
550 singly or in small groups.
Guido van Rossum1bf0bf41997-08-20 23:50:51[diff] [blame]551
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]552SGI: SGI's standard "make" utility (/bin/make or /usr/bin/make)
553 does not check whether a command actually changed the file it
554 is supposed to build. This means that whenever you say "make"
555 it will redo the link step. The remedy is to use SGI's much
556 smarter "smake" utility (/usr/sbin/smake), or GNU make. If
557 you set the first line of the Makefile to #!/usr/sbin/smake
558 smake will be invoked by make (likewise for GNU make).
Guido van Rossum0078aaf1997-08-21 03:05:11[diff] [blame]559
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]560 WARNING: There are bugs in the optimizer of some versions of
561 SGI's compilers that can cause bus errors or other strange
562 behavior, especially on numerical operations. To avoid this,
563 try building with "make OPT=".
Fred Drake0b5fb2b2000-09-29 17:45:05[diff] [blame]564
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]565OS/2: If you are running Warp3 or Warp4 and have IBM's VisualAge C/C++
566 compiler installed, just change into the pc\os2vacpp directory
567 and type NMAKE. Threading and sockets are supported by default
568 in the resulting binaries of PYTHON15.DLL and PYTHON.EXE.
569
Fred Drake36ed5602000-10-06 01:58:48[diff] [blame]570Monterey (64-bit AIX): The current Monterey C compiler (Visual Age)
571 uses the OBJECT_MODE={32|64} environment variable to set the
572 compilation mode to either 32-bit or 64-bit (32-bit mode is
573 the default). Presumably you want 64-bit compilation mode for
574 this 64-bit OS. As a result you must first set OBJECT_MODE=64
575 in your environment before configuring (./configure) or
576 building (make) Python on Monterey.
577
578Reliant UNIX: The thread support does not compile on Reliant UNIX, and
579 there is a (minor) problem in the configure script for that
580 platform as well. This should be resolved in time for a
581 future release.
Trent Mick635f6fb2000-08-23 21:33:05[diff] [blame]582
Barry Warsaw6a8557d2002-10-14 18:04:39[diff] [blame]583MacOSX: The tests will crash on both 10.1 and 10.2 with SEGV in
Barry Warsaw6e753642002-10-14 18:15:35[diff] [blame]584 test_re and test_sre due to the small default stack size. If
585 you set the stack size to 2048 before doing a "make test" the
Georg Brandl2230d982007-03-06 18:47:40[diff] [blame]586 failure can be avoided. If you're using the tcsh or csh shells,
587 use "limit stacksize 2048" and for the bash shell (the default
588 as of OSX 10.3), use "ulimit -s 2048".
Barry Warsaw06557452001-12-14 20:47:12[diff] [blame]589
Jack Jansen6a77cf12001-12-06 21:49:02[diff] [blame]590 On naked Darwin you may want to add the configure option
Jack Jansen4c398fd2001-10-08 13:21:15[diff] [blame]591 "--disable-toolbox-glue" to disable the glue code for the Carbon
Jack Jansen6a77cf12001-12-06 21:49:02[diff] [blame]592 interface modules. The modules themselves are currently only built
593 if you add the --enable-framework option, see below.
Barry Warsaw06557452001-12-14 20:47:12[diff] [blame]594
595 On a clean OSX /usr/local does not exist. Do a
Jack Jansen61fec302002-01-04 15:59:57[diff] [blame]596 "sudo mkdir -m 775 /usr/local"
Jack Jansen84a01642002-08-04 21:19:55[diff] [blame]597 before you do a make install. It is probably not a good idea to
598 do "sudo make install" which installs everything as superuser,
599 as this may later cause problems when installing distutils-based
600 additions.
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]601
Jack Jansen84a01642002-08-04 21:19:55[diff] [blame]602 Some people have reported problems building Python after using "fink"
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]603 to install additional unix software. Disabling fink (remove all
604 references to /sw from your .profile or .login) should solve this.
Barry Warsaw06557452001-12-14 20:47:12[diff] [blame]605
606 You may want to try the configure option "--enable-framework"
607 which installs Python as a framework. The location can be set
608 as argument to the --enable-framework option (default
Jack Jansen84a01642002-08-04 21:19:55[diff] [blame]609 /Library/Frameworks). A framework install is probably needed if you
610 want to use any Aqua-based GUI toolkit (whether Tkinter, wxPython,
611 Carbon, Cocoa or anything else).
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]612
Ronald Oussoren988117f2006-04-29 11:31:35[diff] [blame]613 You may also want to try the configure option "--enable-universalsdk"
614 which builds Python as a universal binary with support for the
Anthony Baxter347f3cc2006-09-18 06:45:59[diff] [blame]615 i386 and PPC architectures. This requires Xcode 2.1 or later to build.
Ronald Oussoren988117f2006-04-29 11:31:35[diff] [blame]616
617 See Mac/OSX/README for more information on framework and
618 universal builds.
Fred Drakedabed752001-02-01 19:41:13[diff] [blame]619
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]620Cygwin: With recent (relative to the time of writing, 2001-12-19)
621 Cygwin installations, there are problems with the interaction
622 of dynamic linking and fork(). This manifests itself in build
623 failures during the execution of setup.py.
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]624
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]625 There are two workarounds that both enable Python (albeit
626 without threading support) to build and pass all tests on
627 NT/2000 (and most likely XP as well, though reports of testing
628 on XP would be appreciated).
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]629
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]630 The workarounds:
631
632 (a) the band-aid fix is to link the _socket module statically
633 rather than dynamically (which is the default).
634
635 To do this, run "./configure --with-threads=no" including any
636 other options you need (--prefix, etc.). Then in Modules/Setup
Walter Dörwald7dc65d62003-02-02 23:37:05[diff] [blame]637 uncomment the lines:
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]638
Michael W. Hudsonbe5ce182001-12-12 11:46:52[diff] [blame]639 #SSL=/usr/local/ssl
640 #_socket socketmodule.c \
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]641 # -DUSE_SSL -I$(SSL)/include -I$(SSL)/include/openssl \
642 # -L$(SSL)/lib -lssl -lcrypto
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]643
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]644 and remove "local/" from the SSL variable. Finally, just run
645 "make"!
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]646
Michael W. Hudsonb6554642001-12-19 22:09:09[diff] [blame]647 (b) The "proper" fix is to rebase the Cygwin DLLs to prevent
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]648 base address conflicts. Details on how to do this can be
649 found in the following mail:
650
651 http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html
652
653 It is hoped that a version of this solution will be
Michael W. Hudsonb6554642001-12-19 22:09:09[diff] [blame]654 incorporated into the Cygwin distribution fairly soon.
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]655
656 Two additional problems:
657
658 (1) Threading support should still be disabled due to a known
659 bug in Cygwin pthreads that causes test_threadedtempfile to
660 hang.
661
Neal Norwitz4ebde092001-12-19 20:44:13[diff] [blame]662 (2) The _curses module does not build. This is a known
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]663 Cygwin ncurses problem that should be resolved the next time
664 that this package is released.
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]665
Michael W. Hudsonbe5ce182001-12-12 11:46:52[diff] [blame]666 On older versions of Cygwin, test_poll may hang and test_strftime
667 may fail.
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]668
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]669 The situation on 9X/Me is not accurately known at present.
670 Some time ago, there were reports that the following
671 regression tests failed:
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]672
673 test_pwd
674 test_select (hang)
675 test_socket
676
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]677 Due to the test_select hang on 9X/Me, one should run the
678 regression test using the following:
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]679
Michael W. Hudsonbe5ce182001-12-12 11:46:52[diff] [blame]680 make TESTOPTS='-l -x test_select' test
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]681
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]682 News regarding these platforms with more recent Cygwin
683 versions would be appreciated!
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]684
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]685AtheOS: From Octavian Cerna <tavy at ylabs.com>:
686
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]687 Before building:
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]688
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]689 Make sure you have shared versions of the libraries you
690 want to use with Python. You will have to compile them
691 yourself, or download precompiled packages.
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]692
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]693 Recommended libraries:
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]694
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]695 ncurses-4.2
696 readline-4.2a
697 zlib-1.1.4
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]698
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]699 Build:
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]700
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]701 $ ./configure --prefix=/usr/python
702 $ make
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]703
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]704 Python is always built as a shared library, otherwise
705 dynamic loading would not work.
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]706
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]707 Testing:
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]708
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]709 $ make test
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]710
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]711 Install:
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]712
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]713 # make install
714 # pkgmanager -a /usr/python
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]715
716
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]717 AtheOS issues:
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]718
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]719 - large file support: due to a stdio bug in glibc/libio,
720 access to large files may not work correctly. fseeko()
721 tries to seek to a negative offset. ftello() returns a
722 negative offset, it looks like a 32->64bit
723 sign-extension issue. The lowlevel functions (open,
724 lseek, etc) are OK.
725 - sockets: AF_UNIX is defined in the C library and in
726 Python, but not implemented in the system.
727 - select: poll is available in the C library, but does not
728 work (It does not return POLLNVAL for bad fds and
729 hangs).
730 - posix: statvfs and fstatvfs always return ENOSYS.
731 - disabled modules:
732 - mmap: not yet implemented in AtheOS
733 - nis: broken (on an unconfigured system
734 yp_get_default_domain() returns junk instead of
735 error)
736 - dl: dynamic loading doesn't work via dlopen()
737 - resource: getrimit and setrlimit are not yet
738 implemented
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]739
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]740 - if you are getting segmentation faults, you probably are
741 low on memory. AtheOS doesn't handle very well an
742 out-of-memory condition and simply SEGVs the process.
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]743
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]744 Tested on:
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]745
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]746 AtheOS-0.3.7
747 gcc-2.95
748 binutils-2.10
749 make-3.78
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]750
751
Skip Montanaro57454e52002-06-14 20:30:31[diff] [blame]752Configuring the bsddb and dbm modules
753-------------------------------------
754
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]755Beginning with Python version 2.3, the PyBsddb package
756<http://pybsddb.sf.net/> was adopted into Python as the bsddb package,
757exposing a set of package-level functions which provide
Gregory P. Smith3adc4aa2006-04-13 19:19:01[diff] [blame]758backwards-compatible behavior. Only versions 3.3 through 4.4 of
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]759Sleepycat's libraries provide the necessary API, so older versions
760aren't supported through this interface. The old bsddb module has
761been retained as bsddb185, though it is not built by default. Users
762wishing to use it will have to tweak Modules/Setup to build it. The
763dbm module will still be built against the Sleepycat libraries if
Gregory P. Smith3adc4aa2006-04-13 19:19:01[diff] [blame]764other preferred alternatives (ndbm, gdbm) are not found.
Guido van Rossum62d45c02002-12-02 15:49:13[diff] [blame]765
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]766Building the sqlite3 module
767---------------------------
768
769To build the sqlite3 module, you'll need the sqlite3 or libsqlite3
770packages installed, including the header files. Many modern operating
771systems distribute the headers in a separate package to the library -
772often it will be the same name as the main package, but with a -dev or
773-devel suffix.
774
775The version of pysqlite2 that's including in Python needs sqlite3 3.0.8
776or later. setup.py attempts to check that it can find a correct version.
Skip Montanaro57454e52002-06-14 20:30:31[diff] [blame]777
Guido van Rossumdc8a3cb1998-05-12 15:29:18[diff] [blame]778Configuring threads
779-------------------
780
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]781As of Python 2.0, threads are enabled by default. If you wish to
782compile without threads, or if your thread support is broken, pass the
783--with-threads=no switch to configure. Unfortunately, on some
784platforms, additional compiler and/or linker options are required for
785threads to work properly. Below is a table of those options,
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]786collected by Bill Janssen. We would love to automate this process
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]787more, but the information below is not enough to write a patch for the
788configure.in file, so manual intervention is required. If you patch
789the configure.in file and are confident that the patch works, please
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]790send in the patch. (Don't bother patching the configure script itself
Walter Dörwald7dc65d62003-02-02 23:37:05[diff] [blame]791-- it is regenerated each time the configure.in file changes.)
Guido van Rossumdc8a3cb1998-05-12 15:29:18[diff] [blame]792
793Compiler switches for threads
794.............................
795
Martin v. Löwis1c24f9c2001-03-06 12:07:27[diff] [blame]796The definition of _REENTRANT should be configured automatically, if
797that does not work on your system, or if _REENTRANT is defined
798incorrectly, please report that as a bug.
Guido van Rossumdc8a3cb1998-05-12 15:29:18[diff] [blame]799
Martin v. Löwis1c24f9c2001-03-06 12:07:27[diff] [blame]800 OS/Compiler/threads Switches for use with threads
801 (POSIX is draft 10, DCE is draft 4) compile & link
802
803 SunOS 5.{1-5}/{gcc,SunPro cc}/solaris -mt
804 SunOS 5.5/{gcc,SunPro cc}/POSIX (nothing)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]805 DEC OSF/1 3.x/cc/DCE -threads
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]806 (butenhof@zko.dec.com)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]807 Digital UNIX 4.x/cc/DCE -threads
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]808 (butenhof@zko.dec.com)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]809 Digital UNIX 4.x/cc/POSIX -pthread
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]810 (butenhof@zko.dec.com)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]811 AIX 4.1.4/cc_r/d7 (nothing)
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]812 (buhrt@iquest.net)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]813 AIX 4.1.4/cc_r4/DCE (nothing)
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]814 (buhrt@iquest.net)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]815 IRIX 6.2/cc/POSIX (nothing)
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]816 (robertl@cwi.nl)
Guido van Rossumdc8a3cb1998-05-12 15:29:18[diff] [blame]817
818
819Linker (ld) libraries and flags for threads
820...........................................
821
822 OS/threads Libraries/switches for use with threads
823
824 SunOS 5.{1-5}/solaris -lthread
825 SunOS 5.5/POSIX -lpthread
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]826 DEC OSF/1 3.x/DCE -lpthreads -lmach -lc_r -lc
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]827 (butenhof@zko.dec.com)
Guido van Rossumdc8a3cb1998-05-12 15:29:18[diff] [blame]828 Digital UNIX 4.x/DCE -lpthreads -lpthread -lmach -lexc -lc
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]829 (butenhof@zko.dec.com)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]830 Digital UNIX 4.x/POSIX -lpthread -lmach -lexc -lc
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]831 (butenhof@zko.dec.com)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]832 AIX 4.1.4/{draft7,DCE} (nothing)
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]833 (buhrt@iquest.net)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]834 IRIX 6.2/POSIX -lpthread
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]835 (jph@emilia.engr.sgi.com)
Guido van Rossumdc8a3cb1998-05-12 15:29:18[diff] [blame]836
837
Martin v. Löwis1142de32002-03-29 16:28:31[diff] [blame]838Building a shared libpython
839---------------------------
840
841Starting with Python 2.3, the majority of the interpreter can be built
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]842into a shared library, which can then be used by the interpreter
Martin v. Löwis1142de32002-03-29 16:28:31[diff] [blame]843executable, and by applications embedding Python. To enable this feature,
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]844configure with --enable-shared.
Martin v. Löwis65069672002-08-03 21:38:27[diff] [blame]845
Fred Drake55512dc2003-07-21 16:01:39[diff] [blame]846If you enable this feature, the same object files will be used to create
847a static library. In particular, the static library will contain object
848files using position-independent code (PIC) on platforms where PIC flags
849are needed for the shared library.
Martin v. Löwis1142de32002-03-29 16:28:31[diff] [blame]850
851
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]852Configuring additional built-in modules
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]853---------------------------------------
Guido van Rossum19e0c261995-01-17 16:36:34[diff] [blame]854
Andrew M. Kuchlingebe30452001-02-27 03:24:24[diff] [blame]855Starting with Python 2.1, the setup.py script at the top of the source
856distribution attempts to detect which modules can be built and
857automatically compiles them. Autodetection doesn't always work, so
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]858you can still customize the configuration by editing the Modules/Setup
859file; but this should be considered a last resort. The rest of this
860section only applies if you decide to edit the Modules/Setup file.
861You also need this to enable static linking of certain modules (which
862is needed to enable profiling on some systems).
863
Andrew M. Kuchlingebe30452001-02-27 03:24:24[diff] [blame]864This file is initially copied from Setup.dist by the configure script;
865if it does not exist yet, create it by copying Modules/Setup.dist
866yourself (configure will never overwrite it). Never edit Setup.dist
867-- always edit Setup or Setup.local (see below). Read the comments in
868the file for information on what kind of edits are allowed. When you
869have edited Setup in the Modules directory, the interpreter will
Guido van Rossumef67ded2001-10-05 17:04:42[diff] [blame]870automatically be rebuilt the next time you run make (in the toplevel
871directory).
Guido van Rossum627b2d71993-12-24 10:39:16[diff] [blame]872
Andrew M. Kuchlingebe30452001-02-27 03:24:24[diff] [blame]873Many useful modules can be built on any Unix system, but some optional
874modules can't be reliably autodetected. Often the quickest way to
875determine whether a particular module works or not is to see if it
876will build: enable it in Setup, then if you get compilation or link
Fred Drake6caae142000-10-25 17:51:02[diff] [blame]877errors, disable it -- you're either missing support or need to adjust
878the compilation and linking parameters for that module.
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]879
880On SGI IRIX, there are modules that interface to many SGI specific
Andrew M. Kuchlingebe30452001-02-27 03:24:24[diff] [blame]881system libraries, e.g. the GL library and the audio hardware. These
882modules will not be built by the setup.py script.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]883
Guido van Rossum4462e931997-01-22 21:00:32[diff] [blame]884In addition to the file Setup, you can also edit the file Setup.local.
885(the makesetup script processes both). You may find it more
886convenient to edit Setup.local and leave Setup alone. Then, when
887installing a new Python version, you can copy your old Setup.local
888file.
889
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]890
891Setting the optimization/debugging options
892------------------------------------------
893
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]894If you want or need to change the optimization/debugging options for
895the C compiler, assign to the OPT variable on the toplevel make
896command; e.g. "make OPT=-g" will build a debugging version of Python
897on most platforms. The default is OPT=-O; a value for OPT in the
898environment when the configure script is run overrides this default
899(likewise for CC; and the initial value for LIBS is used as the base
900set of libraries to link with).
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]901
Fred Drake6caae142000-10-25 17:51:02[diff] [blame]902When compiling with GCC, the default value of OPT will also include
903the -Wall and -Wstrict-prototypes options.
904
905Additional debugging code to help debug memory management problems can
906be enabled by using the --with-pydebug option to the configure script.
907
Brett Cannon08cd5982005-04-24 22:26:38[diff] [blame]908For flags that change binary compatibility, use the EXTRA_CFLAGS
909variable.
910
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]911
Guido van Rossuma68ca352001-10-04 17:00:07[diff] [blame]912Profiling
913---------
914
915If you want C profiling turned on, the easiest way is to run configure
916with the CC environment variable to the necessary compiler
917invocation. For example, on Linux, this works for profiling using
918gprof(1):
919
920 CC="gcc -pg" ./configure
921
922Note that on Linux, gprof apparently does not work for shared
923libraries. The Makefile/Setup mechanism can be used to compile and
Walter Dörwald7dc65d62003-02-02 23:37:05[diff] [blame]924link most extension modules statically.
Guido van Rossuma68ca352001-10-04 17:00:07[diff] [blame]925
926
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]927Testing
928-------
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]929
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]930To test the interpreter, type "make test" in the top-level directory.
Guido van Rossum6ae5d3d1997-05-14 21:39:05[diff] [blame]931This runs the test set twice (once with no compiled files, once with
932the compiled files left by the previous test run). The test set
933produces some output. You can generally ignore the messages about
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]934skipped tests due to optional features which can't be imported.
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]935If a message is printed about a failed test or a traceback or core
936dump is produced, something is wrong. On some Linux systems (those
Guido van Rossum24df6841997-12-30 04:32:30[diff] [blame]937that are not yet using glibc 6), test_strftime fails due to a
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]938non-standard implementation of strftime() in the C library. Please
939ignore this, or upgrade to glibc version 6.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]940
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]941IMPORTANT: If the tests fail and you decide to mail a bug report,
942*don't* include the output of "make test". It is useless. Run the
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]943failing test manually, as follows:
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]944
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]945 ./python ./Lib/test/test_whatever.py
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]946
Andrew M. Kuchling83d042d2002-03-21 23:52:20[diff] [blame]947(substituting the top of the source tree for '.' if you built in a
Guido van Rossum6ae5d3d1997-05-14 21:39:05[diff] [blame]948different directory). This runs the test in verbose mode.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]949
950
951Installing
952----------
953
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]954To install the Python binary, library modules, shared library modules
955(see below), include files, configuration files, and the manual page,
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]956just type
957
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]958 make install
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]959
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]960This will install all platform-independent files in subdirectories of
961the directory given with the --prefix option to configure or to the
962`prefix' Make variable (default /usr/local). All binary and other
963platform-specific files will be installed in subdirectories if the
964directory given by --exec-prefix or the `exec_prefix' Make variable
965(defaults to the --prefix directory) is given.
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]966
Martin v. Löwis3b8ee082003-05-11 20:25:35[diff] [blame]967If DESTDIR is set, it will be taken as the root directory of the
968installation, and files will be installed into $(DESTDIR)$(prefix),
969$(DESTDIR)$(exec_prefix), etc.
970
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]971All subdirectories created will have Python's version number in their
972name, e.g. the library modules are installed in
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]973"/usr/local/lib/python<version>/" by default, where <version> is the
974<major>.<minor> release number (e.g. "2.1"). The Python binary is
975installed as "python<version>" and a hard link named "python" is
976created. The only file not installed with a version number in its
977name is the manual page, installed as "/usr/local/man/man1/python.1"
978by default.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]979
Neil Schemenauer0e91dc72001-02-16 04:18:08[diff] [blame]980If you have a previous installation of Python that you don't
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]981want to replace yet, use
982
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]983 make altinstall
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]984
985This installs the same set of files as "make install" except it
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]986doesn't create the hard link to "python<version>" named "python" and
987it doesn't install the manual page at all.
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]988
989The only thing you may have to install manually is the Python mode for
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]990Emacs found in Misc/python-mode.el. (But then again, more recent
991versions of Emacs may already have it.) Follow the instructions that
992came with Emacs for installation of site-specific files.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]993
Jack Jansena39ef862001-08-19 21:17:03[diff] [blame]994On Mac OS X, if you have configured Python with --enable-framework, you
995should use "make frameworkinstall" to do the installation. Note that this
996installs the Python executable in a place that is not normally on your
997PATH, you may want to set up a symlink in /usr/local/bin.
998
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]999
1000Configuration options and variables
1001-----------------------------------
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1002
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1003Some special cases are handled by passing options to the configure
1004script.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1005
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1006WARNING: if you rerun the configure script with different options, you
1007must run "make clean" before rebuilding. Exceptions to this rule:
1008after changing --prefix or --exec-prefix, all you need to do is remove
Guido van Rossumb06df271997-08-05 21:50:20[diff] [blame]1009Modules/getpath.o.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1010
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1011--with(out)-gcc: The configure script uses gcc (the GNU C compiler) if
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1012 it finds it. If you don't want this, or if this compiler is
1013 installed but broken on your platform, pass the option
1014 --without-gcc. You can also pass "CC=cc" (or whatever the
1015 name of the proper C compiler is) in the environment, but the
1016 advantage of using --without-gcc is that this option is
1017 remembered by the config.status script for its --recheck
1018 option.
Guido van Rossum76be6ed1995-01-02 18:33:54[diff] [blame]1019
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1020--prefix, --exec-prefix: If you want to install the binaries and the
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1021 Python library somewhere else than in /usr/local/{bin,lib},
1022 you can pass the option --prefix=DIRECTORY; the interpreter
1023 binary will be installed as DIRECTORY/bin/python and the
1024 library files as DIRECTORY/lib/python/*. If you pass
1025 --exec-prefix=DIRECTORY (as well) this overrides the
1026 installation prefix for architecture-dependent files (like the
1027 interpreter binary). Note that --prefix=DIRECTORY also
1028 affects the default module search path (sys.path), when
1029 Modules/config.c is compiled. Passing make the option
1030 prefix=DIRECTORY (and/or exec_prefix=DIRECTORY) overrides the
1031 prefix set at configuration time; this may be more convenient
1032 than re-running the configure script if you change your mind
1033 about the install prefix.
Guido van Rossum76be6ed1995-01-02 18:33:54[diff] [blame]1034
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]1035--with-readline: This option is no longer supported. GNU
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1036 readline is automatically enabled by setup.py when present.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1037
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]1038--with-threads: On most Unix systems, you can now use multiple
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1039 threads, and support for this is enabled by default. To
1040 disable this, pass --with-threads=no. If the library required
1041 for threads lives in a peculiar place, you can use
1042 --with-thread=DIRECTORY. IMPORTANT: run "make clean" after
1043 changing (either enabling or disabling) this option, or you
1044 will get link errors! Note: for DEC Unix use
1045 --with-dec-threads instead.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1046
1047--with-sgi-dl: On SGI IRIX 4, dynamic loading of extension modules is
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1048 supported by the "dl" library by Jack Jansen, which is
1049 ftp'able from ftp://ftp.cwi.nl/pub/dynload/dl-1.6.tar.Z.
1050 This is enabled (after you've ftp'ed and compiled the dl
1051 library) by passing --with-sgi-dl=DIRECTORY where DIRECTORY
1052 is the absolute pathname of the dl library. (Don't bother on
1053 IRIX 5, it already has dynamic linking using SunOS style
1054 shared libraries.) THIS OPTION IS UNSUPPORTED.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1055
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1056--with-dl-dld: Dynamic loading of modules is rumored to be supported
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1057 on some other systems: VAX (Ultrix), Sun3 (SunOS 3.4), Sequent
1058 Symmetry (Dynix), and Atari ST. This is done using a
1059 combination of the GNU dynamic loading package
1060 (ftp://ftp.cwi.nl/pub/dynload/dl-dld-1.1.tar.Z) and an
1061 emulation of the SGI dl library mentioned above (the emulation
1062 can be found at
1063 ftp://ftp.cwi.nl/pub/dynload/dld-3.2.3.tar.Z). To
1064 enable this, ftp and compile both libraries, then call
1065 configure, passing it the option
1066 --with-dl-dld=DL_DIRECTORY,DLD_DIRECTORY where DL_DIRECTORY is
1067 the absolute pathname of the dl emulation library and
1068 DLD_DIRECTORY is the absolute pathname of the GNU dld library.
1069 (Don't bother on SunOS 4 or 5, they already have dynamic
1070 linking using shared libraries.) THIS OPTION IS UNSUPPORTED.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1071
1072--with-libm, --with-libc: It is possible to specify alternative
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1073 versions for the Math library (default -lm) and the C library
1074 (default the empty string) using the options
1075 --with-libm=STRING and --with-libc=STRING, respectively. For
1076 example, if your system requires that you pass -lc_s to the C
1077 compiler to use the shared C library, you can pass
1078 --with-libc=-lc_s. These libraries are passed after all other
1079 libraries, the C library last.
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]1080
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1081--with-libs='libs': Add 'libs' to the LIBS that the python interpreter
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1082 is linked against.
Guido van Rossumd02ba451996-07-31 17:36:01[diff] [blame]1083
Martin v. Löwis0f48d982006-04-14 14:34:26[diff] [blame]1084--with-cxx-main=<compiler>: If you plan to use C++ extension modules,
1085 then -- on some platforms -- you need to compile python's main()
1086 function with the C++ compiler. With this option, make will use
1087 <compiler> to compile main() *and* to link the python executable.
1088 It is likely that the resulting executable depends on the C++
1089 runtime library of <compiler>. (The default is --without-cxx-main.)
1090
1091 There are platforms that do not require you to build Python
1092 with a C++ compiler in order to use C++ extension modules.
1093 E.g., x86 Linux with ELF shared binaries and GCC 3.x, 4.x is such
1094 a platform. We recommend that you configure Python
1095 --without-cxx-main on those platforms because a mismatch
1096 between the C++ compiler version used to build Python and to
1097 build a C++ extension module is likely to cause a crash at
1098 runtime.
1099
1100 The Python installation also stores the variable CXX that
1101 determines, e.g., the C++ compiler distutils calls by default
1102 to build C++ extensions. If you set CXX on the configure command
1103 line to any string of non-zero length, then configure won't
1104 change CXX. If you do not preset CXX but pass
1105 --with-cxx-main=<compiler>, then configure sets CXX=<compiler>.
1106 In all other cases, configure looks for a C++ compiler by
1107 some common names (c++, g++, gcc, CC, cxx, cc++, cl) and sets
1108 CXX to the first compiler it finds. If it does not find any
1109 C++ compiler, then it sets CXX="".
1110
1111 Similarly, if you want to change the command used to link the
1112 python executable, then set LINKCC on the configure command line.
Martin v. Löwis2df66942000-12-13 14:14:32[diff] [blame]1113
1114
Fred Drake6caae142000-10-25 17:51:02[diff] [blame]1115--with-pydebug: Enable additional debugging code to help track down
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1116 memory management problems. This allows printing a list of all
1117 live objects when the interpreter terminates.
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]1118
Jack Jansen7b8c7542002-04-14 20:12:41[diff] [blame]1119--with(out)-universal-newlines: enable reading of text files with
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1120 foreign newline convention (default: enabled). In other words,
1121 any of \r, \n or \r\n is acceptable as end-of-line character.
1122 If enabled import and execfile will automatically accept any newline
1123 in files. Python code can open a file with open(file, 'U') to
1124 read it in universal newline mode. THIS OPTION IS UNSUPPORTED.
Fred Drake6caae142000-10-25 17:51:02[diff] [blame]1125
Martin v. Löwisf30d60e2004-06-08 08:17:44[diff] [blame]1126--with-tsc: Profile using the Pentium timestamping counter (TSC).
Guido van Rossum76be6ed1995-01-02 18:33:54[diff] [blame]1127
Martin v. Löwis9176fc12006-04-11 11:12:43[diff] [blame]1128--with-system-ffi: Build the _ctypes extension module using an ffi
1129 library installed on the system.
1130
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]1131
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1132Building for multiple architectures (using the VPATH feature)
1133-------------------------------------------------------------
1134
1135If your file system is shared between multiple architectures, it
1136usually is not necessary to make copies of the sources for each
1137architecture you want to support. If the make program supports the
1138VPATH feature, you can create an empty build directory for each
1139architecture, and in each directory run the configure script (on the
1140appropriate machine with the appropriate options). This creates the
1141necessary subdirectories and the Makefiles therein. The Makefiles
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1142contain a line VPATH=... which points to a directory containing the
Guido van Rossumd0fe8451996-08-26 03:02:37[diff] [blame]1143actual sources. (On SGI systems, use "smake -J1" instead of "make" if
1144you use VPATH -- don't try gnumake.)
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1145
1146For example, the following is all you need to build a minimal Python
1147in /usr/tmp/python (assuming ~guido/src/python is the toplevel
1148directory and you want to build in /usr/tmp/python):
1149
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1150 $ mkdir /usr/tmp/python
1151 $ cd /usr/tmp/python
1152 $ ~guido/src/python/configure
1153 [...]
1154 $ make
1155 [...]
1156 $
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1157
Neil Schemenauer73f8ab22001-01-26 22:18:55[diff] [blame]1158Note that configure copies the original Setup file to the build
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1159directory if it finds no Setup file there. This means that you can
1160edit the Setup file for each architecture independently. For this
1161reason, subsequent changes to the original Setup file are not tracked
1162automatically, as they might overwrite local changes. To force a copy
1163of a changed original Setup file, delete the target Setup file. (The
1164makesetup script supports multiple input files, so if you want to be
1165fancy you can change the rules to create an empty Setup.local if it
1166doesn't exist and run it with arguments $(srcdir)/Setup Setup.local;
1167however this assumes that you only need to add modules.)
1168
1169
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1170Building on non-UNIX systems
1171----------------------------
1172
Martin v. Löwis2eba0d62004-12-24 08:28:28[diff] [blame]1173For Windows (2000/NT/ME/98/95), assuming you have MS VC++ 7.1, the
Tim Peters03444242000-09-19 00:38:35[diff] [blame]1174project files are in PCbuild, the workspace is pcbuild.dsw. See
1175PCbuild\readme.txt for detailed instructions.
Guido van Rossum8d7d4ed1996-07-30 21:41:07[diff] [blame]1176
Martin v. Löwis2eba0d62004-12-24 08:28:28[diff] [blame]1177For other non-Unix Windows compilers, in particular MS VC++ 6.0 and
Guido van Rossum31ae2071999-04-12 14:47:30[diff] [blame]1178for OS/2, enter the directory "PC" and read the file "readme.txt".
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1179
1180For the Mac, a separate source distribution will be made available,
1181for use with the CodeWarrior compiler. If you are interested in Mac
1182development, join the PythonMac Special Interest Group
1183(http://www.python.org/sigs/pythonmac-sig/, or send email to
1184pythonmac-sig-request@python.org).
1185
1186Of course, there are also binary distributions available for these
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]1187platforms -- see http://www.python.org/.
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1188
1189To port Python to a new non-UNIX system, you will have to fake the
1190effect of running the configure script manually (for Mac and PC, this
1191has already been done for you). A good start is to copy the file
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]1192pyconfig.h.in to pyconfig.h and edit the latter to reflect the actual
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1193configuration of your system. Most symbols must simply be defined as
11941 only if the corresponding feature is present and can be left alone
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]1195otherwise; however the *_t type symbols must be defined as some
1196variant of int if they need to be defined at all.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1197
Barry Warsaw06557452001-12-14 20:47:12[diff] [blame]1198For all platforms, it's important that the build arrange to define the
1199preprocessor symbol NDEBUG on the compiler command line in a release
1200build of Python (else assert() calls remain in the code, hurting
1201release-build performance). The Unix, Windows and Mac builds already
1202do this.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1203
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1204
1205Miscellaneous issues
1206====================
1207
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1208Emacs mode
1209----------
1210
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]1211There's an excellent Emacs editing mode for Python code; see the file
1212Misc/python-mode.el. Originally written by the famous Tim Peters, it
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1213is now maintained by the equally famous Barry Warsaw (it's no
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]1214coincidence that they now both work on the same team). The latest
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]1215version, along with various other contributed Python-related Emacs
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]1216goodies, is online at http://www.python.org/emacs/python-mode. And
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]1217if you are planning to edit the Python C code, please pick up the
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]1218latest version of CC Mode http://www.python.org/emacs/cc-mode; it
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]1219contains a "python" style used throughout most of the Python C source
1220files. (Newer versions of Emacs or XEmacs may already come with the
1221latest version of python-mode.)
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1222
1223
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]1224Tkinter
1225-------
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1226
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]1227The setup.py script automatically configures this when it detects a
1228usable Tcl/Tk installation. This requires Tcl/Tk version 8.0 or
Fred Drake8179a9d2000-09-28 23:28:30[diff] [blame]1229higher.
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1230
Guido van Rossum973e4dc2000-07-01 00:34:39[diff] [blame]1231For more Tkinter information, see the Tkinter Resource page:
1232http://www.python.org/topics/tkinter/
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]1233
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]1234There are demos in the Demo/tkinter directory.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1235
1236Note that there's a Python module called "Tkinter" (capital T) which
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]1237lives in Lib/lib-tk/Tkinter.py, and a C module called "_tkinter"
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1238(lower case t and leading underscore) which lives in
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1239Modules/_tkinter.c. Demos and normal Tk applications import only the
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]1240Python Tkinter module -- only the latter imports the C _tkinter
1241module. In order to find the C _tkinter module, it must be compiled
1242and linked into the Python interpreter -- the setup.py script does
1243this. In order to find the Python Tkinter module, sys.path must be
1244set correctly -- normal installation takes care of this.
Guido van Rossum84c8c7f1995-08-28 02:44:24[diff] [blame]1245
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1246
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1247Distribution structure
1248----------------------
1249
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1250Most subdirectories have their own README files. Most files have
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1251comments.
1252
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1253BeOS/ Files specific to the BeOS port
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1254Demo/ Demonstration scripts, modules and programs
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1255Doc/ Documentation sources (LaTeX)
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1256Grammar/ Input for the parser generator
1257Include/ Public header files
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1258LICENSE Licensing information
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1259Lib/ Python library modules
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]1260Mac/ Macintosh specific resources
Neil Schemenauer3f52da52001-01-24 17:49:59[diff] [blame]1261Makefile.pre.in Source from which config.status creates the Makefile.pre
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]1262Misc/ Miscellaneous useful files
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1263Modules/ Implementation of most built-in modules
1264Objects/ Implementation of most built-in object types
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]1265PC/ Files specific to PC ports (DOS, Windows, OS/2)
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1266PCbuild/ Build directory for Microsoft Visual C++
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1267Parser/ The parser and tokenizer and their input handling
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1268Python/ The byte-compiler and interpreter
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1269README The file you're reading now
1270Tools/ Some useful programs written in Python
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]1271pyconfig.h.in Source from which pyconfig.h is created (GNU autoheader output)
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1272configure Configuration shell script (GNU autoconf output)
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]1273configure.in Configuration specification (input for GNU autoconf)
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1274install-sh Shell script used to install files
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1275setup.py Python script used to build extension modules
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1276
1277The following files will (may) be created in the toplevel directory by
1278the configuration and build processes:
1279
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1280Makefile Build rules
Neil Schemenauer3f52da52001-01-24 17:49:59[diff] [blame]1281Makefile.pre Build rules before running Modules/makesetup
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]1282buildno Keeps track of the build number
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]1283config.cache Cache of configuration variables
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]1284pyconfig.h Configuration header
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]1285config.log Log from last configure run
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1286config.status Status from last run of the configure script
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1287getbuildinfo.o Object file from Modules/getbuildinfo.c
1288libpython<version>.a The library archive
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1289python The executable interpreter
1290tags, TAGS Tags files for vi and Emacs
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1291
1292
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]1293That's all, folks!
1294------------------
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1295
1296
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]1297--Guido van Rossum (home page: http://www.python.org/~guido/)