3 At the time of this release (SXEmacs 22.1.15), SXEmacs has the
4 following idiosyncrasies:
9 ** User init file (C-h v user-init-file)
11 SXEmacs looks for user init files in `user-init-directory'. The
12 preferred directory is: ${XDG_CONFIG_HOME}/sxemacs but it can fall
13 back to the old ~/.sxemacs directory.
17 ${XDG_CONFIG_HOME}/sxemacs
18 ${HOME}/.config/sxemacs # if $XDG_CONFIG_HOME is not set
19 ${HOME}/.sxemacs # if other dirs don't exist
21 You can also force the use of ~/.sxemacs regardless of the existence
22 of the XDG dir/var by setting $SXE_USE_LEGACY environment variable to
25 If you're coming from XEmacs, symlinking your old ~/.xemacs directory
26 to a SXEmacs location should be enough to get you up and running:
28 $ ln -svfn ${HOME}/.xemacs ${XDG_CONFIG_HOME}/sxemacs
30 BTW, unlike XEmacs, SXEmacs doesn't attempt to "migrate" your old init
31 file or Gnu/Emacs .emacs file.
35 The default location that SXEmacs searches for packages is
36 `$prefix/share/sxemacs/'. The same as for the user-init-file, a
37 symlink is all you need to get up and running.
39 $ ln -s /usr/local/lib/xemacs /usr/local/share/sxemacs RET
47 *** FFI is not included with your distro
49 Sadly, some Linux distributions (hello Fedora) don't ship a libffi
50 package, and their GCC does NOT include libffi or FFI headers either.
51 In this instance you have 2 options...
53 1) Get the standalone package of libffi at
54 <http://sourceware.org/libffi/>.
56 2) Compile your own GCC from source, making sure you enable the java
57 compiler. Enabling java in your GCC build is the only way to get
60 Obviously, option #1 there is the easiest and quickest path to
61 FFI-enabled SXEmacsen, and it is the option that we recommend.
63 Oh, and please nag your distro to have FFI included by default.
65 *** FFI is included in your GCC but you see missing header errors
67 Often libffi headers aren't completely installed. If you are getting
68 errors in effi.c that seem to be hinged from something like...
70 /usr/include/ffi.h:63:23: ffitarget.h: No such file or directory
72 You need to find `ffitarget.h' and put it in the same directory as
73 your `ffi.h'. Your libffi came with GCC, so you'll find it within
76 $ dirname $(gcc -print-libgcc-file-name)
77 /usr/lib/gcc-lib/i586-pc-linux-gnu/3.3.1
79 Using that example, ffitarget.h would be in...
81 /usr/lib/gcc-lib/i586-pc-linux-gnu/3.3.1/libffi/
83 Just copy or symlink the ffitarget.h there to /usr/include
85 *** FFI on SELinux enabled machines
87 If you are running with SELinux enabled and configure fails with
88 messages like the following in `config.log'...
90 error while loading shared libraries: /usr/local/lib/libffi.so.1:
91 cannot restore segment prot after reloc: Permission denied
93 You need to correct the default security context for `libffi.so'.
95 $ chcon -t textrel_shlib_t /usr/local/lib/libffi.so
99 The autoconf tests for PostgreSQL support have changed. SXEmacs'
100 configure script now uses `pg_config' to determine whether or not to
101 enable PostgreSQL. Because of this you may have to set you $PATH
102 environment to include the pgsql bin directory. It is normally
103 `/usr/local/pgsql/bin/'. Another popular directory on Solaris 9 is
104 `/opt/crw/postgresql/bin/'. Check with your site administrator.
106 Bash users can do it like this...
108 export PATH=/usr/local/pgsql/bin:$PATH
110 *** Solaris 9 with 64-bit PostgreSQL
112 There has also been a report that on Solaris 9 you may also need to
113 configure with `--with-cflags='-mcpu=ultrasparc -m64''. Apparently
114 GCC on Solaris 9 defaults to building 32-bit, so you lose if you have
117 ** 64-bit test suite failure
119 We have had a couple of reports of the test suite failing on 64-bit
120 systems. The error is like this (or similar)...
122 Testing /usr/src/sxemacs/modules/ase/ase-heap-tests.el...
123 Loading ase_heap v0.0.0 (SXEmacs module: ase-heap)
124 Loaded module ase_heap v0.0.0 (SXEmacs module: ase-heap)Fatal error: assertion failed, file alloc.c, line 298, block != (void*)0xCAFEBABEDEADBEEF
125 make[3]: *** [check-am] Aborted
127 At this point we are not too sure exactly what the issue is. It looks
128 like it might be a bug in the malloc or free code of the libc. We do
129 know that not all 64-bit systems are affected, so far, only Fedora
130 Core 7, and Gentoo on x86_64.
132 One user has reported that using `-O1' in CFLAGS prevents it.
134 But even with this test failure, SXEmacs still runs and operates
135 without incident. In fact, the failure can't be reproduced when
136 running the test suite interactively. With that in mind, it should be
137 safe to install if you see this failure.
139 We'll endeavour to get to the bottom of this one ASAP, if you think
140 you can help, let us know.
142 ** m4, libtool, autoconf, automake, and whatnot
144 SXEmacs tries to cope with any combination of versions of the above
145 programs. However, there is one lower bound, autoconf 2.60, and
146 unfortunately this has an impact on the other parts of the build
149 To cut it really short, here is the minimum known-to-work combination:
150 - autoconf 2.62, automake 1.9.6, libtool 1.5.22, m4 1.4.6
152 In general we support (as of April, 2010):
153 - autoconf >= 2.62, including current git versions
154 - automake 1.9.6, 1.10, 1.10a, 1.11.1, and current git versions
155 - libtool 1.5.N with N >= 22, libtool >= 2.1a (current CVS version)
156 - m4 1.4.M with M >= 6 plus current git versions
158 Note that many libtool packages shipped with the distros (OpenSuSE,
159 Debian, just to name two) are _broken_. Make sure you compile
160 your own libtool in case you want to rerun autogen.sh or bootstrap
161 the build chain, and double check that you use --enable-ltdl-install
164 If you are on a platform that has its own _non_gnu libtool (like OS/X
165 Leopard) add --program-prefix=g to your gnu libtool configure so it
166 installs as glibtool and doesn't clobber your other one.
168 Sometimes it helps just to copy over the libtool script manually:
169 cp -a $(type -p libtool) ${top_builddir}
171 *** ylwrap fails with sed errors
173 Some versions of the ylwrap script provided by autotools uses commas
174 as separators in sed commands. As such if your build path uses commas
175 the ylwrap will fail.
177 Sample message (where the build path was /Users/njsf/Projects/SXEmacs/nsx-up/,,mac):
179 /Users/njsf/Projects/SXEmacs/nsx-up/,,mac/lib-src/make-docfile --modname cl-loop -E cl-loop.doc.c ../../../modules/cl/cl-loop.c
180 /bin/sh ../../../ylwrap ../../../modules/cl/cl-loop-parser.y y.tab.c cl-loop-parser.c y.tab.h cl-loop-parser.h y.output
181 cl-loop-parser.output -- bison -y -d
182 sed: 1: "s,/Users/njsf/Projects/ ...": bad flag in substitute command: 'm'
183 sed: 1: "s,/Users/njsf/Projects/ ...": bad flag in substitute command: 'm'
185 The workaround is to use a path without commas in it.
188 *** Missing libltdl.la (Solaris 2.8)
190 We've had a report that missing libtool on Solaris 2.8 isn't detected
191 and so the included libtool still isn't used. If you see an error
192 about a missing libltdl.la all you need to do is configure SXEmacs
199 *** configure on FreeBSD, NetBSD, OpenBSD, etc.
201 Building SXEmacs on *BSD as far as we know requires the GNU Bourne
202 Again SHell (bash) versions 3 or 4.
204 bash is available for all tier 1 architectures as a binary package and
205 and for tier 2/3 as a port.
207 To run configure successfully...
209 CONFIG_SHELL=/path/to/bash $CONFIG_SHELL configure [option, ...]
211 *** configure on FreeBSD
213 Turning on the use of libssp and -fstack-protector from configure
214 ( --with-error-checking=stack ) will result in a broken build.
216 Do not, under any circumstances, add -fstack-protector to CFLAGS, even
217 independently of the stack error checking option.
219 ** bdwgc and gcc and code optimisation
221 There are some weird optimisation issues with the Boehm-Demers-Weiser
222 garbage collector (hereafter BDWGC) and the GCC C compilers of the 2 and
223 3 series. The build will crash like this:
225 Loading build-autoloads.el...
226 Loading loadup-el.el...
227 Loading loadup.el...make[3]: *** [auto-autoloads.el] Segmentation fault
229 make[3]: Leaving directory
231 The C backtrace will look like:
233 #0 0xbff9a2f0 in ?? ()
234 #1 0xb7eaf7d6 in GC_invoke_finalizers () at finalize.c:787
235 #2 0xb7eaf8ed in GC_notify_or_invoke_finalizers () at finalize.c:844
236 #3 0xb7eb2c8c in GC_generic_malloc (lb=32, k=0) at malloc.c:190
238 If this is true for you, you may want to try another optimisation level:
240 ./configure CFLAGS="-g -O2"
242 If this still does not work out either dispense with BDWGC support or
243 use a recent C compiler. ATTOW, all GCC 4.x compilers (including SVN)
248 ENT is basically a conglomerate of internally and externally implemented
249 arithmetics. Hence it supports a number of libraries, some of which
250 overlap in their functionality, some others do not but then break at the
253 One of the most likely problems is the GMP vs. MPFR issue. In past
254 times, mpfr (a multiprecision library for floats with exact rounding
255 facilities) has been a part of the GMP distribution. Later on, mpfr got
256 separated from it and has been developed independently while the version
257 of mpfr which ships with GMP stayed the same. Now that scenario is
260 Inattentive distributions (like Fedora) still deliver packages of GMP
261 with the old'n'incompatible mpfr library. SXEmacs will disable the MPFR
262 support on such systems by default (at configure time). However, if you
263 install a supported version of mpfr in parallel to the packaged ones on
264 such a system SXEmacs autodetection correctly reports that a sane
265 version of mpfr is available and enables it. Nonetheless, the according
266 build may fail (or the build may even succeed but calling the binary may
269 number-mpfr.o: In function `ent_lt_BIGFR_T':
270 /home/martin/src/edit/sxemacs-main/src/number-mpfr.c:661: undefined
271 reference to `mpfr_less_p'
272 number-mpfr.o: In function `ent_gt_BIGFR_T':
273 /home/martin/src/edit/sxemacs-main/src/number-mpfr.c:671: undefined
274 reference to `mpfr_greater_p'
277 Especially note that we _only_ support the standalone version of MPFR,
278 and not the one distributed with GMP.
283 Badger your distributor and demand separate packages for GMP and
286 Remove the GMP package and install your own build -- available at
287 http://swox.com/gmp -- afterwards install your own build of mpfr (the
288 one from http://www.mpfr.org)
290 Reconfigure and rebuild SXEmacs afterwards.
293 ** Build fails because of missing makeinfo
295 Install the GNU texinfo package on your system. You'll need at least
299 ** MacOS X warns of a crash during configure
301 This is normal, as one of the tests made during configure (for the
302 realpath call correctness) induces as crash.
304 If you are developing SXEmacs and will do lots of runs of configure
305 and that dialog annoys you, consider issuing:
307 # Disable crash reporting
308 sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.ReportCrash.Root.plist
309 # Redo last configure
310 ./config.status --recheck
311 # Enable crash reporting
312 sudo launchctl load -w /System/Library/LaunchDaemons/com.apple.ReportCrash.Root.plist
314 Another alternative (not recommended) is to launch
316 /Developer/Applications/Utilities/CrashReporterPrefs
318 and configure the mode to server, but you will loose notifications of
319 crashes on all applications.
321 In order to give SXEmacs developers with good diagnosis information it
322 is recommended the mode be Developer.
326 SXEmacs does build and run on OpenIndiana (151a) but you will need to
327 install a few files/packages beforehand. Namely...
329 Common Name OpenIndiana Package Name
333 automake automake-110
335 libtool libtool (also install libltdl)
342 Yes, you read that right... to get pkg-config you must install the
343 "gettext" package. :-)
345 In that list, `bison', `gmp', and `mpfr' are not critical, but you
346 will get extra functionality in your SXEmacs if you have them.
348 *** automake additional instructions for OpenIndiana
350 When you install the automake-110 OpenIndiana package it won't set up
351 the symlinks to /usr/bin/automake or /usr/bin/aclocal. Fix that
354 sudo ln -sv automake-1.10 /usr/bin/automake
355 sudo ln -sv aclocal-1.10 /usr/bin/aclocal
357 *** Running SXEmacs configure on OpenIndiana
359 There's one more quirk with OpenIndiana when you try to run SXEmacs'
360 configure... you MUST set $CONFIG_SHELL
362 CONFIG_SHELL=/bin/bash ../configure [opts]
365 ** make does not stop on subdirectory build failure
367 Due to a bug in the make argument parsing in code generated by
368 autoconf it is possible for make not to stop when a subdirectory fails.
370 This failure occurs for instance when the make command line has a variable
371 assignment which has a value with a - and k. Example:
373 make CFLAGS="-Wall -fpacked -fpedantic" build-report
379 We have identified 2 packages so far that don't work "out of the box"
380 with SXEmacs. In both of these the problem is with parsing version
381 information. Patches have been sent to the appropriate maintainer to
382 fix the problem and are included here in case the packages haven't
383 been updated by the time you install SXEmacs.
385 Update: The EFS, and Dired XEmacs packages that are currently
386 available from the "Pre-Releases" area of XEmacs package mirrors are
387 both now compatible with SXEmacs and do not need the patches mentioned
390 * Problems with running SXEmacs
391 ===============================
395 *** ffi-wand.el refuses to load.
397 Can't load library `libMagickWand': libgomp.so.1: shared object cannot be
400 If you get that error when trying to load ffi-wand, it is because you
401 have a ImageMagick that is using OpenMP (currently only svn HEAD). To
402 fix this you will need to rebuild ImageMagick, making sure that you
403 configure it using --disable-openmp.
405 See: <http://issues.sxemacs.org/show_bug.cgi?id=104>
407 ** Multimedia Goodness
409 *** SXEmacs hangs or crashes during (init-asynchronousity).
411 This is most likely a known effect (we do not want to call it bug,
412 since there is no definite location) with certain (g)libc and kernel
413 combinations under Linux. If it crashes analyse the core file, it
414 should look like this:
416 #0 0x4014ebc4 in __sigsuspend (set=0xbffffbb4) at
417 ../sysdeps/unix/sysv/linux/sigsuspend.c:48
418 #1 0x40101b34 in __pthread_wait_for_restart_signal (self=0x401116e0) at
420 #2 0x40101138 in __pthread_create_2_1 (thread=0x206f8dc, attr=0xbffffc58,
421 start_routine=0x20043ac <console>, arg=0xbffffd88) at restart.h:26
423 A definite fault-prone setup is using kernel 2.6.x in conjunction with
426 *** SXEmacs hangs or crashes before it ought to playback sound.
428 As before, this is most likely a suspicious (g)libc/kernel
431 *** SXEmacs dumps core when using the ALSA audio device
433 This has been reported to happen with old ALSA libraries (1.0.3 to be
434 precise). At the moment it is uncertain at which version these
435 problems disappear (no developer wants to downgrade to a non-working
436 ALSA :D). We highly suggest to use the version 1.0.10 and above, or
439 *** SXEmacs in async mode does not play simultaneous sounds with ALSA
441 This is due to missing (hardware-)mixing capabilities of your
442 soundcard. There is a user-space plugin called dmix, which can
443 effectively circumvent this issue.
445 *** SXEmacs crashes when using state sentinels with asynchronous sounds
447 This is a known bug (#13 in our bug database). At the moment the only
448 advise we can give is: do not use sentinels before 22.1.7.
449 Also see our bug database at http://issues.sxemacs.org
451 *** make-media-stream seems to recognise any file as valid audio
453 This is a known issue with fully-featured ffmpeg builds. The current
454 code in SXEmacs blindly relies on FFmpeg when it reports a file or
455 string as valid audio. There is no way to double-check that at the
456 moment. However, you can perform the additional check yourself if
457 you have taglib installed. Use the included ffi-taglib.el.
460 * Original XEmacs PROBLEMS File
461 ===============================
463 The original XEmacs PROBLEMS file may be found in the SXEmacs
464 source distribution as PROBLEMS.XEmacs - while many issues mentioned
465 have since been fixed, it is preserved for posterity.