Commit | Line | Data |
---|---|---|
2e493ffe JB |
1 | This is a list of the status of GNU Emacs on various machines and systems. |
2 | Last updated 14 October 1990. | |
3 | ||
4 | Systems: | |
5 | For each type of system, the name of the appropriate s- header file | |
6 | is given. | |
7 | ||
8 | Berkeley 4.1 (s-bsd4-1.h) | |
9 | ||
10 | Works on vaxes. | |
11 | ||
12 | Berkeley 4.2 (s-bsd4-2.h) | |
13 | ||
14 | Works on several machines. | |
15 | ||
16 | Berkeley 4.3 (s-bsd4-3.h) | |
17 | ||
18 | Works, on Vaxes at least. | |
19 | ||
20 | Microport | |
21 | ||
22 | See under "Intel 386". | |
23 | ||
24 | System V rel 0 (s-usg5-0.h) | |
25 | ||
26 | Works, on Vaxes and 3bxxx's. | |
27 | There are some problems in 18.37 due to shortnames/cccp problems: | |
28 | use the emacs 17 cpp if you have it. | |
29 | ||
30 | System V rel 2 (s-usg5-2.h) | |
31 | ||
32 | Works on various machines. | |
33 | On some (maybe all) machines the library -lPW exists and contains | |
34 | a version of `alloca'. On these machines, to use it, put | |
35 | #define HAVE_ALLOCA | |
36 | #define LIB_STANDARD -lPW -lc | |
37 | in the m-...h file for the machine. | |
38 | ||
39 | If you find that the character Meta-DEL makes Emacs crash, | |
40 | find where function init_sys_modes in sysdep.c sets sg.c_cc[VQUIT] | |
41 | and make it store 7 there. I have as yet no evidence of whether | |
42 | this problem, known in HP-UX, exists in other system V versions. | |
43 | ||
44 | System V rel 2.2 (s-usg5-2-2.h) | |
45 | ||
46 | In 5.2.2 AT&T undid, incompatibly, their previous incompatible | |
47 | change to the way the nlist library is called. A different s- file | |
48 | is used to enable the other interface. | |
49 | ||
50 | They call themselves the right choice--can't they choose? | |
51 | ||
52 | Emacs version 18 unexec is currently not working properly | |
53 | on 5.2.2. Nobody knows why yet. A workaround is to define | |
54 | NO_REMAP. It is not yet known whether this applies to all | |
55 | machines running 5.2.2. | |
56 | ||
57 | System V rel 3 (s-usg5-3.h) | |
58 | ||
59 | Some versions of this system support ptys and BSD-style sockets. | |
60 | On such systems, you should define HAVE_PTYS and HAVE_SOCKETS in config.h. | |
61 | ||
62 | If you want to link Emacs with shared libraries, define | |
63 | USG_SHARED_LIBRARIES. | |
64 | ||
65 | System V rel 4 (s-usg5-4.h) | |
66 | ||
67 | Supported, including shared libraries for ELF, but ptys do not | |
68 | work because TIOCGPGRP fails to work on ptys. | |
69 | This failure is probably due to a misunderstanding of the | |
70 | consequences of the POSIX spec: many system designers mistakenly | |
71 | think that POSIX requires this feature to fail. | |
72 | ||
73 | Ultrix (s-bsd4-2.h) | |
74 | ||
75 | DEC's Ultrix OS is essentially Berkeley 4.2. It does not correctly | |
76 | implement certain features of 4.3. | |
77 | ||
78 | Ultrix 3.0 has incompatibilities in its X library if you have the | |
79 | Ultrix version of X (UWS version 2.0). To solve them, you need to | |
80 | prevent XvmsAlloc.o in Xlib from being used. Israel Pinkas says: | |
81 | ||
82 | I added the following lines to config.h after the X defines: | |
83 | ||
84 | #if defined(ultrix) && defined(X11) | |
85 | #define OBJECTS_SYSTEM calloc.o | |
86 | #endif | |
87 | ||
88 | Then I ran the following: | |
89 | ||
90 | ar x /usr/lib/libc.a calloc.o | |
91 | ||
92 | The problem is said to be gone in UWS version 2.1. | |
93 | ||
94 | Uniplus 5.2 (s-unipl5-2.h) | |
95 | ||
96 | Works, on Dual machines at least. | |
97 | ||
98 | VMS (s-vms.h) | |
99 | ||
100 | Works except for certain features (directory listing, dired, | |
101 | sending and receiving mail) that use synchronous subprocesses. | |
102 | We need people to write alternative implementations of these | |
103 | facilities. | |
104 | ||
105 | Note that Emacs for VMS is usually distributed in a special | |
106 | VMS distribution. See the file ../VMSINSTALL for info on moving | |
107 | Unix distributions to VMS, and other VMS-related topics. | |
108 | ||
109 | Xenix | |
110 | ||
111 | Should work in 18.50, but you will need to edit the files | |
112 | etc/Makefile and src/ymakefile | |
113 | (see the comments that mention "Xenix" for what to change.) | |
114 | Compiling Emacs with -O is said not to work. | |
115 | ||
116 | If you want Emacs to work with Smail (installed as /usr/bin/smail) | |
117 | then add the line #define SMAIL to config.h. | |
118 | ||
119 | The file etc/XENIX suggests some useful things to do to Xenix | |
120 | to make the Emacs meta key work. | |
121 | \f | |
122 | Machines: | |
123 | For each type of machine, the names of the m- and s- header files | |
124 | are given. | |
125 | ||
126 | Alliant (m-alliant4.h or m-alliant.h or m-alliant1.h; s-bsd4-2.h) | |
127 | ||
128 | 18.52 works on system version 4. Previous Emacs versions were | |
129 | known to work on previous system versions. | |
130 | ||
131 | Use m-alliant1.h on version 1 of their operating system | |
132 | and use m-alliant.h on version 2 or 3. | |
133 | Use m-alliant4.h on version 4. | |
134 | ||
135 | Altos 3068 (m-altos.h; s-usg5-2.h) | |
136 | ||
137 | 18.52 is said to work, provided you don't compile unexec.c with -O. | |
138 | ||
139 | Amdahl UTS (m-amdahl.h; s-usg5-2-2.h) | |
140 | ||
141 | Small changes for 18.38 were merged in 18.39. It is mostly | |
142 | working, but at last report a bug sometimes causes Emacs to | |
143 | grab very large amounts of memory. No fix or explanation | |
144 | has yet been reported. It may be possible to find this bug | |
145 | if you find which Emacs command it happens within and then | |
146 | run that command with a breakpoint set at malloc. | |
147 | ||
148 | The 5.2u370 compiler is so brain damaged that it is not | |
149 | even worth trying to use it. Success was obtained with the | |
150 | uts native C compiler on uts version 5.2.5. | |
151 | ||
152 | Apollo running Domain (m-apollo.h; s-bsd4-2.h) | |
153 | ||
154 | 18.52 works, to some extent. | |
155 | Code for dumping Emacs has been written, but we cannot | |
156 | distribute it yet. | |
157 | There are reports of bugs in cc -O on this system. | |
158 | ||
159 | In etc/Makefile, don't expect emacsclient and emacsserver to | |
160 | compile. You might want to remove them from your makefile. | |
161 | ||
162 | Supposedly something in dired.c runs into a compiler bug. | |
163 | Paraphrasing the statement should avoid the problem. | |
164 | I havenot yet received word as to the exact statement this is. | |
165 | ||
166 | AT&T 3b2, 3b5, 3b15, 3b20 (m-att3b.h; s-usg5-2-2.h) | |
167 | ||
168 | Emacs will probably not work with certain kernel constants too small. | |
169 | ||
170 | In param.h CDLIMIT should be at least (1L << 12) in order to allow | |
171 | processes to write up to 2 Mbyte files. If it's too small, | |
172 | you can perhaps fix it using the file `ulimit.hack' in this directory. | |
173 | ||
174 | Note that for V.3.1 and later releases, this parameter is configurable | |
175 | by normal means in /etc/master.d/kernel; examine that file for the | |
176 | symbol CDLIMIT or ULIMIT, and raise it by several powers of 2. Then | |
177 | do normal kernel rebuild things via "cd /boot; mkboot -k KERNEL" and so | |
178 | forth. Do not use ulimit.hack if you have V.3.1 or later. | |
179 | ||
180 | In seg.h NSEGP and STACKSEG should be at least 16 and 4 respectively | |
181 | to allow processes with total size of up to 2Mbytes. | |
182 | However, I'm told it is unlikely this would fail to be true. | |
183 | ||
184 | The MAXMEM may also prevent Emacs from running. The file | |
185 | 3B-MAXMEM in this directory explains how to increase MAXMEM. | |
186 | ||
187 | AT&T 7300 or 3b1 (m-7300.h; s-usg5-2-2.h) | |
188 | ||
189 | 18.52 seems to work. If you have strange troubles with dumping Emacs, | |
190 | delete the last few lines from m-7300.h and recompile. These lines | |
191 | are supposed to produce a sharable executable. | |
192 | ||
193 | m-7300.h defines SHORTNAMES because operating system versions | |
194 | older than 3.5 did not support long symbol names. Version 3.5 does | |
195 | support them, so you can remove the #define SHORTNAMES in that version. | |
196 | ||
197 | Bull sps7 (m-sps7.h; s-usg5-2.h) | |
198 | ||
199 | Changes partially merged in version 19, but some fixes are probably required. | |
200 | ||
201 | CCI 5/32, 6/32 | |
202 | ||
203 | See "Tahoe". | |
204 | ||
205 | Celerity (m-celerity.h; s-bsd4-2.h) | |
206 | ||
207 | Version 18.49 works. | |
208 | ||
209 | Clipper (m-clipper.h; ???) | |
210 | ||
211 | Version 19 has support for some brand of clipper system. | |
212 | Note that the Orion 105 is also a clipper, but some system-related | |
213 | parameters are different. | |
214 | ||
215 | Convex (m-convex.h; s-bsd4-3.h) | |
216 | ||
217 | 18.53 is supposed to work. | |
218 | ||
219 | Cubix QBx/386 (m-intel386.h; s-usg5-3.h) | |
220 | ||
221 | Changes merged in 19.1. Systems before 2/A/0 may fail to compile etags.c | |
222 | due to a compiler bug. | |
223 | ||
224 | Cydra 5 (m-cydra5.h; s-usg5-3.h) | |
225 | ||
226 | 18.51 worked in one version of their operating system but stopped | |
227 | working in a newer version. This has not been fixed. | |
228 | ||
229 | DECstation (m-pmax.h; s-bsd4-2.h) | |
230 | ||
231 | Works, as of 18.55. See under Ultrix for problems using X windows | |
232 | on Ultrix. Note that this is a MIPS machine. | |
233 | ||
234 | Delta (m-delta.h; s-usg5-3.h) | |
235 | ||
236 | Motorola Delta boxes running System V/68 release 3. | |
237 | (tested on sys1147 with SVR3V5). Changes merged in 19.1. | |
238 | ||
239 | Dual running System V (m-dual.h; s-usg5-2.h) | |
240 | ||
241 | As of 17.46, this works except for a few changes | |
242 | needed in unexec.c. | |
243 | ||
244 | Dual running Uniplus (m-dual.h; s-unipl5-2.h) | |
245 | ||
246 | Works, as of 17.51. | |
247 | ||
248 | Elxsi 6400 (m-elxsi; s-usg5-2.h) | |
249 | ||
250 | Changes for 12.0 release are in 19.1. | |
251 | Dumping should work now. | |
252 | ||
253 | Encore machine (m-ns16000.h; s-umax.h) | |
254 | ||
255 | This machine bizarrely uses 4.2BSD modified to use the COFF format | |
256 | for object files. Works (as of 18.40). For the APC processor you | |
257 | must enable two lines at the end of s-umax.h, which are commented | |
258 | out in the file as distributed. | |
259 | ||
260 | WARNING: If you compile Emacs with the "-O" compiler switch, you | |
261 | must also use the "-q enter_exits" switch so that all functions have | |
262 | stack frames. Otherwise routines that call `alloca' all lose. | |
263 | ||
264 | A kernel bug in some system versions causes input characters to be lost | |
265 | occasionally. | |
266 | ||
267 | GEC 93 (m-gec93.h; s-usg5-2.h?) | |
268 | ||
269 | Changes are partially merged in version 18, but | |
270 | certainly require more work. | |
271 | ||
272 | Gould Power Node (m-gould.h; s-bsd4-2.h or s-bsd4-3.h) | |
273 | ||
274 | 18.36 worked on versions 1.2 and 2.0 of the operating system. | |
275 | ||
276 | On UTX/32 2.0, use s-bsd4-3.h. | |
277 | ||
278 | On UTX/32 1.2 and UTX/32S 1.0, use s-bsd4-2.h and note that compiling | |
279 | etc/sorted-doc tickles a compiler bug: remove the -g flag to cc in the | |
280 | makefile. | |
281 | ||
282 | UTX/32 1.3 has a bug in the bcopy library routine. Fix it by | |
283 | #undef BSTRING in m-gould.h. | |
284 | ||
285 | Version 19 incorporates support for releases 2.1 and later of UTX/32. | |
286 | A site running a pre-release of 2.1 should #define RELEASE2_1 in config.h. | |
287 | ||
288 | Gould NP1 (m-gould-np1.h; s-bsd4-3.h) | |
289 | ||
290 | Version 19 supposedly works. | |
291 | ||
292 | Honeywell XPS100 (m-xps100.h; s-usg5-2.h) | |
293 | ||
294 | Config file added in version 19. | |
295 | ||
296 | HP 9000 series 200 or 300 (m-hp9000s300.h; s-hpux.h or s-bsd4-3.h) | |
297 | ||
298 | Version 18 works. | |
299 | These machines are 68000-series CPUs running HP-UX | |
300 | (a derivative of sysV with some BSD features) or BSD 4.3 ported by Utah. | |
301 | The choice of s- file determines which system Emacs is built for. | |
302 | ||
303 | Series 200 HPUX runs Emacs only if it has the "HP-UX upgrade". | |
304 | ||
305 | Note that HP has used two incompatible assembler syntaxes, | |
306 | and has recently changed the format of C function frames. | |
307 | src/crt0.c and src/alloca.s have been conditionalised for the new | |
308 | assembler and new function-entry sequence. You may need to define | |
309 | OLD_HP_ASSEMBLER if you are using an older hpux version. If you | |
310 | have an official (bought from HP) series 300 machine you have | |
311 | the new assembler. Kernels that are 5.+ or later have new | |
312 | assembler. A Series 200 that has been upgraded to a 68010 | |
313 | processor and a 5.+ kernel has the new compiler. | |
314 | ||
315 | Define C_SWITCH_MACHINE to be +X to make a version of Emacs that | |
316 | runs on both 68010 and 68020 based hp-ux's. | |
317 | ||
318 | Define HPUX_68010 if you are using the new assembler, for | |
319 | a system that has a 68010 without a 68881. This is to say, | |
320 | a s200 (upgraded) or s310. | |
321 | ||
322 | Define the symbol HPUX_NET if you have the optional network features | |
323 | that include the `netunam' system call. This is refered to as | |
324 | Network Services (NS/9000) in HP literature. | |
325 | ||
326 | HP 9000 series 300 running BSD 4.3 (m-hp300bsd.h; s-bsd4-3.h) | |
327 | ||
328 | Version 18.55 works. | |
329 | ||
330 | HP 9000 series 500: not supported. | |
331 | ||
332 | The series 500 has a seriously incompatible memory architecture | |
333 | which relocates data in memory during execution of a program, | |
334 | and support for it would be difficult to implement. | |
335 | ||
336 | HP 9000 series 800 (Spectrum) (m-hp9000s800.h; s-hpux.h) | |
337 | ||
338 | These files support HP's Precision Architecture machines | |
339 | running HP-UX. It has been moderately tested on the Series | |
340 | 840. | |
341 | ||
342 | High Level Hardware Orion (m-orion.h; s-bsd4-2.h) | |
343 | ||
344 | This is the original microprogrammed hardware. | |
345 | m- file ought to work. | |
346 | ||
347 | High Level Hardware Orion 1/05 (m-orion105.h; s-bsd4-2.h) | |
348 | ||
349 | Changes merged in 18.52. This is the one with the Clipper cpu. | |
350 | Note that systems which lack NFS need LOAD_AVE_TYPE changed to `double'. | |
351 | ||
352 | C compiler has a bug; it loops compiling eval.c. | |
353 | Compile it by hand without optimization. | |
354 | ||
355 | IBM PS/2 (m-ibmps2-aix.h; s-usg5-2-2.h) | |
356 | ||
357 | Changes merged in version 19. You may need to copy | |
358 | /usr/lib/samples/hft/hftctl.c to the Emacs src directory. | |
359 | ||
360 | IBM RS/6000 (m-ibmrs6000.h; s-aix3-1.h) | |
361 | ||
362 | Changes merged in version 19. | |
363 | ||
364 | Note that Emacs can store the pure Lisp data in a shared segment on | |
365 | this machine, but only if you put this command in one of the boot | |
366 | scripts | |
367 | ||
368 | emacs -batch -q | |
369 | ||
370 | so that it is run as root when the machine is booted. This command | |
371 | creates a read-only shared segment owned by root. Otherwise the | |
372 | pure data is not shared. The pure data size is around 100K, which | |
373 | may not be enough to matter for most users of this machine. | |
374 | ||
375 | IBM RT/PC (m-ibmrt.h or m-ibmrt-aix.h; s-bsd4-2.h or s-usg5-2-2.h) | |
376 | ||
377 | 18.52 works on both operating systems. | |
378 | Use s-bsd-4-2.h for the 4.2-like system and s-usg5-2-2.h for AIX. | |
379 | ||
380 | On BSD, if you have trouble, try compiling with a different compiler. | |
381 | ||
382 | On AIX, the file /usr/lib/samples/hft/hftctl.c must be compiled into | |
383 | hftctl.o, with this result left in the src directory (hftctl.c is | |
384 | part of the standard AIX distribution). | |
385 | ||
386 | window.c must not be compiled with -O on AIX. | |
387 | ||
388 | Integrated Solutions `Optimum V' (m-isi-ov.h; s-bsd4-2.h or s-bsd4-3.h) | |
389 | ||
390 | 18.52 said to work on some sort of ISI machine. | |
391 | Version 18.45 worked (running on a Optimum V (VME bus, 68020) | |
392 | BSD 4.2 (3.05e) system). 18.42 is reported to work on | |
393 | a Qbus 68010 system. Has not been tried on `WorkStation' `Cluster | |
394 | Compute Node' `Cluster WorkStation' or `Server Node' (Love the | |
395 | StudLYCaps) | |
396 | ||
397 | Compilation with -O is rumored to break something. | |
398 | ||
399 | On recent system versions, you may need to undefine the macro UMAX | |
400 | in etc/loadst.c and src/getpagesize.h. They stupidly defined this | |
401 | in a system header file, which confuses Emacs (which thinks that UMAX | |
402 | indicates the Umax operating system). | |
403 | ||
404 | Intel 80386 (m-intel386.h or m-is386.h; s-bsd4-2.h, s-usg5-2-2.h, s-usg5-3.h, | |
405 | s-386-ix.h or s-xenix.h) | |
406 | ||
407 | Changes merged in 18.50 for all three operating systems. | |
408 | If you are using Xenix, see notes above under Xenix. | |
409 | ||
410 | Some sysV.3 systems seem to have bugs in `opendir'; | |
411 | for them, alter `config.h' to define NONSYSTEM_DIR_LIBRARY | |
412 | and undefine SYSV_SYSTEM_DIR. | |
413 | ||
414 | If you use optimization on V.3, you may need the option -W2,'-y 0' | |
415 | to prevent certain faulty optimization. | |
416 | ||
417 | There is no consistency in the handling of certain system header files | |
418 | on V.3. | |
419 | ||
420 | Some versions have sys/sioctl.h, and require it in sysdep.c. | |
421 | But some versions do not have sys/sioctl.h. | |
422 | Some versions have been known to need sys/ttold.h, sys/stream.h, | |
423 | and sys/ptem.h included in sysdep.c. | |
424 | Some versions convince sysdep.c to try to use `struct tchars' | |
425 | but define `struct tc' instead; add `#define tchars tc' | |
426 | to config.h to solve this problem. | |
427 | ||
428 | The file m-is386.h is used for an Integrated Solutions 386 machine. | |
429 | It may also be correct for Microport systems. | |
430 | ||
431 | Iris 2500 (m-irist.h; s-iris3-5.h or s-iris3-6.h) | |
432 | ||
433 | Version 18 said to work; use s-irist3-5.h for system version 2.5 | |
434 | and s-iris3-6.h for system version 3.6. | |
435 | ||
436 | Iris 2500 Turbo (m-irist.h; s-iris3-5.h or s-iris3-6.h) | |
437 | ||
438 | 18.49 works. Use s-iris3-6.h for system versions 3.6 and up. | |
439 | Note that the 3030 is the same machine as this. | |
440 | ||
441 | Iris 4D (m-iris4d.h; s-iris3-6.h or s-irix3-3.h) | |
442 | ||
443 | Changes for 18.49 are merged in 18.50. 18.52 seems to work. | |
444 | The system version called Irix 3.3 is newer than what was called 3.6. | |
445 | ||
446 | Note that the compilation of etc/loadst.c requires editing etc/Makefile. | |
447 | See the comment in that file. | |
448 | ||
449 | Someone says that with the yellow pages you must change two definitions | |
450 | in m-iris4d.h as follows: | |
451 | ||
452 | #define LIBS_MACHINE -lsun -lbsd -lPW -lmld | |
453 | #define C_SWITCH_MACHINE -I/usr/include/sun -I/usr/include/bsd | |
454 | ||
455 | Macintosh | |
456 | ||
457 | We are boycotting Apple because of Apple's efforts to take away | |
458 | our freedom to write compatible imitations of existing software. | |
459 | If you value your freedom to write such programs, we urge you | |
460 | not to buy from Apple, not to develop software for Apple, and | |
461 | certainly not to accept a job with Apple. | |
462 | ||
463 | See the file APPLE in this directory for more information. | |
464 | ||
465 | Masscomp (m-masscomp.h; s-rtu.h) | |
466 | ||
467 | 18.36 worked on a 5500DP running RTU v3.1a and compiler version 3.2 | |
468 | with minor fixes that are included in 18.37. However, bizarre behavior | |
469 | was reported for 18.36 on a Masscomp (model and version unknown but probably | |
470 | a 68020 system). The report sounds like a compiler bug. | |
471 | ||
472 | A compiler bug affecting statements like | |
473 | unsigned char k; unsigned char *p;... x = p[k]; | |
474 | has been reported for "C version 1.2 under RTU 3.1". We do not wish | |
475 | to take the time to install the numerous workarounds required to | |
476 | compensate for this bug; go complain to Masscomp. | |
477 | ||
478 | For RTU version 3.1, define FIRST_PTY_LETTER to be 'p' in s-rtu.h | |
479 | (or #undef and redefine it in config.h) so that ptys will be used. | |
480 | ||
481 | GNU Emacs is said to have no chance of compiling on RTU versions | |
482 | prior to v3.0. | |
483 | ||
484 | Megatest (m-mega68.h; s-bsd4-2.h) | |
485 | ||
486 | Emacs 15 worked; do not have any reports about Emacs 16 or 17 | |
487 | but any new bugs are probably not difficult. | |
488 | ||
489 | Mips (m-mips.h or m-mips4.h; s-usg5-2-2.h) | |
490 | ||
491 | Changes merged in 18.39. | |
492 | ||
493 | Use m-mips4.h for RISCOS version 4. | |
494 | ||
495 | Motorola Delta (m-delta.h; s-usg5-3.h) | |
496 | ||
497 | m- file added in version 18.56. | |
498 | ||
499 | National Semiconductor 32000 (m-ns32000.h; s-usg5-2.h) | |
500 | ||
501 | This is for a complete machine from National Semiconductor, | |
502 | running Genix. Changes merged in version 19. | |
503 | ||
504 | NCR Tower 32 (m-tower32.h; s-usg5-2.h) | |
505 | ||
506 | Works as of 17.56. | |
507 | ||
508 | Nixdorf Targon 31 (m-targon31.h; s-usg5-2-2.h) | |
509 | ||
510 | m- file for version 17 is included in 18 | |
511 | but whether it works is not known. | |
512 | src/unexec.c bombs if compiled with -O. | |
513 | Note that the "Targon 35" is really a Pyramid. | |
514 | ||
515 | Nu (TI or LMI) (m-nu.h; s-usg5-2.h) | |
516 | ||
517 | Version 18 is believed to work. | |
518 | ||
519 | Plexus (m-plexus.h; s-usg5-2.h) | |
520 | ||
521 | Works as of 17.56. | |
522 | ||
523 | Pmax (DEC Mips) (m-pmax.h; s-bsd4-2.h) | |
524 | ||
525 | See under DECstation, above. | |
526 | ||
527 | Prime EXL (m-intel386.h; s-usg5-3.h) | |
528 | ||
529 | Minor changes merged in 19.1. | |
530 | ||
531 | Pyramid (m-pyramid.h; s-bsd4-2.h) | |
532 | ||
533 | In OSx 4.0, it seems necessary to add the following two lines | |
534 | to m-pyramid.h: | |
535 | #define _longjmp longjmp | |
536 | #define _setjmp setjmp | |
537 | ||
538 | In Pyramid system 2.5 there has been a compiler bug making | |
539 | Emacs crash just after screen-splitting with Qnil containing 0. | |
540 | A compiler that fixes this is Pyramid customer number 8494, | |
541 | internal number 1923. | |
542 | ||
543 | Some versions of the pyramid compiler get fatal | |
544 | errors when the -gx compiler switch is used; if this | |
545 | happens to you, change src/m-pyramid.h to define | |
546 | C_DEBUG_SWITCH with an empty definition. | |
547 | ||
548 | Some old system versions may require you to define PYRAMID_OLD | |
549 | in when alloca.s is preprocessed, in order to define _longjmp and _setjmp. | |
550 | ||
551 | Sequent Balance (m-sequent.h; s-bsd4-2.h, or s-bsd4-3.h on newer systems) | |
552 | ||
553 | Emacs 18.51 should work on system version 3.0. 18.52 is said to work. | |
554 | Delete some lines at the end of m-sequent.h for earlier system versions. | |
555 | ||
556 | Sequent Symmetry (m-symmetry.h; s-bsd4-3.h) | |
557 | ||
558 | Emacs 19 should work. | |
559 | ||
560 | SONY News (m-news.h; s-bsd4-2.h, or s-bsd4-3.h for system release 3) | |
561 | ||
562 | 18.52 should work. | |
563 | ||
564 | SONY News 3000 series (RISC NEWS) (m-news-risc.h; s-bsd4-3.h) | |
565 | ||
566 | Works, as of 18.56. Note that this is a MIPS architecture machine. | |
567 | ||
568 | Stardent 1500 or 3000 | |
569 | ||
570 | See Titan. | |
571 | ||
572 | Stride (m-stride.h; s-usg5-2.h) | |
573 | ||
574 | Works (most recent news for 18.30) on their release 2.0. | |
575 | For release 2.2, see the end of m-stride.h. | |
576 | It may be possible to run on their V.1 system but changes | |
577 | in the s- file would be needed. | |
578 | ||
579 | Sun 1, 2 and 3 (m-sun1.h, m-sun2.h, m-sun3.h; s-bsd4-2.h or s-sunos4.h) | |
580 | ||
581 | There are three m- files for different versions of SunOS. | |
582 | All are derived from Berkeley 4.2. Emacs 17 has run on all of them. | |
583 | Whether you should use m-sun1.h, m-sun2.h or m-sun3.h depends | |
584 | on the VERSION OF THE OPERATING SYSTEM you have. | |
585 | You will need to use m-sun3.h on Sun 2's running SunOS release 3. | |
586 | ||
587 | For SunOS release 4 on a Sun 3, use m-sun3.h and s-sunos4.h. | |
588 | See the file etc/SUNBUG for how to solve problems caused by bugs in | |
589 | the "export" version of SunOS 4. | |
590 | ||
591 | If you have trouble using open-network-stream, get the | |
592 | distribution of `bind' (the BSD name-server), build libresolv.a, | |
593 | and link Emacs with -lresolv. This problem is due to obsolete | |
594 | software in the nonshared standard library. | |
595 | ||
596 | If you want to use SunWindows, define HAVE_SUN_WINDOWS | |
597 | in config.h to enable a special interface called `emacstool'. | |
598 | The definition must *precede* the #include "m-sun3.h". | |
599 | System version 3.2 is required for this facility to work. | |
600 | ||
601 | We recommend that you instead use the X window system, which | |
602 | has technical advantages, is an industry standard, and is also | |
603 | free software. | |
604 | ||
605 | If you are compiling for X windows, and the X window library was | |
606 | compiled to use the 68881, then you must edit config.h according | |
607 | the comments at the end of m-sun3.h. | |
608 | ||
609 | Note that Emacs on a Sun is not really as big as it looks. | |
610 | As dumped, it includes around 200k of zeros between the | |
611 | original text section and the original data section | |
612 | (now remapped as part of the text). These are never | |
613 | swapped in. | |
614 | ||
615 | To build a single Emacs that will run on Sun 2 and Sun 3 | |
616 | HARDWARE, just build it on the Sun 2. | |
617 | ||
618 | Sun 4 (m-sparc.h; s-bsd4-2.h or s-sunos4.h) | |
619 | ||
620 | Changes merged in 18.50. Some people say optimizing compilation | |
621 | does not work; some say that -O2 (whatever that is) works | |
622 | perhaps with a small change. | |
623 | ||
624 | Use s-sunos4.h for operating system version 4. | |
625 | See the file etc/SUNBUG for how to solve problems caused by bugs in | |
626 | the "export" version of SunOS 4. | |
627 | ||
628 | Sun Roadrunner (m-sun386.h; s-sunos4.h) | |
629 | ||
630 | Changes merged in 18.51. | |
631 | ||
632 | Tahoe (m-tahoe.h; s-bsd4-2.h or s-bsd4-3.h) | |
633 | ||
634 | 18.52 known to work on some Tahoes, but a compiler bug intervenes | |
635 | on others. Some Emacs versions have worked in Unisys 1r4 | |
636 | (not in 1r3) and CCI I.21. | |
637 | ||
638 | If you have trouble compiling etc/loadst.c, turn off the definition | |
639 | of DKSTAT_HEADER_FILE in m-tahoe.h. | |
640 | ||
641 | Tektronix 16000 box (6130?) (m-ns16000.h; s-bsd4-2.h) | |
642 | ||
643 | Emacs 17.61 worked. | |
644 | ||
645 | Tektronix 4300 (m-tex4300.h; s-bsd4-3.h) | |
646 | ||
647 | Emacs 18.51 should work. | |
648 | ||
649 | Titan P2 or P3 (m-titan.h; s-usg5-3.h) | |
650 | ||
651 | Changes probably merged in version 19. | |
652 | ||
653 | Ustation E30 (SS5E) (m-ustation.h; s-unipl5-2.h). | |
654 | ||
655 | Changes merged in 18.52; don't know whether they work. | |
656 | ||
657 | Vax running Berkeley Unix (m-vax.h; s-bsd4-1.h, s-bsd4-2.h or s-bsd4-3.h) | |
658 | ||
659 | Works. | |
660 | ||
661 | Vax running Ultrix (m-vax.h; s-bsd4-2.h) | |
662 | ||
663 | Works. See under Ultrix for problems using X windows on Ultrix. | |
664 | ||
665 | Vax running System V rel 2 (m-vax.h; s-usg5-2.h) | |
666 | ||
667 | 18.27 Works. | |
668 | ||
669 | Vax running System V rel 0 (m-vax.h; s-usg5-0.h) | |
670 | ||
671 | Works as of 18.36. | |
672 | ||
673 | Vax running VMS (m-vax.h; s-vms.h but on VMS they are m_vax.h and s_vms.h) | |
674 | ||
675 | 18.36 believed to work. Addition of features is necessary to make | |
676 | this Emacs version more usable. | |
677 | ||
678 | Whitechapel MG1 (m-mg1.h, s-?) | |
679 | ||
680 | May work. Supposedly no changes were needed except the m- file. | |
681 | I do not know what Unix version runs on them. | |
682 | ||
683 | Wicat (m-wicat.h; s-usg5-2.h) | |
684 | ||
685 | Changes merged as of 18.6; whether they work is unknown. | |
686 | See comments in m-wicat.h for things you should change | |
687 | depending on the system and compiler version you have. | |
688 | \f | |
689 | Local variables: | |
690 | mode: text | |
691 | fill-prefix: " " | |
692 | End: |