Mercurial > hg > xemacs-beta
comparison PROBLEMS @ 1245:6981ff72175a
[xemacs-hg @ 2003-01-31 12:14:00 by stephent]
Motif PROBLEMS <87znphy0uo.fsf@tleepslib.sk.tsukuba.ac.jp>
| author | stephent |
|---|---|
| date | Fri, 31 Jan 2003 12:14:00 +0000 |
| parents | 0350fea81e65 |
| children | b531bf8658e9 |
comparison
equal
deleted
inserted
replaced
| 1244:78c3f60ba757 | 1245:6981ff72175a |
|---|---|
| 38 | 38 |
| 39 * Problems with building XEmacs | 39 * Problems with building XEmacs |
| 40 =============================== | 40 =============================== |
| 41 | 41 |
| 42 ** General | 42 ** General |
| 43 | |
| 43 Much general information is in INSTALL. If it's covered in | 44 Much general information is in INSTALL. If it's covered in |
| 44 INSTALL, we don't repeat it here. | 45 INSTALL, we don't repeat it here. |
| 45 | 46 |
| 46 *** How do I configure to get the buffer tabs/progress bars? | 47 *** How do I configure to get the buffer tabs/progress bars? |
| 47 | 48 |
| 168 more modern version of libz might be installed. This will cause problems | 169 more modern version of libz might be installed. This will cause problems |
| 169 when attempting to link against libMagick. The fix is to remove the old | 170 when attempting to link against libMagick. The fix is to remove the old |
| 170 libz.a in the X11 binary directory. | 171 libz.a in the X11 binary directory. |
| 171 | 172 |
| 172 | 173 |
| 173 ** AIX | 174 ** Motif |
| 174 *** IBM compiler fails: "The character # is not a valid C source character." | 175 |
| 175 | 176 Motif is the X11 version of the Gnus torture test: if there's a way to |
| 176 Most recently observed in 21.5.9, due to USE_KKCC ifdefs (they just | 177 crash, Motif will find it. With the open source release of Motif, it |
| 177 happen to tickle the implementation). | 178 seems like a good idea to collect all Motif-related issues in one |
| 178 | 179 place. |
| 179 Valdis Kletnieks says: | 180 |
| 180 | 181 You should also look in your OS's section, as it may not be Motif's |
| 181 The problem is that IBM defines a *MACRO* called 'memcpy', and we | 182 fault. |
| 182 have stuck a #ifdef/#endif inside the macro call. As a workaround, | 183 |
| 183 try adding '-U__STR__' to your CFLAGS - this will cause string.h to | 184 *** XEmacs crashes on exit (#1). |
| 184 not do a #define for strcpy() to __strcpy() - it uses this for | 185 |
| 185 automatic inlining support. | 186 The backtrace is something like: |
| 186 | 187 |
| 187 (For the record, the same issue affects a number of other functions | 188 (gdb) where |
| 188 defined in string.h - basically anything the compiler knows how to | 189 #0 0xfeb9a480 in _libc_kill () from /usr/lib/libc.so.1 |
| 189 inline.) | 190 #1 0x000b0388 in fatal_error_signal () |
| 190 | 191 #2 <signal handler called> |
| 191 *** On AIX 4.3, you must specify --with-dialogs=athena with configure | 192 #3 YowIter (ht=0xb, id=0x0, v=0x74682074, client=0x47e3c0) |
| 192 | 193 at ImageCache.c:1159 |
| 193 *** The libXt shipped with AIX 4.3 up to 4.3.2 is broken. This causes | 194 #4 0xff26cc5c in _LTHashTableForEachItem (ht=0x4725e8, |
| 194 xemacs -nw to fail in various ways. The official APAR is this: | 195 iter=0xff26dda0 <YowIter>, ClientData=0x47e3c0) at Hash.c:671 |
| 195 | 196 #5 0xff2a4664 in destroy (w=0x496550) at Screen.c:352 |
| 196 APAR NUMBER: <IX89470> RESOLVED AS: PROGRAM ERROR | 197 #6 0xfef92118 in Phase2Destroy () from /usr/openwin/lib/libXt.so.4 |
| 197 | 198 #7 0xfef91940 in Recursive () from /usr/openwin/lib/libXt.so.4 |
| 198 ABSTRACT: | 199 #8 0xfef91e44 in XtPhase2Destroy () from /usr/openwin/lib/libXt.so.4 |
| 199 <IX89470>: LIBXT.A INCORRECT HANDLING OF EXCEPTIONS IN XTAPPADDINPUT | 200 #9 0xfef91ae8 in _XtDoPhase2Destroy () from /usr/openwin/lib/libXt.so.4 |
| 200 | 201 #10 0xfef918cc in XtDestroyWidget () from /usr/openwin/lib/libXt.so.4 |
| 201 The solution is to install X11.base.lib at version >=4.3.2.5. | 202 #11 0xfef91438 in CloseDisplay () from /usr/openwin/lib/libXt.so.4 |
| 202 | 203 #12 0xfef91394 in XtCloseDisplay () from /usr/openwin/lib/libXt.so.4 |
| 203 *** On AIX, you get this compiler error message: | 204 #13 0x0025b8b0 in x_delete_device () |
| 204 | 205 #14 0x000940b0 in delete_device_internal () |
| 205 Processing include file ./XMenuInt.h | 206 #15 0x000806a0 in delete_console_internal () |
| 206 1501-106: (S) Include file X11/Xlib.h not found. | 207 |
| 207 | 208 This is known to happen with Lesstif version 0.93.36. Similar |
| 208 This means your system was installed with only the X11 runtime i.d | 209 backtraces have also been observed on HP/UX and Solaris. There is a |
| 209 libraries. You have to find your sipo (bootable tape) and install | 210 patch for Lesstif. (This is not a solution; it just stops the crash. |
| 210 X11Dev... with smit. | 211 It may or may not be harmless, but "it works for the author".) |
| 211 | 212 |
| 212 *** On AIX 4.1.2, linker error messages such as | 213 Note that this backtrace looks a lot like the one in the next item. |
| 213 ld: 0711-212 SEVERE ERROR: Symbol .__quous, found in the global symbol table | 214 However, this one is invulnerable to the Solaris patches mentioned there. |
| 214 of archive /usr/lib/libIM.a, was not defined in archive member shr.o. | 215 |
| 215 | 216 Frank McIngvale <frankm@hiwaay.net> says: |
| 216 This is a problem in libIM.a. You can work around it by executing | 217 |
| 217 these shell commands in the src subdirectory of the directory where | 218 Ok, 0.93.34 works, and I tracked down the crash to a section |
| 218 you build Emacs: | 219 marked "experimental" in 0.93.36. Patch attached, "works for me". |
| 219 | 220 |
| 220 cp /usr/lib/libIM.a . | 221 diff -u -r lesstif-0.93.36/lib/Xm/ImageCache.c lesstif-0.93.36-mod/lib/Xm/ImageCache.c |
| 221 chmod 664 libIM.a | 222 --- lesstif-0.93.36/lib/Xm/ImageCache.c 2002-08-05 14:53:24.000000000 -0500 |
| 222 ranlib libIM.a | 223 +++ lesstif-0.93.36-mod/lib/Xm/ImageCache.c 2002-11-11 11:13:12.000000000 -0600 |
| 223 | 224 @@ -1166,5 +1166,4 @@ |
| 224 Then change -lIM to ./libIM.a in the command to link temacs (in | 225 DEBUGOUT(_LtDebug0(__FILE__, NULL, "_LtImageCacheScreenDestroy (XmGetPixmapByDepth) %p\n", |
| 225 Makefile). | 226 s)); |
| 226 | 227 |
| 227 *** Excessive optimization on AIX 4.2 can lead to compiler failure. | 228 - (void) _LTHashTableForEachItem(PixmapCache, YowIter, (XtPointer)s); |
| 228 | 229 } |
| 229 Valdis.Kletnieks@vt.edu writes: | 230 |
| 230 At least at the b34 level, and the latest-and-greatest IBM xlc | 231 *** XEmacs crashes on exit (#2) |
| 231 (3.1.4.4), there are problems with -O3. I haven't investigated | 232 |
| 232 further. | 233 Especially frequent with multiple frames. Crashes that produce C |
| 233 | 234 backtraces like this: |
| 234 | |
| 235 ** SunOS/Solaris | |
| 236 *** Crashes when using Motif libraries, especially with multiple frames. | |
| 237 | |
| 238 Crashes that produce C-backtraces like this: | |
| 239 | 235 |
| 240 #0 0xfec9a118 in _libc_kill () from /usr/lib/libc.so.1 | 236 #0 0xfec9a118 in _libc_kill () from /usr/lib/libc.so.1 |
| 241 #1 0x77f48 in fatal_error_signal (sig=11) | 237 #1 0x77f48 in fatal_error_signal (sig=11) |
| 242 at /codes/rpluim/xemacs-21.4/src/emacs.c:539 | 238 at /codes/rpluim/xemacs-21.4/src/emacs.c:539 |
| 243 #2 <signal handler called> | 239 #2 <signal handler called> |
| 258 For information (although they have not been confirmed to work), the | 254 For information (although they have not been confirmed to work), the |
| 259 equivalent patches for Solaris 2.8 are: | 255 equivalent patches for Solaris 2.8 are: |
| 260 | 256 |
| 261 108940-33 108652-25 | 257 108940-33 108652-25 |
| 262 | 258 |
| 259 *** On HP-UX 11.0 XEmacs causes excessive X11 errors when running. | |
| 260 (also appears on AIX as reported in comp.emacs.xemacs) | |
| 261 | |
| 262 Marcus Thiessel <marcus@xemacs.org> | |
| 263 | |
| 264 Unfortunately, XEmacs releases prior to 21.0 don't work with | |
| 265 Motif2.1. It will compile but you will get excessive X11 errors like | |
| 266 | |
| 267 xemacs: X Error of failed request: BadGC (invalid GC parameter) | |
| 268 | |
| 269 and finally XEmacs gets killed. A workaround is to use the | |
| 270 Motif1.2_R6 libraries. You can the following line to your call to | |
| 271 configure: | |
| 272 | |
| 273 --x-libraries="/usr/lib/Motif1.2_R6 -L/usr/lib/X11R6" | |
| 274 | |
| 275 Make sure /usr/lib/Motif1.2_R6/libXm.sl is a link to | |
| 276 /usr/lib/Motif1.2_R6/libXm.3. | |
| 277 | |
| 278 *** On HP-UX 11.0: Object "" does not have windowed ancestor | |
| 279 | |
| 280 Marcus Thiessel <marcus@xemacs.org> | |
| 281 | |
| 282 XEmacs dies without core file and reports: | |
| 283 | |
| 284 Error: Object "" does not have windowed ancestor. | |
| 285 | |
| 286 This is a bug. Please apply the patch PHSS_19964 (check if | |
| 287 superseded). The other alternative is to link with Motif1.2_R6 (see | |
| 288 previous item). | |
| 289 | |
| 290 *** Motif dialog boxes lose on Irix. | |
| 291 | |
| 292 Larry Auton <lda@control.att.com> writes: | |
| 293 Beware of not specifying | |
| 294 | |
| 295 --with-dialogs=athena | |
| 296 | |
| 297 if it builds with the motif dialogs [boom!] you're a dead man. | |
| 298 | |
| 299 | |
| 300 ** AIX | |
| 301 *** IBM compiler fails: "The character # is not a valid C source character." | |
| 302 | |
| 303 Most recently observed in 21.5.9, due to USE_KKCC ifdefs (they just | |
| 304 happen to tickle the implementation). | |
| 305 | |
| 306 Valdis Kletnieks says: | |
| 307 | |
| 308 The problem is that IBM defines a *MACRO* called 'memcpy', and we | |
| 309 have stuck a #ifdef/#endif inside the macro call. As a workaround, | |
| 310 try adding '-U__STR__' to your CFLAGS - this will cause string.h to | |
| 311 not do a #define for strcpy() to __strcpy() - it uses this for | |
| 312 automatic inlining support. | |
| 313 | |
| 314 (For the record, the same issue affects a number of other functions | |
| 315 defined in string.h - basically anything the compiler knows how to | |
| 316 inline.) | |
| 317 | |
| 318 *** On AIX 4.3, you must specify --with-dialogs=athena with configure | |
| 319 | |
| 320 *** The libXt shipped with AIX 4.3 up to 4.3.2 is broken. This causes | |
| 321 xemacs -nw to fail in various ways. The official APAR is this: | |
| 322 | |
| 323 APAR NUMBER: <IX89470> RESOLVED AS: PROGRAM ERROR | |
| 324 | |
| 325 ABSTRACT: | |
| 326 <IX89470>: LIBXT.A INCORRECT HANDLING OF EXCEPTIONS IN XTAPPADDINPUT | |
| 327 | |
| 328 The solution is to install X11.base.lib at version >=4.3.2.5. | |
| 329 | |
| 330 *** On AIX, you get this compiler error message: | |
| 331 | |
| 332 Processing include file ./XMenuInt.h | |
| 333 1501-106: (S) Include file X11/Xlib.h not found. | |
| 334 | |
| 335 This means your system was installed with only the X11 runtime i.d | |
| 336 libraries. You have to find your sipo (bootable tape) and install | |
| 337 X11Dev... with smit. | |
| 338 | |
| 339 *** On AIX 4.1.2, linker error messages such as | |
| 340 ld: 0711-212 SEVERE ERROR: Symbol .__quous, found in the global symbol table | |
| 341 of archive /usr/lib/libIM.a, was not defined in archive member shr.o. | |
| 342 | |
| 343 This is a problem in libIM.a. You can work around it by executing | |
| 344 these shell commands in the src subdirectory of the directory where | |
| 345 you build Emacs: | |
| 346 | |
| 347 cp /usr/lib/libIM.a . | |
| 348 chmod 664 libIM.a | |
| 349 ranlib libIM.a | |
| 350 | |
| 351 Then change -lIM to ./libIM.a in the command to link temacs (in | |
| 352 Makefile). | |
| 353 | |
| 354 *** Excessive optimization on AIX 4.2 can lead to compiler failure. | |
| 355 | |
| 356 Valdis.Kletnieks@vt.edu writes: | |
| 357 At least at the b34 level, and the latest-and-greatest IBM xlc | |
| 358 (3.1.4.4), there are problems with -O3. I haven't investigated | |
| 359 further. | |
| 360 | |
| 361 | |
| 362 ** SunOS/Solaris | |
| 263 *** Dumping error when using GNU binutils / GNU ld on a Sun. | 363 *** Dumping error when using GNU binutils / GNU ld on a Sun. |
| 264 | 364 |
| 265 Errors similar to the following: | 365 Errors similar to the following: |
| 266 | 366 |
| 267 Dumping under the name xemacs unexec(): | 367 Dumping under the name xemacs unexec(): |
| 561 | 661 |
| 562 Marcus Thiessel <marcus@xemacs.org> | 662 Marcus Thiessel <marcus@xemacs.org> |
| 563 | 663 |
| 564 This might be a sed problem. For your own safety make sure to use | 664 This might be a sed problem. For your own safety make sure to use |
| 565 GNU sed while dumping XEmacs. | 665 GNU sed while dumping XEmacs. |
| 566 | |
| 567 *** On HP-UX 11.0 XEmacs causes excessive X11 errors when running. | |
| 568 (also appears on AIX as reported in comp.emacs.xemacs) | |
| 569 | |
| 570 Marcus Thiessel <marcus@xemacs.org> | |
| 571 | |
| 572 Unfortunately, XEmacs releases prior to 21.0 don't work with | |
| 573 Motif2.1. It will compile but you will get excessive X11 errors like | |
| 574 | |
| 575 xemacs: X Error of failed request: BadGC (invalid GC parameter) | |
| 576 | |
| 577 and finally XEmacs gets killed. A workaround is to use the | |
| 578 Motif1.2_R6 libraries. You can the following line to your call to | |
| 579 configure: | |
| 580 | |
| 581 --x-libraries="/usr/lib/Motif1.2_R6 -L/usr/lib/X11R6" | |
| 582 | |
| 583 Make sure /usr/lib/Motif1.2_R6/libXm.sl is a link to | |
| 584 /usr/lib/Motif1.2_R6/libXm.3. | |
| 585 | |
| 586 *** On HP-UX 11.0: Object "" does not have windowed ancestor | |
| 587 | |
| 588 Marcus Thiessel <marcus@xemacs.org> | |
| 589 | |
| 590 XEmacs dies without core file and reports: | |
| 591 | |
| 592 Error: Object "" does not have windowed ancestor. | |
| 593 | |
| 594 This is a bug. Please apply the patch PHSS_19964 (check if | |
| 595 superseded). The other alternative is to link with Motif1.2_R6 (see | |
| 596 previous item). | |
| 597 | 666 |
| 598 | 667 |
| 599 ** SCO OpenServer | 668 ** SCO OpenServer |
| 600 *** Native cc on SCO OpenServer 5 is now OK. Icc may still throw you | 669 *** Native cc on SCO OpenServer 5 is now OK. Icc may still throw you |
| 601 a curve. Here is what Robert Lipe <robertl@arnet.com> says: | 670 a curve. Here is what Robert Lipe <robertl@arnet.com> says: |
| 1478 compatible include files into the compilation. In particular this | 1547 compatible include files into the compilation. In particular this |
| 1479 affected virtually all ioctl() calls. | 1548 affected virtually all ioctl() calls. |
| 1480 | 1549 |
| 1481 | 1550 |
| 1482 ** Linux | 1551 ** Linux |
| 1483 *** XEmacs crashes on exit. | |
| 1484 | |
| 1485 This is known to happen with Lesstif version 0.93.36. It is | |
| 1486 apparently due to breakage in Lesstif. There is a patch for Lesstif. | |
| 1487 | |
| 1488 Frank McIngvale <frankm@hiwaay.net> says: | |
| 1489 | |
| 1490 Ok, 0.93.34 works, and I tracked down the crash to a section | |
| 1491 marked "experimental" in 0.93.36. Patch attached, "works for me". | |
| 1492 | |
| 1493 diff -u -r lesstif-0.93.36/lib/Xm/ImageCache.c lesstif-0.93.36-mod/lib/Xm/ImageCache.c | |
| 1494 --- lesstif-0.93.36/lib/Xm/ImageCache.c 2002-08-05 14:53:24.000000000 -0500 | |
| 1495 +++ lesstif-0.93.36-mod/lib/Xm/ImageCache.c 2002-11-11 11:13:12.000000000 -0600 | |
| 1496 @@ -1166,5 +1166,4 @@ | |
| 1497 DEBUGOUT(_LtDebug0(__FILE__, NULL, "_LtImageCacheScreenDestroy (XmGetPixmapByDepth) %p\n", | |
| 1498 s)); | |
| 1499 | |
| 1500 - (void) _LTHashTableForEachItem(PixmapCache, YowIter, (XtPointer)s); | |
| 1501 } | |
| 1502 | |
| 1503 | |
| 1504 *** XEmacs crashes on startup, in make-frame. | 1552 *** XEmacs crashes on startup, in make-frame. |
| 1505 | 1553 |
| 1506 Typically the Lisp backtrace includes | 1554 Typically the Lisp backtrace includes |
| 1507 | 1555 |
| 1508 make-frame(nil #<x-device on ":0.0" 0x2558>) | 1556 make-frame(nil #<x-device on ":0.0" 0x2558>) |
| 1686 *** Trouble using ptys on IRIX, or running out of ptys. | 1734 *** Trouble using ptys on IRIX, or running out of ptys. |
| 1687 | 1735 |
| 1688 The program mkpts (which may be in `/usr/adm' or `/usr/sbin') needs to | 1736 The program mkpts (which may be in `/usr/adm' or `/usr/sbin') needs to |
| 1689 be set-UID to root, or non-root programs like Emacs will not be able | 1737 be set-UID to root, or non-root programs like Emacs will not be able |
| 1690 to allocate ptys reliably. | 1738 to allocate ptys reliably. |
| 1691 | |
| 1692 *** Motif dialog boxes lose on Irix. | |
| 1693 | |
| 1694 Larry Auton <lda@control.att.com> writes: | |
| 1695 Beware of not specifying | |
| 1696 | |
| 1697 --with-dialogs=athena | |
| 1698 | |
| 1699 if it builds with the motif dialogs [boom!] you're a dead man. | |
| 1700 | 1739 |
| 1701 *** Beware of the default image & graphics library on Irix | 1740 *** Beware of the default image & graphics library on Irix |
| 1702 | 1741 |
| 1703 Richard Cognot <cognot@ensg.u-nancy.fr> writes: | 1742 Richard Cognot <cognot@ensg.u-nancy.fr> writes: |
| 1704 | 1743 |
