Commit | Line | Data |
---|---|---|
c906108c | 1 | If you find inaccuracies in this list, please send mail to |
2a00c9ce AC |
2 | gdb-patches@sourceware.cygnus.com. If you would like to work on any |
3 | of these, you should consider sending mail to the same address, to | |
4 | find out whether anyone else is working on it. | |
c906108c | 5 | |
138f88c0 | 6 | |
4afc966c AC |
7 | Known problems in GDB 5.0 |
8 | ========================= | |
138f88c0 | 9 | |
bc9e5bbf AC |
10 | Below is a list of problems identified during the GDB 5.0 release |
11 | cycle. People hope to have these problems fixed in a follow-on | |
12 | release. | |
138f88c0 | 13 | |
138f88c0 AC |
14 | -- |
15 | ||
4fd99b5a AC |
16 | The BFD directory requires bug-fixed AUTOMAKE et.al. |
17 | ||
18 | AUTOMAKE 1.4 incorrectly set the TEXINPUTS environment variable. It | |
19 | contained the full path to texinfo.tex when it should have only | |
20 | contained the directory. The bug has been fixed in the current | |
21 | AUTOMAKE sources. Automake snapshots can be found in: | |
22 | ftp://sourceware.cygnus.com/pub/gdb/snapshots | |
23 | and ftp://sourceware.cygnus.com/pub/binutils | |
24 | ||
25 | -- | |
26 | ||
bc9e5bbf | 27 | RFD: infrun.c: No bpstat_stop_status call after proceed over break? |
138f88c0 AC |
28 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00665.html |
29 | ||
bc9e5bbf AC |
30 | GDB misses watchpoint triggers after proceeding over a breakpoint on |
31 | x86 targets. | |
32 | ||
138f88c0 AC |
33 | -- |
34 | ||
67edb2c6 AC |
35 | x86 linux GDB and SIGALRM (???) |
36 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00803.html | |
37 | ||
37d4dc74 MK |
38 | This problem has been fixed, but a regression test still needs to be |
39 | added to the testsuite: | |
40 | http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00309.html | |
67edb2c6 | 41 | |
bc9e5bbf | 42 | Mark |
67edb2c6 AC |
43 | |
44 | -- | |
45 | ||
9d6d78f2 AC |
46 | Revised UDP support (was: Re: [Fwd: [patch] UDP transport support]) |
47 | http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00000.html | |
48 | ||
bc9e5bbf AC |
49 | (Broken) support for GDB's remote protocol across UDP is to be |
50 | included in the follow-on release. | |
138f88c0 | 51 | |
2e4e9e68 AC |
52 | It should be noted that UDP can only work when the [Gg] packet fits in |
53 | a single UDP packet. | |
54 | ||
b2f4b24d AC |
55 | -- |
56 | ||
57 | Can't build IRIX -> arm GDB. | |
58 | http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00356.html | |
59 | ||
60 | David Whedon writes: | |
61 | > Now I'm building for an embedded arm target. If there is a way of turning | |
62 | > remote-rdi off, I couldn't find it. It looks like it gets built by default | |
63 | > in gdb/configure.tgt(line 58) Anyway, the build dies in | |
64 | > gdb/rdi-share/unixcomm.c. SERPORT1 et. al. never get defined because we | |
65 | > aren't one of the architectures supported. | |
66 | ||
6bc37a96 AC |
67 | -- |
68 | ||
69 | Problem with weak functions | |
70 | http://sourceware.cygnus.com/ml/gdb/2000-05/msg00060.html | |
71 | ||
72 | Dan Nicolaescu writes: | |
73 | > It seems that gdb-4.95.1 does not display correctly the function when | |
74 | > stoping in weak functions. | |
75 | > | |
76 | > It stops in a function that is defined as weak, not in the function | |
77 | > that is actualy run... | |
b2f4b24d | 78 | |
26099b4a AC |
79 | -- |
80 | ||
81 | GDB 5.0 doesn't work on Linux/SPARC | |
82 | ||
3fffcb5e AC |
83 | -- |
84 | ||
4afc966c AC |
85 | Code Cleanups: Next Release |
86 | =========================== | |
3fffcb5e | 87 | |
4afc966c AC |
88 | The following are small cleanups that will hopefully be completed by |
89 | the follow on to 5.0. | |
3fffcb5e | 90 | |
138f88c0 AC |
91 | -- |
92 | ||
4afc966c | 93 | Delete macro TARGET_BYTE_ORDER_SELECTABLE. |
bc9e5bbf | 94 | |
4afc966c AC |
95 | Patches in the database. |
96 | ||
97 | -- | |
98 | ||
26099b4a | 99 | Purge PARAMS. |
4afc966c AC |
100 | |
101 | Eliminate all uses of PARAMS in GDB's source code. | |
102 | ||
103 | -- | |
104 | ||
78566ebe AC |
105 | Fix copyright notices. |
106 | ||
107 | Turns out that ``1998-2000'' isn't considered valid :-( | |
108 | ||
109 | http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00467.html | |
110 | ||
1ba607ad AC |
111 | -- |
112 | ||
113 | Find something better than DEFAULT_BFD_ARCH, DEFAULT_BFD_VEC to | |
114 | determine the default isa/byte-order. | |
115 | ||
116 | -- | |
117 | ||
118 | Rely on BFD_BIG_ENDIAN and BFD_LITTLE_ENDIAN instead of host dependant | |
119 | BIG_ENDIAN and LITTLE_ENDIAN. | |
120 | ||
e2ad119d AC |
121 | -- |
122 | ||
123 | printcmd.c (print_address_numeric): | |
124 | ||
125 | NOTE: This assumes that the significant address information is kept in | |
126 | the least significant bits of ADDR - the upper bits were either zero | |
127 | or sign extended. Should ADDRESS_TO_POINTER() or some | |
128 | ADDRESS_TO_PRINTABLE() be used to do the conversion? | |
129 | ||
4afc966c AC |
130 | -- |
131 | ||
132 | Code Cleanups: General | |
133 | ====================== | |
134 | ||
135 | The following are more general cleanups and fixes. They are not tied | |
136 | to any specific release. | |
bc9e5bbf AC |
137 | |
138 | -- | |
139 | ||
d8038014 AC |
140 | Eliminate more compiler warnings. |
141 | ||
142 | Of course there also needs to be the usual debate over which warnings | |
143 | are valid and how to best go about this. | |
144 | ||
145 | One method: choose a single option; get agreement that it is | |
146 | reasonable; try it out to see if there isn't anything silly about it | |
147 | (-Wunused-parameters is an example of that) then incrementally hack | |
148 | away. | |
149 | ||
150 | The other method is to enable all warnings and eliminate them from one | |
151 | file at a time. | |
152 | ||
153 | -- | |
154 | ||
4afc966c | 155 | Elimination of ``(catch_errors_ftype *) func''. |
bc9e5bbf | 156 | |
4afc966c | 157 | Like make_cleanup_func it isn't portable. |
6ecce94d AC |
158 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00791.html |
159 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00814.html | |
160 | ||
4afc966c AC |
161 | |
162 | -- | |
163 | ||
164 | Nuke USG define. | |
165 | ||
166 | -- | |
167 | ||
168 | [PATCH/5] src/intl/Makefile.in:distclean additions | |
169 | http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00363.html | |
170 | ||
171 | Do not forget to merge the patch back into the trunk. | |
172 | ||
173 | -- | |
174 | ||
175 | Rationalize the host-endian code (grep for HOST_BYTE_ORDER). | |
176 | ||
177 | At present defs.h includes <endian.h> (which is linux specific) yet | |
178 | almost nothing depends on it. Suggest "gdb_endian.h" which can also | |
179 | handle <machine/endian.h> and only include that where it is really | |
180 | needed. | |
181 | ||
182 | -- | |
183 | ||
184 | Replace asprintf() calls with xasprintf() calls. | |
185 | ||
186 | As with things like strdup() most calls to asprintf() don't check the | |
187 | return value. | |
188 | ||
189 | -- | |
190 | ||
191 | Replace strsave() + mstrsave() with libiberty:xstrdup(). | |
192 | ||
193 | -- | |
194 | ||
195 | Replace savestring() with something from libiberty. | |
196 | ||
197 | An xstrldup()? but that would have different semantics. | |
198 | ||
199 | -- | |
200 | ||
201 | Rationalize use of floatformat_unknown in GDB sources. | |
202 | ||
203 | Instead of defaulting to floatformat_unknown, should hosts/targets | |
204 | specify the value explicitly? | |
205 | ||
206 | http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00447.html | |
207 | ||
208 | -- | |
209 | ||
210 | Add a ``name'' member to include/floatformat.h:struct floatformat. | |
211 | Print that name in gdbarch.c. | |
212 | ||
213 | -- | |
214 | ||
215 | Sort out the harris mess in include/floatformat.h (it hardwires two | |
216 | different floating point formats). | |
217 | ||
218 | -- | |
219 | ||
220 | See of the GDB local floatformat_do_doublest() and libiberty's | |
221 | floatformat_to_double (which was once GDB's ...) can be merged some | |
222 | how. | |
223 | ||
224 | -- | |
225 | ||
226 | Eliminate mmalloc() from GDB. | |
227 | ||
228 | Also eliminate it from defs.h. | |
229 | ||
230 | -- | |
231 | ||
232 | Eliminate PTR. ISO-C allows ``void *''. | |
233 | ||
234 | -- | |
235 | ||
236 | Eliminate abort (). | |
237 | ||
238 | GDB should never abort. GDB should either throw ``error ()'' or | |
239 | ``internal_error ()''. Better still GDB should naturally unwind with | |
240 | an error status. | |
241 | ||
242 | -- | |
243 | ||
244 | GDB probably doesn't build on FreeBSD pre 2.2.x | |
245 | http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00378.html | |
246 | ||
247 | Fixes to get FreeBSD working on 2.2.x, 3.x and 4.x caused the code to | |
248 | suffer bit rot. | |
bc9e5bbf AC |
249 | |
250 | -- | |
251 | ||
78566ebe AC |
252 | Deprecate "fg". Apparently ``fg'' is actually continue. |
253 | ||
254 | http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00417.html | |
255 | ||
256 | -- | |
257 | ||
258 | Deprecate current use of ``floatformat_unknown''. | |
259 | ||
260 | Require all targets to explicitly provide their float format instead | |
261 | of defaulting to floatformat unknown. Doing the latter leads to nasty | |
262 | bugs. | |
263 | ||
264 | http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00447.html | |
bc9e5bbf AC |
265 | |
266 | -- | |
267 | ||
78566ebe AC |
268 | Rationalize floatformat_to_double() vs floatformat_to_doublest(). |
269 | ||
270 | Looks like GDB migrated floatformat_to_double() to libiberty but then | |
271 | turned around and created a ..._to_doublest() the latter containing | |
272 | several bug fixes. | |
273 | ||
274 | http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00472.html | |
275 | ||
276 | -- | |
277 | ||
278 | Move floatformat_ia64_ext to libiberty/include floatformat.[ch]. | |
279 | ||
280 | http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00466.html | |
281 | ||
282 | -- | |
283 | ||
78566ebe AC |
284 | Follow through `make check' with --enable-shared. |
285 | ||
286 | When the srcware tree is configured with --enable-shared, the `expect' | |
287 | program won't run properly. Jim Wilson found out gdb has a local hack | |
288 | to set LD_LIBRARY_PATH, but, AFAIK, no other project has been hacked | |
289 | similarly. | |
290 | ||
291 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00845.html | |
292 | ||
b4a20239 AC |
293 | -- |
294 | ||
b4a20239 AC |
295 | The ``maintenance deprecate set endian big'' command doesn't notice |
296 | that it is deprecating ``set endian'' and not ``set endian big'' (big | |
297 | is implemented using an enum). Is anyone going to notice this? | |
298 | ||
299 | -- | |
300 | ||
301 | When tab expanding something like ``set arch<tab>'' ignore the | |
302 | deprecated ``set archdebug'' and expand to ``set architecture''. | |
303 | ||
78566ebe AC |
304 | -- |
305 | ||
53904c9e AC |
306 | Eliminate ``arm_register_names[j] = (char *) regnames[j]'' and the |
307 | like from arm-tdep.c. | |
308 | ||
309 | -- | |
310 | ||
311 | Fix uses of ->function.cfunc = set_function(). | |
312 | ||
313 | The command.c code calls sfunc() when a set command. Rather than | |
314 | change it suggest fixing the callback function so that it is more | |
315 | useful. See: | |
316 | ||
317 | http://sourceware.cygnus.com/ml/gdb-patches/2000-06/msg00062.html | |
318 | ||
319 | See also ``Fix implementation of ``target xxx''.'' below. | |
320 | ||
bf64bfd6 AC |
321 | -- |
322 | ||
323 | IRIX 3.x support is probably broken. | |
324 | ||
5d35f0ac AC |
325 | -- |
326 | ||
327 | Delete sim/SIM_HAVE_BREAKPOINTS and gdb/SIM_HAS_BREAKPOINTS. | |
328 | http://sourceware.cygnus.com/ml/gdb-patches/2000-07/msg00042.html | |
329 | ||
330 | Apart from the d30v, are there any sim/common simulators that make use | |
331 | of this? | |
332 | ||
333 | A brief summary of what happended is that sim/common/sim-break.c was | |
334 | created as a good idea. It turned out a better idea was to use | |
335 | SIM_SIGBREAK and have GDB pass back sim_resume (..., SIGBREAK). | |
336 | ||
337 | -- | |
338 | ||
339 | parse.c:build_parse() has a buffer overrun. | |
340 | ||
53904c9e | 341 | -- |
78566ebe | 342 | |
4afc966c AC |
343 | New Features and Fixes |
344 | ====================== | |
bc9e5bbf | 345 | |
4afc966c AC |
346 | These are harder than cleanups but easier than work involving |
347 | fundamental architectural change. | |
bc9e5bbf AC |
348 | |
349 | -- | |
350 | ||
4afc966c AC |
351 | Add built-by, build-date, tm, xm, nm and anything else into gdb binary |
352 | so that you can see how the GDB was created. | |
bc9e5bbf | 353 | |
4afc966c AC |
354 | Some of these (*m.h) would be added to the generated config.h. That |
355 | in turn would fix a long standing bug where by the build process many | |
356 | not notice a changed tm.h file. Since everything depends on config.h, | |
357 | a change to *m.h forces a change to config.h and, consequently forces | |
358 | a rebuild. | |
bc9e5bbf AC |
359 | |
360 | -- | |
361 | ||
4afc966c AC |
362 | Add an "info bfd" command that displays supported object formats, |
363 | similarly to objdump -i. | |
5683e87a | 364 | |
4afc966c | 365 | Is there a command already? |
5683e87a AC |
366 | |
367 | -- | |
368 | ||
4afc966c | 369 | Fix ``I'm sorry, Dave, I can't do that.'' from symfile.c. |
bc9e5bbf | 370 | |
4afc966c | 371 | This requires internationalization. |
bc9e5bbf | 372 | |
4afc966c | 373 | -- |
bc9e5bbf | 374 | |
2e4e9e68 AC |
375 | Add support for: |
376 | ||
377 | (gdb) p fwprintf(stdout,L"%S\n", f) | |
378 | No symbol "L" in current context. | |
379 | ||
380 | -- | |
381 | ||
4afc966c | 382 | Convert GDB build process to AUTOMAKE. |
bc9e5bbf | 383 | |
4afc966c | 384 | See also sub-directory configure below. |
138f88c0 | 385 | |
b4a20239 AC |
386 | The current convention is (kind of) to use $(<header>_h) in all |
387 | dependency lists. It isn't done in a consistent way. | |
388 | ||
67edb2c6 AC |
389 | -- |
390 | ||
4afc966c | 391 | Cleanup configury support for optional sub-directories. |
7ae38352 | 392 | |
4afc966c AC |
393 | Check how GCC handles multiple front ends for an example of how things |
394 | could work. A tentative first step is to rationalize things so that | |
395 | all sub directories are handled in a fashion similar to gdb/mi. | |
396 | ||
397 | See also automake above. | |
7ae38352 AC |
398 | |
399 | -- | |
400 | ||
4afc966c AC |
401 | Restructure gdb directory tree so that it avoids any 8.3 and 14 |
402 | filename problems. | |
67edb2c6 | 403 | |
4afc966c | 404 | -- |
67edb2c6 | 405 | |
4afc966c AC |
406 | Add a transcript mechanism to GDB. |
407 | ||
408 | Such a mechanism might log all gdb input and output to a file in a | |
409 | form that would allow it to be replayed. It could involve ``gdb | |
410 | --transcript=FILE'' or it could involve ``(gdb) transcript file''. | |
67edb2c6 AC |
411 | |
412 | -- | |
413 | ||
4afc966c | 414 | Can the xdep files be replaced by autoconf? |
bc9e5bbf | 415 | |
4afc966c | 416 | -- |
bc9e5bbf | 417 | |
4afc966c | 418 | Document trace machinery |
bc9e5bbf | 419 | |
4afc966c AC |
420 | -- |
421 | ||
78566ebe AC |
422 | Document ui-out and ui-file. |
423 | ||
424 | http://sourceware.cygnus.com/ml/gdb/2000-04/msg00121.html | |
425 | ||
426 | -- | |
427 | ||
428 | Update texinfo.tex to latest? | |
429 | ||
430 | ||
431 | ||
432 | -- | |
433 | ||
434 | Incorporate agentexpr.texi into gdb.texinfo | |
435 | ||
436 | agentexpr.texi mostly describes the details of the byte code used for | |
437 | tracepoints, not the internals of the support for this in GDB. So it | |
438 | looks like gdb.texinfo is a better place for this information. | |
439 | ||
440 | http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00566.html | |
441 | ||
442 | -- | |
443 | ||
4afc966c | 444 | Document overlay machinery. |
bc9e5bbf | 445 | |
7ae38352 AC |
446 | -- |
447 | ||
4afc966c | 448 | ``(gdb) catch signal SIGNAL'' |
7ae38352 | 449 | |
4afc966c AC |
450 | Overlaps with ``handle SIGNAL'' but the implied behavour is different. |
451 | You can attach commands to a catch but not a handle. A handle has a | |
452 | limited number of hardwired actions. | |
7ae38352 AC |
453 | |
454 | -- | |
455 | ||
4afc966c | 456 | Get the TUI working on all platforms. |
7ae38352 | 457 | |
bc9e5bbf AC |
458 | -- |
459 | ||
4afc966c AC |
460 | Add support for ``gdb --- PROGRAM ARGS ...''. |
461 | Add support for ``gdb -cmd=...'' | |
9debab2f | 462 | |
4afc966c AC |
463 | Along with many variations. Check: |
464 | ||
465 | ????? for a full discussion. | |
466 | ||
467 | for a discussion. | |
9debab2f AC |
468 | |
469 | -- | |
470 | ||
4afc966c | 471 | Implement ``(gdb) !ls''. |
e55e8cee | 472 | |
4afc966c AC |
473 | Which is very different from ``(gdb) ! ls''. Implementing the latter |
474 | is trivial. | |
475 | ||
476 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00034.html | |
e55e8cee AC |
477 | |
478 | -- | |
479 | ||
b4a20239 AC |
480 | Change the (char *list[]) to (const char (*)[]) so that dynamic lists can |
481 | be passed. | |
482 | ||
483 | -- | |
484 | ||
485 | When tab expanding something like ``set arch<tab>'' ignore the | |
486 | deprecated ``set archdebug'' and expand to ``set architecture''. | |
487 | ||
488 | -- | |
489 | ||
4afc966c AC |
490 | Replace the code that uses the host FPU with an emulator of the target |
491 | FPU. | |
7ae38352 | 492 | |
4afc966c AC |
493 | -- |
494 | ||
495 | Thread Support | |
496 | ============== | |
7ae38352 AC |
497 | |
498 | -- | |
499 | ||
4afc966c AC |
500 | Generic: lin-thread cannot handle thread exit (Mark Kettenis, Michael |
501 | Snyder) http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00525.html | |
7ae38352 | 502 | |
4afc966c AC |
503 | The thread_db assisted debugging code doesn't handle exiting threads |
504 | properly, at least in combination with glibc 2.1.3 (the framework is | |
505 | there, just not the actual code). There are at least two problems | |
506 | that prevent this from working. | |
507 | ||
508 | As an additional reference point, the pre thread_db code did not work | |
509 | either. | |
7ae38352 AC |
510 | |
511 | -- | |
512 | ||
4afc966c AC |
513 | GNU/Linux/x86 and random thread signals (and Solaris/SPARC but not |
514 | Solaris/x86). | |
515 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00336.html | |
7ae38352 | 516 | |
4afc966c AC |
517 | Christopher Blizzard writes: |
518 | ||
519 | So, I've done some more digging into this and it looks like Jim | |
520 | Kingdon has reported this problem in the past: | |
521 | ||
522 | http://sourceware.cygnus.com/ml/bug-gdb/1999-10/msg00058.html | |
523 | ||
524 | I can reproduce this problem both with and without Tom's patch. Has | |
525 | anyone seen this before? Maybe have a solution for it hanging around? | |
526 | :) | |
527 | ||
528 | There's a test case for this documented at: | |
529 | ||
530 | when debugging threaded applications you get extra SIGTRAPs | |
531 | http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=9565 | |
532 | ||
533 | [There should be a GDB testcase - cagney] | |
7ae38352 AC |
534 | |
535 | -- | |
536 | ||
4afc966c AC |
537 | GDB5 TOT on unixware 7 |
538 | http://sourceware.cygnus.com/ml/gdb/2000-04/msg00119.html | |
7ae38352 | 539 | |
4afc966c AC |
540 | Robert Lipe writes: |
541 | > I just spun the top of tree of the GDB5 branch on UnixWare 7. As a | |
542 | > practical matter, the current thread support is somewhat more annoying | |
543 | > than when GDB was thread-unaware. | |
7ae38352 AC |
544 | |
545 | -- | |
546 | ||
4afc966c | 547 | Migrate qfThreadInfo packet -> qThreadInfo. (Andrew Cagney) |
7ae38352 | 548 | |
4afc966c AC |
549 | Add support for packet enable/disable commands with these thread |
550 | packets. General cleanup. | |
551 | ||
552 | [PATCH] Document the ThreadInfo remote protocol queries | |
553 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00832.html | |
554 | ||
555 | [PATCH] "info threads" queries for remote.c | |
556 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00831.html | |
7ae38352 AC |
557 | |
558 | -- | |
559 | ||
4afc966c AC |
560 | Language Support |
561 | ================ | |
7ae38352 | 562 | |
4afc966c | 563 | New languages come onto the scene all the time. |
7ae38352 AC |
564 | |
565 | -- | |
566 | ||
4afc966c | 567 | Pascal (Pierre Muller, David Taylor) |
7ae38352 | 568 | |
4afc966c AC |
569 | Pierre Muller has contributed patches for adding Pascal Language |
570 | support to GDB. | |
571 | ||
572 | 2 pascal language patches inserted in database | |
573 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00521.html | |
574 | ||
575 | Indent -gnu ? | |
576 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00496.html | |
577 | ||
578 | -- | |
579 | ||
580 | Java (Anthony Green, David Taylor) | |
581 | ||
582 | Anthony Green has a number of Java patches that did not make it into | |
02d44fdb AG |
583 | the 5.0 release. The first two are in cvs now, but the third needs |
584 | some fixing up before it can go in. | |
4afc966c AC |
585 | |
586 | Patch: java tests | |
587 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00512.html | |
588 | ||
589 | Patch: java booleans | |
590 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00515.html | |
591 | ||
592 | Patch: handle N_MAIN stab | |
593 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00527.html | |
594 | ||
595 | -- | |
596 | ||
597 | [Comming...] | |
598 | ||
599 | Modify gdb to work correctly with Pascal. | |
600 | ||
601 | -- | |
602 | ||
603 | Re: Various C++ things | |
604 | ||
26099b4a AC |
605 | value_headof/value_from_vtable_info are worthless, and should be |
606 | removed. The one place in printcmd.c that uses it should use the RTTI | |
607 | functions. | |
4afc966c | 608 | |
26099b4a AC |
609 | RTTI for g++ should be using the typeinfo functions rather than the |
610 | vtables. The typeinfo functions are always at offset 4 from the | |
611 | beginning of the vtable, and are always right. The vtables will have | |
612 | weird names like E::VB sometimes. The typeinfo function will always | |
613 | be "E type_info function", or somesuch. | |
4afc966c | 614 | |
26099b4a AC |
615 | value_virtual_fn_field needs to be fixed so there are no failures for |
616 | virtual functions for C++ using g++. | |
4afc966c | 617 | |
26099b4a AC |
618 | Testsuite cases are the major priority right now for C++ support, |
619 | since i have to make a lot of changes that could potentially break | |
620 | each other. | |
7ae38352 AC |
621 | |
622 | -- | |
623 | ||
4afc966c | 624 | Add support for Modula3 |
7ae38352 | 625 | |
4afc966c | 626 | Get DEC/Compaq to contribute their Modula-3 support. |
7ae38352 AC |
627 | |
628 | -- | |
629 | ||
4afc966c AC |
630 | Remote Protocol Support |
631 | ======================= | |
7ae38352 AC |
632 | |
633 | -- | |
634 | ||
4afc966c | 635 | set/show remote X-packet ... |
7ae38352 | 636 | |
4afc966c AC |
637 | ``(gdb) help set remote X-packet'' doesn't list the applicable |
638 | responses. The help message needs to be expanded. | |
7ae38352 AC |
639 | |
640 | -- | |
641 | ||
642 | Remote protocol doco feedback. | |
643 | ||
644 | Too much feedback to mention needs to be merged in (901660). Search | |
645 | for the word ``remote''. | |
646 | ||
4afc966c AC |
647 | |
648 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00023.html | |
649 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00056.html | |
650 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00382.html | |
651 | ||
7ae38352 AC |
652 | -- |
653 | ||
4afc966c | 654 | GDB doesn't recover gracefully from remote protocol errors. |
7ae38352 | 655 | |
4afc966c AC |
656 | GDB wasn't checking for NAKs from the remote target. Instead a NAK is |
657 | ignored and a timeout is required before GDB retries. A pre-cursor to | |
658 | fixing this this is making GDB's remote protocol packet more robust. | |
659 | ||
660 | While downloading to a remote protocol target, gdb ignores packet | |
661 | errors in so far as it will continue to edownload with chunk N+1 even | |
662 | if chunk N was not correctly sent. This causes gdb.base/remote.exp to | |
663 | take a painfully long time to run. As a PS that test needs to be | |
664 | fixed so that it builds on 16 bit machines. | |
7ae38352 AC |
665 | |
666 | -- | |
667 | ||
4afc966c | 668 | Add the cycle step command. |
7ae38352 | 669 | |
4afc966c | 670 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00237.html |
7ae38352 | 671 | |
26099b4a AC |
672 | -- |
673 | ||
674 | Resolve how to scale things to support very large packets. | |
675 | ||
676 | -- | |
677 | ||
678 | Resolve how to handle a target that changes things like its endianess | |
679 | on the fly - should it be returned in the ``T'' packet? | |
680 | ||
681 | Underlying problem is that the register file is target endian. If the | |
682 | target endianess changes gdb doesn't know. | |
683 | ||
2e4e9e68 AC |
684 | -- |
685 | ||
686 | Use the recently added add_set_auto_boolean_cmd() function to | |
687 | implement the X-packet commands. | |
688 | ||
689 | -- | |
690 | ||
691 | Rename read_register{,_pid}() to read_unsigned_register{,_pid}(). | |
692 | ||
7ae38352 AC |
693 | -- |
694 | ||
4afc966c AC |
695 | Symbol Support |
696 | ============== | |
7ae38352 | 697 | |
4afc966c AC |
698 | If / when GDB starts to support the debugging of multi-processor |
699 | (rather than multi-thread) applications the symtab code will need to | |
700 | be updated a little so that several independant symbol tables are | |
701 | active at a given time. | |
702 | ||
703 | The other interesting change is a clarification of the exact meaning | |
704 | of CORE_ADDR and that has had consequences for a few targets (that | |
705 | were abusing that data type). | |
7ae38352 | 706 | |
d8038014 AC |
707 | -- |
708 | ||
4afc966c | 709 | Investiagate ways of reducing memory. |
d8038014 AC |
710 | |
711 | -- | |
712 | ||
4afc966c | 713 | Investigate ways of improving load time. |
d8038014 | 714 | |
4afc966c AC |
715 | -- |
716 | ||
717 | Get the d10v to use POINTER_TO_ADDRESS and ADDRESS_TO_POINTER. | |
718 | ||
719 | Consequence of recent symtab clarification. No marks for figuring out | |
720 | who maintains the d10v. | |
d8038014 | 721 | |
0aaf65d7 AC |
722 | -- |
723 | ||
4afc966c AC |
724 | Get the MIPS to correctly sign extend all address <-> pointer |
725 | conversions. | |
0aaf65d7 | 726 | |
4afc966c AC |
727 | Consequence of recent symtab clarification. No marks for figuring out |
728 | who maintains the MIPS. | |
0aaf65d7 | 729 | |
5d35f0ac AC |
730 | -- |
731 | ||
732 | GDB truncates 64 bit enums. | |
733 | ||
734 | http://sourceware.cygnus.com/ml/gdb-patches/2000-06/msg00290.html | |
735 | ||
26099b4a AC |
736 | -- |
737 | ||
738 | Testsuite Support | |
739 | ================= | |
740 | ||
741 | There are never to many testcases. | |
742 | ||
743 | -- | |
744 | ||
745 | Better thread testsuite. | |
746 | ||
747 | -- | |
748 | ||
749 | Better C++ testsuite. | |
750 | ||
751 | -- | |
752 | ||
753 | Look at adding a GDB specific testsuite directory so that white box | |
754 | tests of key internals can be added (eg ui_file). | |
755 | ||
756 | -- | |
757 | ||
758 | Separate out tests that involve the floating point (FP). | |
759 | ||
760 | (Something for people brining up new targets). FP and non-fp tests | |
761 | are combined. I think there should be set of basic tests that | |
762 | exercise pure integer support and then a more expanded set that | |
763 | exercise FP and FP/integer interactions. | |
764 | ||
765 | As an example, the MIPS, for n32 as problems with passing FP's and | |
766 | structs. Since most inferior call tests include FP it is difficult to | |
767 | determine of the integer tests are ok. | |
768 | ||
7ae38352 AC |
769 | -- |
770 | ||
4afc966c AC |
771 | Architectural Changes: General |
772 | ============================== | |
7ae38352 AC |
773 | |
774 | These are harder than simple cleanups / fixes and, consequently | |
775 | involve more work. Typically an Architectural Change will be broken | |
776 | down into a more digestible set of cleanups and fixes. | |
777 | ||
778 | -- | |
779 | ||
4afc966c AC |
780 | Cleanup software single step. |
781 | ||
782 | At present many targets implement software single step by directly | |
783 | blatting memory (see rs6000-tdep.c). Those targets should register | |
784 | the applicable breakpoints using the breakpoint framework. Perhaphs a | |
785 | new internal breakpoint class ``step'' is needed. | |
786 | ||
787 | -- | |
788 | ||
7ae38352 AC |
789 | Replace READ_FP() with FRAME_HANDLE(). |
790 | ||
791 | READ_FP() is a hangover from the days of the vax when the ABI really | |
792 | did have a frame pointer register. Modern architectures typically | |
793 | construct a virtual frame-handle from the stack pointer and various | |
794 | other bits of string. | |
795 | ||
796 | Unfortunatly GDB still treats this synthetic FP register as though it | |
797 | is real. That in turn really confuses users (arm and ``print $fp'' VS | |
798 | ``info registers fp''). The synthetic FP should be separated out of | |
799 | the true register set presented to the user. | |
800 | ||
801 | -- | |
802 | ||
2a00c9ce AC |
803 | Register Cache Cleanup (below from Andrew Cagney) |
804 | ||
805 | I would depict the current register architecture as something like: | |
806 | ||
807 | High GDB --> Low GDB | |
808 | | | | |
809 | \|/ \|/ | |
810 | --- REG NR ----- | |
811 | | | |
812 | register + REGISTER_BYTE(reg_nr) | |
813 | | | |
814 | \|/ | |
815 | ------------------------- | |
816 | | extern register[] | | |
817 | ------------------------- | |
818 | ||
819 | where neither the high (valops.c et.al.) or low gdb (*-tdep.c) are | |
820 | really clear on what mechanisms they should be using to manipulate that | |
821 | buffer. Further, much code assumes, dangerously, that registers are | |
822 | contigious. Having got mips-tdep.c to support multiple ABIs, believe | |
823 | me, that is a bad assumption. Finally, that register cache layout is | |
824 | determined by the current remote/local target and _not_ the less | |
825 | specific target ISA. In fact, in many cases it is determined by the | |
826 | somewhat arbitrary layout of the [gG] packets! | |
827 | ||
828 | ||
829 | How I would like the register file to work is more like: | |
830 | ||
831 | ||
832 | High GDB | |
833 | | | |
834 | \|/ | |
835 | pseudo reg-nr | |
836 | | | |
837 | map pseudo <-> | |
838 | random cache | |
839 | bytes | |
840 | | | |
841 | \|/ | |
842 | ------------ | |
843 | | register | | |
844 | | cache | | |
845 | ------------ | |
846 | /|\ | |
847 | | | |
848 | map random cache | |
849 | bytes to target | |
850 | dependant i-face | |
851 | /|\ | |
852 | | | |
853 | target dependant | |
854 | such as [gG] packet | |
855 | or ptrace buffer | |
856 | ||
857 | The main objectives being: | |
858 | ||
859 | o a clear separation between the low | |
860 | level target and the high level GDB | |
861 | ||
862 | o a mechanism that solves the general | |
863 | problem of register aliases, overlaps | |
864 | etc instead of treating them as optional | |
865 | extras that can be wedged in as an after | |
866 | thought (that is a reasonable description | |
867 | of the current code). | |
868 | ||
869 | Identify then solve the hard case and the | |
870 | rest just falls out. GDB solved the easy | |
871 | case and then tried to ignore the real | |
872 | world :-) | |
873 | ||
874 | o a removal of the assumption that the | |
875 | mapping between the register cache | |
876 | and virtual registers is largely static. | |
877 | If you flip the USR/SSR stack register | |
878 | select bit in the status-register then | |
879 | the corresponding stack registers should | |
880 | reflect the change. | |
881 | ||
882 | o a mechanism that clearly separates the | |
883 | gdb internal register cache from any | |
884 | target (not architecture) dependant | |
885 | specifics such as [gG] packets. | |
886 | ||
887 | Of course, like anything, it sounds good in theory. In reality, it | |
888 | would have to contend with many<->many relationships at both the | |
889 | virt<->cache and cache<->target level. For instance: | |
890 | ||
891 | virt<->cache | |
892 | Modifying an mmx register may involve | |
893 | scattering values across both FP and | |
894 | mmpx specific parts of a buffer | |
895 | ||
896 | cache<->target | |
897 | When writing back a SP it may need to | |
898 | both be written to both SP and USP. | |
899 | ||
900 | ||
901 | Hmm, | |
902 | ||
903 | Rather than let this like the last time it was discussed, just slip, I'm | |
904 | first going to add this e-mail (+ references) to TODO. I'd then like to | |
905 | sketch out a broad strategy I think could get us there. | |
906 | ||
907 | ||
908 | First thing I'd suggest is separating out the ``extern registers[]'' | |
909 | code so that we can at least identify what is using it. At present | |
910 | things are scattered across many files. That way we can at least | |
911 | pretend that there is a cache instead of a global array :-) | |
912 | ||
913 | I'd then suggest someone putting up a proposal for the pseudo-reg / | |
914 | high-level side interface so that code can be adopted to it. For old | |
915 | code, initially a blanket rename of write_register_bytes() to | |
916 | deprecated_write_register_bytes() would help. | |
917 | ||
918 | Following that would, finaly be the corresponding changes to the target. | |
bc9e5bbf AC |
919 | |
920 | -- | |
921 | ||
67edb2c6 AC |
922 | Check that GDB can handle all BFD architectures (Andrew Cagney) |
923 | ||
924 | There should be a test that checks that BFD/GDB are in sync with | |
925 | regard to architecture changes. Something like a test that first | |
926 | queries GDB for all supported architectures and then feeds each back | |
927 | to GDB.. Anyone interested in learning how to write tests? :-) | |
928 | ||
929 | -- | |
930 | ||
4afc966c AC |
931 | Architectural Change: Multi-arch et al. |
932 | ======================================= | |
2a00c9ce | 933 | |
4afc966c AC |
934 | The long term objective is to remove all assumptions that there is a |
935 | single target with a single address space with a single instruction | |
936 | set architecture and single application binary interface. | |
2a00c9ce | 937 | |
4afc966c AC |
938 | This is an ongoing effort. The first milestone is to enable |
939 | ``multi-arch'' where by all architectural decisions are made at | |
940 | runtime. | |
7ae38352 | 941 | |
4afc966c AC |
942 | It should be noted that ``gdbarch'' is really ``gdbabi'' and |
943 | ``gdbisa''. Once things are multi-arched breaking that down correctly | |
944 | will become much easier. | |
7ae38352 AC |
945 | |
946 | -- | |
947 | ||
4afc966c | 948 | GDBARCH cleanup (Andrew Cagney) |
7ae38352 | 949 | |
4afc966c AC |
950 | The non-generated parts of gdbarch.{sh,h,c} should be separated out |
951 | into arch-utils.[hc]. | |
952 | ||
953 | Document that gdbarch_init_ftype could easily fail because it didn't | |
954 | identify an architecture. | |
ed952ac5 AC |
955 | |
956 | -- | |
957 | ||
4afc966c | 958 | Fix BELIEVE_PPC_PROMOTION. Change it to BELIEVE_PPC_PROMOTION_P? |
ed952ac5 | 959 | |
4afc966c AC |
960 | At present there is still #ifdef BELIEVE_PPC_PROMOTION code in the |
961 | symtab file. | |
ed952ac5 | 962 | |
4afc966c AC |
963 | -- |
964 | ||
8e6a3c35 AC |
965 | Fix target_signal_from_host() etc. |
966 | ||
967 | The name is wrong for starters. ``target_signal'' should probably be | |
968 | ``gdb_signal''. ``from_host'' should be ``from_target_signal''. | |
969 | After that it needs to be multi-arched and made independant of any | |
970 | host signal numbering. | |
971 | ||
972 | -- | |
973 | ||
4afc966c AC |
974 | Update ALPHA so that it uses ``struct frame_extra_info'' instead of |
975 | EXTRA_FRAME_INFO. | |
976 | ||
977 | This is a barrier to replacing mips_extra_func_info with something | |
978 | that works with multi-arch. | |
7ae38352 AC |
979 | |
980 | -- | |
981 | ||
4afc966c AC |
982 | Multi-arch mips_extra_func_info. |
983 | ||
984 | This first needs the alpha to be updated so that it uses ``struct | |
985 | frame_extra_info''. | |
7ae38352 AC |
986 | |
987 | -- | |
988 | ||
4afc966c | 989 | Rationalize TARGET_SINGLE_FORMAT and TARGET_SINGLE_BIT et al. |
7ae38352 | 990 | |
4afc966c | 991 | Surely one of them is redundant. |
7ae38352 AC |
992 | |
993 | -- | |
994 | ||
4afc966c | 995 | Convert ALL architectures to MULTI-ARCH. |
7ae38352 AC |
996 | |
997 | -- | |
998 | ||
999 | Select the initial multi-arch ISA / ABI based on --target or similar. | |
1000 | ||
1001 | At present the default is based on what ever is first in the BFD | |
1002 | archures table. It should be determined based on the ``--target=...'' | |
1003 | name. | |
1004 | ||
1005 | -- | |
1006 | ||
bf64bfd6 AC |
1007 | Make MIPS pure multi-arch. |
1008 | ||
1009 | It is only at the multi-arch enabled stage. | |
1010 | ||
1011 | -- | |
1012 | ||
7ae38352 AC |
1013 | Truly multi-arch. |
1014 | ||
1015 | Enable the code to recognize --enable-targets=.... like BINUTILS does. | |
1016 | ||
4afc966c AC |
1017 | Can the tm.h and nm.h files be eliminated by multi-arch. |
1018 | ||
7ae38352 AC |
1019 | -- |
1020 | ||
4afc966c AC |
1021 | Architectural Change: MI, LIBGDB and scripting languages |
1022 | ======================================================== | |
7ae38352 | 1023 | |
4afc966c AC |
1024 | See also architectural changes related to the event loop. LIBGDB |
1025 | can't be finished until there is a generic event loop being used by | |
1026 | all targets. | |
1027 | ||
1028 | The long term objective is it to be possible to integrate GDB into | |
1029 | scripting languages. | |
7ae38352 AC |
1030 | |
1031 | -- | |
1032 | ||
4afc966c | 1033 | Implement generic ``(gdb) commmand > file'' |
7ae38352 | 1034 | |
4afc966c AC |
1035 | Once everything is going through ui_file it should be come fairly |
1036 | easy. | |
1037 | ||
1038 | http://sourceware.cygnus.com/ml/gdb/2000-04/msg00104.html | |
1039 | ||
1040 | -- | |
1041 | ||
1042 | Replace gdb_stdtarg with gdb_targout (and possibly gdb_targerr). | |
1043 | ||
1044 | gdb_stdtarg is easily confused with gdb_stdarg. | |
1045 | ||
1046 | -- | |
1047 | ||
1048 | Extra ui_file methods - dump. | |
1049 | ||
26099b4a | 1050 | Very useful for whitebox testing. |
4afc966c AC |
1051 | |
1052 | -- | |
1053 | ||
1054 | Eliminate error_begin(). | |
1055 | ||
1056 | With ui_file, there is no need for the statefull error_begin () | |
1057 | function. | |
1058 | ||
1059 | -- | |
1060 | ||
1061 | Send normal output to gdb_stdout. | |
1062 | Send error messages to gdb_stderror. | |
1063 | Send debug and log output log gdb_stdlog. | |
1064 | ||
1065 | GDB still contains many cases where (f)printf or printf_filtered () is | |
1066 | used when it should be sending the messages to gdb_stderror or | |
1067 | gdb_stdlog. The thought of #defining printf to something has crossed | |
1068 | peoples minds ;-) | |
7ae38352 AC |
1069 | |
1070 | -- | |
1071 | ||
1072 | Re-do GDB's output pager. | |
1073 | ||
1074 | GDB's output pager still relies on people correctly using *_filtered | |
1075 | for gdb_stdout and *_unfiltered for gdb_stdlog / gdb_stderr. | |
1076 | Hopefully, with all normal output going to gdb_stdout, the pager can | |
1077 | just look at the ui_file that the output is on and then use that to | |
1078 | decide what to do about paging. Sounds good in theory. | |
1079 | ||
1080 | -- | |
1081 | ||
4afc966c AC |
1082 | Check/cleanup MI documentation. |
1083 | ||
1084 | The list of commands specified in the documentation needs to be | |
1085 | checked against the mi-cmds.c table in a mechanical way (so that they | |
1086 | two can be kept up-to-date). | |
1087 | ||
1088 | -- | |
1089 | ||
1090 | Convert MI into libgdb | |
1091 | ||
1092 | MI provides a text interface into what should be many of the libgdb | |
1093 | functions. The implementation of those functions should be separated | |
1094 | into the MI interface and the functions proper. Those functions being | |
1095 | moved to gdb/lib say. | |
1096 | ||
1097 | -- | |
1098 | ||
1099 | Create libgdb.h | |
1100 | ||
1101 | The first part can already be found in defs.h. | |
1102 | ||
1103 | -- | |
1104 | ||
1105 | MI's input does not use buffering. | |
1106 | ||
1107 | At present the MI interface reads raw characters of from an unbuffered | |
1108 | FD. This is to avoid several nasty buffer/race conditions. That code | |
1109 | should be changed so that it registers its self with the event loop | |
1110 | (on the input FD) and then push commands up to MI as they arrive. | |
1111 | ||
1112 | The serial code already does this. | |
1113 | ||
1114 | -- | |
1115 | ||
1116 | Make MI interface accessable from existing CLI. | |
1117 | ||
1118 | -- | |
1119 | ||
1120 | Add a breakpoint-edit command to MI. | |
1121 | ||
1122 | It would be similar to MI's breakpoint create but would apply to an | |
1123 | existing breakpoint. It saves the need to delete/create breakpoints | |
1124 | when ever they are changed. | |
1125 | ||
1126 | -- | |
1127 | ||
1128 | Add directory path to MI breakpoint. | |
1129 | ||
1130 | That way the GUI's task of finding the file within which the | |
1131 | breakpoint was set is simplified. | |
1132 | ||
1133 | -- | |
1134 | ||
1135 | Add a mechanism to reject certain expression classes to MI | |
7ae38352 AC |
1136 | |
1137 | There are situtations where you don't want GDB's expression | |
1138 | parser/evaluator to perform inferior function calls or variable | |
4afc966c AC |
1139 | assignments. A way of restricting the expression parser so that such |
1140 | operations are not accepted would be very helpful. | |
7ae38352 AC |
1141 | |
1142 | -- | |
1143 | ||
1144 | Remove sideffects from libgdb breakpoint create function. | |
1145 | ||
1146 | The user can use the CLI to create a breakpoint with partial | |
1147 | information - no file (gdb would use the file from the last | |
1148 | breakpoint). | |
1149 | ||
1150 | The libgdb interface currently affects that environment which can lead | |
1151 | to confusion when a user is setting breakpoints via both the MI and | |
1152 | the CLI. | |
1153 | ||
1154 | This is also a good example of how getting the CLI ``right'' will be | |
1155 | hard. | |
1156 | ||
1157 | -- | |
1158 | ||
4afc966c | 1159 | Move gdb_lasterr to ui_out? |
7ae38352 | 1160 | |
4afc966c AC |
1161 | The way GDB throws errors and records them needs a re-think. ui_out |
1162 | handles the correct output well. It doesn't resolve what to do with | |
1163 | output / error-messages when things go wrong. | |
7ae38352 | 1164 | |
97c3646f AC |
1165 | -- |
1166 | ||
1167 | do_setshow_command contains a 1024 byte buffer. | |
1168 | ||
1169 | The function assumes that there will never be any more than 1024 bytes | |
1170 | of enum. It should use mem_file. | |
1171 | ||
1172 | -- | |
1173 | ||
1174 | Should struct cmd_list_element . completer take the command as an | |
1175 | argument? | |
1176 | ||
1177 | -- | |
1178 | ||
1179 | Should the bulk of top.c:line_completion_function() be moved to | |
1180 | command.[hc]? complete_on_cmdlist() and complete_on_enums() could | |
1181 | then be made private. | |
1182 | ||
1183 | -- | |
1184 | ||
1185 | top.c (execute_command): Should a command being valid when the target | |
1186 | is running be made an attribute (predicate) to the command rather than | |
1187 | an explicit set of tests. | |
1188 | ||
1189 | -- | |
1190 | ||
1191 | top.c (execute_command): Should the bulk of this function be moved | |
1192 | into command.[hc] so that top.c doesn't grub around in the command | |
1193 | internals? | |
1194 | ||
4afc966c AC |
1195 | -- |
1196 | ||
1197 | Architectural Change: Async | |
1198 | =========================== | |
1199 | ||
1200 | While GDB uses an event loop when prompting the user for input. That | |
1201 | event loop is not exploited by targets when they allow the target | |
1202 | program to continue. Typically targets still block in (target_wait()) | |
1203 | until the program again halts. | |
1204 | ||
1205 | The closest a target comes to supporting full asynchronous mode are | |
1206 | the remote targets ``async'' and ``extended-async''. | |
7ae38352 AC |
1207 | |
1208 | -- | |
1209 | ||
4afc966c | 1210 | Asynchronous expression evaluator |
7ae38352 | 1211 | |
4afc966c | 1212 | Inferior function calls hang GDB. |
7ae38352 AC |
1213 | |
1214 | -- | |
1215 | ||
1216 | Fix implementation of ``target xxx''. | |
1217 | ||
1218 | At present when the user specifies ``target xxxx'', the CLI maps that | |
1219 | directly onto a target open method. It is then assumed that the | |
1220 | target open method should do all sorts of complicated things as this | |
1221 | is the only chance it has. Check how the various remote targets | |
1222 | duplicate the target operations. Check also how the various targets | |
1223 | behave differently for purely arbitrary reasons. | |
1224 | ||
1225 | What should happen is that ``target xxxx'' should call a generic | |
1226 | ``target'' function and that should then co-ordinate the opening of | |
1227 | ``xxxx''. This becomes especially important when you're trying to | |
1228 | open an asynchronous target that may need to perform background tasks | |
1229 | as part of the ``attach'' phase. | |
1230 | ||
1231 | Unfortunatly, due to limitations in the old/creaking command.h | |
1232 | interface, that isn't possible. The function being called isn't told | |
1233 | of the ``xxx'' or any other context information. | |
1234 | ||
1235 | Consequently a precursor to fixing ``target xxxx'' is to clean up the | |
1236 | CLI code so that it passes to the callback function (attatched to a | |
1237 | command) useful information such as the actual command and a context | |
1238 | for that command. Other changes such as making ``struct command'' | |
1239 | opaque may also help. | |
1240 | ||
53904c9e AC |
1241 | See also: |
1242 | http://sourceware.cygnus.com/ml/gdb-patches/2000-06/msg00062.html | |
1243 | ||
7ae38352 AC |
1244 | -- |
1245 | ||
4afc966c AC |
1246 | Make "target xxx" command interruptible. |
1247 | ||
1248 | As things become async this becomes possible. A target would start | |
1249 | the connect and then return control to the event loop. A cntrl-c | |
1250 | would notify the target that the operation is to be abandoned and the | |
1251 | target code could respond. | |
7ae38352 AC |
1252 | |
1253 | -- | |
1254 | ||
4afc966c AC |
1255 | Add a "suspend" subcommand of the "continue" command to suspend gdb |
1256 | while continuing execution of the subprocess. Useful when you are | |
1257 | debugging servers and you want to dodge out and initiate a connection | |
1258 | to a server running under gdb. | |
1259 | ||
1260 | [hey async!!] | |
7ae38352 | 1261 | |
2a00c9ce AC |
1262 | -- |
1263 | ||
26099b4a AC |
1264 | TODO FAQ |
1265 | ======== | |
1266 | ||
1267 | Frequently requested but not approved requests. | |
1268 | ||
1269 | -- | |
1270 | ||
1271 | Eliminate unused argument warnings using ATTRIBUTE_UNUSED. | |
1272 | ||
1273 | The benefits on this one are thought to be marginal - GDBs design | |
1274 | means that unused parameters are very common. GCC 3.0 will also | |
1275 | include the option -Wno-unused-parameter which means that ``-Wall | |
1276 | -Wno-unused-parameters -Werror'' can be specified. | |
1277 | ||
1278 | -- | |
1279 | ||
1280 | ||
1281 | ||
2a00c9ce AC |
1282 | Legacy Wish List |
1283 | ================ | |
1284 | ||
1285 | This list is not up to date, and opinions vary about the importance or | |
1286 | even desirability of some of the items. If you do fix something, it | |
1287 | always pays to check the below. | |
1288 | ||
1289 | -- | |
c906108c | 1290 | |
b83266a0 SS |
1291 | @c This does not work (yet if ever). FIXME. |
1292 | @c @item --parse=@var{lang} @dots{} | |
1293 | @c Configure the @value{GDBN} expression parser to parse the listed languages. | |
1294 | @c @samp{all} configures @value{GDBN} for all supported languages. To get a | |
1295 | @c list of all supported languages, omit the argument. Without this | |
1296 | @c option, @value{GDBN} is configured to parse all supported languages. | |
1297 | ||
7ae38352 | 1298 | -- |
c906108c SS |
1299 | |
1300 | START_INFERIOR_TRAPS_EXPECTED need never be defined to 2, since that | |
1301 | is its default value. Clean this up. | |
1302 | ||
7ae38352 AC |
1303 | -- |
1304 | ||
c906108c SS |
1305 | It should be possible to use symbols from shared libraries before we know |
1306 | exactly where the libraries will be loaded. E.g. "b perror" before running | |
1307 | the program. This could maybe be done as an extension of the "breakpoint | |
1308 | re-evaluation" after new symbols are loaded. | |
1309 | ||
7ae38352 AC |
1310 | -- |
1311 | ||
c906108c SS |
1312 | Make single_step() insert and remove breakpoints in one operation. |
1313 | ||
26099b4a AC |
1314 | [If this is talking about having single_step() insert the breakpoints, |
1315 | run the target then pull the breakpoints then it is wrong. The | |
1316 | function has to return as control has to eventually be passed back to | |
1317 | the main event loop.] | |
1318 | ||
7ae38352 AC |
1319 | -- |
1320 | ||
c906108c SS |
1321 | Speed up single stepping by avoiding extraneous ptrace calls. |
1322 | ||
7ae38352 AC |
1323 | -- |
1324 | ||
c906108c SS |
1325 | Speed up single stepping by not inserting and removing breakpoints |
1326 | each time the inferior starts and stops. | |
1327 | ||
1328 | Breakpoints should not be inserted and deleted all the time. Only the | |
1329 | one(s) there should be removed when we have to step over one. Support | |
1330 | breakpoints that don't have to be removed to step over them. | |
1331 | ||
7ae38352 | 1332 | [this has resulted in numerous debates. The issue isn't clear cut] |
c906108c | 1333 | |
7ae38352 | 1334 | -- |
c906108c SS |
1335 | |
1336 | Provide "voodoo" debugging of core files. This creates a zombie | |
1337 | process as a child of the debugger, and loads it up with the data, | |
1338 | stack, and regs of the core file. This allows you to call functions | |
1339 | in the executable, to manipulate the data in the core file. | |
1340 | ||
7ae38352 | 1341 | [you wish] |
c906108c | 1342 | |
7ae38352 | 1343 | -- |
c906108c | 1344 | |
7ae38352 | 1345 | GDB reopens the source file on every line, as you "next" through it. |
c906108c | 1346 | |
7ae38352 | 1347 | [still true? I've a memory of this being fixed] |
c906108c | 1348 | |
7ae38352 | 1349 | -- |
c906108c SS |
1350 | |
1351 | Perhaps "i source" should take an argument like that of "list". | |
1352 | ||
7ae38352 | 1353 | -- |
c906108c SS |
1354 | |
1355 | Remove "at 0xnnnn" from the "b foo" response, if `print address off' and if | |
1356 | it matches the source line indicated. | |
1357 | ||
7ae38352 | 1358 | -- |
c906108c | 1359 | |
7ae38352 | 1360 | The prompt at end of screen should accept space as well as CR. |
c906108c | 1361 | |
7ae38352 | 1362 | -- |
c906108c SS |
1363 | |
1364 | Backtrace should point out what the currently selected frame is, in | |
1365 | its display, perhaps showing "@3 foo (bar, ...)" or ">3 foo (bar, | |
1366 | ...)" rather than "#3 foo (bar, ...)". | |
1367 | ||
7ae38352 AC |
1368 | -- |
1369 | ||
c906108c SS |
1370 | "i program" should work for core files, and display more info, like what |
1371 | actually caused it to die. | |
1372 | ||
7ae38352 AC |
1373 | -- |
1374 | ||
c906108c SS |
1375 | "x/10i" should shorten the long name, if any, on subsequent lines. |
1376 | ||
7ae38352 | 1377 | -- |
c906108c SS |
1378 | |
1379 | "next" over a function that longjumps, never stops until next time you happen | |
1380 | to get to that spot by accident. E.g. "n" over execute_command which has | |
1381 | an error. | |
1382 | ||
7ae38352 AC |
1383 | -- |
1384 | ||
c906108c SS |
1385 | "set zeroprint off", don't bother printing members of structs which |
1386 | are entirely zero. Useful for those big structs with few useful | |
1387 | members. | |
1388 | ||
7ae38352 AC |
1389 | -- |
1390 | ||
c906108c SS |
1391 | GDB does four ioctl's for every command, probably switching terminal modes |
1392 | to/from inferior or for readline or something. | |
1393 | ||
7ae38352 AC |
1394 | -- |
1395 | ||
c906108c SS |
1396 | terminal_ours versus terminal_inferior: cache state. Switch should be a noop |
1397 | if the state is the same, too. | |
1398 | ||
7ae38352 | 1399 | -- |
c906108c SS |
1400 | |
1401 | "i frame" shows wrong "arglist at" location, doesn't show where the args | |
1402 | should be found, only their actual values. | |
1403 | ||
7ae38352 AC |
1404 | -- |
1405 | ||
c906108c SS |
1406 | There should be a way for "set" commands to validate the new setting |
1407 | before it takes effect. | |
1408 | ||
7ae38352 | 1409 | -- |
c906108c SS |
1410 | |
1411 | "ena d" is ambiguous, why? "ena delete" seems to think it is a command! | |
1412 | ||
7ae38352 | 1413 | -- |
c906108c SS |
1414 | |
1415 | i line VAR produces "Line number not known for symbol ``var''.". I | |
1416 | thought we were stashing that info now! | |
1417 | ||
7ae38352 AC |
1418 | -- |
1419 | ||
c906108c SS |
1420 | We should be able to write to random files at hex offsets like adb. |
1421 | ||
7ae38352 AC |
1422 | -- |
1423 | ||
7ae38352 AC |
1424 | [elena - delete this] |
1425 | ||
c906108c SS |
1426 | Handle add_file with separate text, data, and bss addresses. Maybe |
1427 | handle separate addresses for each segment in the object file? | |
1428 | ||
7ae38352 AC |
1429 | -- |
1430 | ||
1431 | [Jimb/Elena delete this one] | |
1432 | ||
c906108c SS |
1433 | Handle free_named_symtab to cope with multiply-loaded object files |
1434 | in a dynamic linking environment. Should remember the last copy loaded, | |
1435 | but not get too snowed if it finds references to the older copy. | |
1436 | ||
7ae38352 AC |
1437 | -- |
1438 | ||
1439 | [elena delete this also] | |
c906108c SS |
1440 | |
1441 | Remove all references to: | |
1442 | text_offset | |
1443 | data_offset | |
1444 | text_data_start | |
1445 | text_end | |
1446 | exec_data_offset | |
1447 | ... | |
1448 | now that we have BFD. All remaining are in machine dependent files. | |
1449 | ||
7ae38352 | 1450 | -- |
c906108c SS |
1451 | |
1452 | Re-organize help categories into things that tend to fit on a screen | |
1453 | and hang together. | |
1454 | ||
7ae38352 | 1455 | -- |
c906108c SS |
1456 | |
1457 | Add in commands like ADB's for searching for patterns, etc. We should | |
1458 | be able to examine and patch raw unsymboled binaries as well in gdb as | |
1459 | we can in adb. (E.g. increase the timeout in /bin/login without source). | |
1460 | ||
7ae38352 AC |
1461 | [actually, add ADB interface :-] |
1462 | ||
1463 | -- | |
c906108c SS |
1464 | |
1465 | When doing "step" or "next", if a few lines of source are skipped between | |
1466 | the previous line and the current one, print those lines, not just the | |
1467 | last line of a multiline statement. | |
1468 | ||
7ae38352 | 1469 | -- |
c906108c SS |
1470 | |
1471 | Handling of "&" address-of operator needs some serious overhaul | |
1472 | for ANSI C and consistency on arrays and functions. | |
1473 | For "float point[15];": | |
1474 | ptype &point[4] ==> Attempt to take address of non-lvalue. | |
1475 | For "char *malloc();": | |
1476 | ptype malloc ==> "char *()"; should be same as | |
1477 | ptype &malloc ==> "char *(*)()" | |
1478 | call printf ("%x\n", malloc) ==> weird value, should be same as | |
1479 | call printf ("%x\n", &malloc) ==> correct value | |
1480 | ||
7ae38352 AC |
1481 | -- |
1482 | ||
c906108c SS |
1483 | Fix dbxread.c symbol reading in the presence of interrupts. It |
1484 | currently leaves a cleanup to blow away the entire symbol table when a | |
1485 | QUIT occurs. (What's wrong with that? -kingdon, 28 Oct 1993). | |
1486 | ||
7ae38352 AC |
1487 | [I suspect that the grype was that, on a slow system, you might want |
1488 | to cntrl-c and get just half the symbols and then load the rest later | |
1489 | - scary to be honest] | |
1490 | ||
1491 | -- | |
1492 | ||
c906108c SS |
1493 | Mipsread.c reads include files depth-first, because the dependencies |
1494 | in the psymtabs are way too inclusive (it seems to me). Figure out what | |
1495 | really depends on what, to avoid recursing 20 or 30 times while reading | |
1496 | real symtabs. | |
1497 | ||
7ae38352 AC |
1498 | -- |
1499 | ||
c906108c SS |
1500 | value_add() should be subtracting the lower bound of arrays, if known, |
1501 | and possibly checking against the upper bound for error reporting. | |
1502 | ||
7ae38352 | 1503 | -- |
c906108c SS |
1504 | |
1505 | When listing source lines, check for a preceding \n, to verify that | |
1506 | the file hasn't changed out from under us. | |
1507 | ||
7ae38352 AC |
1508 | [fixed by some other means I think. That hack wouldn't actually work |
1509 | reliably - the file might move such that another \n appears. ] | |
c906108c | 1510 | |
7ae38352 | 1511 | -- |
c906108c SS |
1512 | |
1513 | Get all the remote systems (where the protocol allows it) to be able to | |
1514 | stop the remote system when the GDB user types ^C (like remote.c | |
1515 | does). For ebmon, use ^Ak. | |
1516 | ||
7ae38352 AC |
1517 | -- |
1518 | ||
c906108c SS |
1519 | Possible feature: A version of the "disassemble" command which shows |
1520 | both source and assembly code ("set symbol-filename on" is a partial | |
1521 | solution). | |
1522 | ||
7ae38352 AC |
1523 | [has this been done? It was certainly done for MI and GDBtk] |
1524 | ||
1525 | -- | |
1526 | ||
c906108c SS |
1527 | investigate "x/s 0" (right now stops early) (I think maybe GDB is |
1528 | using a 0 address for bad purposes internally). | |
1529 | ||
7ae38352 AC |
1530 | -- |
1531 | ||
c906108c SS |
1532 | Make "info path" and path_command work again (but independent of the |
1533 | environment either of gdb or that we'll pass to the inferior). | |
1534 | ||
7ae38352 AC |
1535 | -- |
1536 | ||
c906108c SS |
1537 | Make GDB understand the GCC feature for putting octal constants in |
1538 | enums. Make it so overflow on an enum constant does not error_type | |
1539 | the whole type. Allow arbitrarily large enums with type attributes. | |
1540 | Put all this stuff in the testsuite. | |
1541 | ||
7ae38352 AC |
1542 | -- |
1543 | ||
c906108c SS |
1544 | Make TYPE_CODE_ERROR with a non-zero TYPE_LENGTH more useful (print |
1545 | the value in hex; process type attributes). Add this to the | |
1546 | testsuite. This way future compilers can add new types and old | |
1547 | versions of GDB can do something halfway reasonable. | |
1548 | ||
7ae38352 | 1549 | -- |
c906108c SS |
1550 | |
1551 | Fix mdebugread.c:parse_type to do fundamental types right (see | |
1552 | rs6000_builtin_type in stabsread.c for what "right" is--the point is | |
1553 | that the debug format fixes the sizes of these things and it shouldn't | |
1554 | depend on stuff like TARGET_PTR_BIT and so on. For mdebug, there seem | |
1555 | to be separate bt* codes for 64 bit and 32 bit things, and GDB should | |
1556 | be aware of that). Also use a switch statement for clarity and speed. | |
1557 | ||
7ae38352 AC |
1558 | -- |
1559 | ||
c906108c SS |
1560 | Investigate adding symbols in target_load--some targets do, some |
1561 | don't. | |
1562 | ||
7ae38352 AC |
1563 | -- |
1564 | ||
c906108c SS |
1565 | Put dirname in psymtabs and change lookup*symtab to use dirname (so |
1566 | /foo/bar.c works whether compiled by cc /foo/bar.c, or cd /foo; cc | |
1567 | bar.c). | |
1568 | ||
7ae38352 AC |
1569 | -- |
1570 | ||
c906108c SS |
1571 | Merge xcoffread.c and coffread.c. Use breakpoint_re_set instead of |
1572 | fixup_breakpoints. | |
1573 | ||
7ae38352 | 1574 | -- |
c906108c SS |
1575 | |
1576 | Make a watchpoint which contains a function call an error (it is | |
1577 | broken now, making it work is probably not worth the effort). | |
1578 | ||
7ae38352 | 1579 | -- |
c906108c SS |
1580 | |
1581 | New test case based on weird.exp but in which type numbers are not | |
1582 | renumbered (thus multiply defining a type). This currently causes an | |
1583 | infinite loop on "p v_comb". | |
1584 | ||
7ae38352 | 1585 | -- |
c906108c | 1586 | |
7ae38352 | 1587 | [Hey! Hint Hint Delete Delete!!!] |
c906108c SS |
1588 | |
1589 | Fix 386 floating point so that floating point registers are real | |
1590 | registers (but code can deal at run-time if they are missing, like | |
1591 | mips and 68k). This would clean up "info float" and related stuff. | |
1592 | ||
7ae38352 | 1593 | -- |
c906108c SS |
1594 | |
1595 | gcc -g -c enummask.c then gdb enummask.o, then "p v". GDB complains | |
1596 | about not being able to access memory location 0. | |
1597 | ||
1598 | -------------------- enummask.c | |
1599 | enum mask | |
1600 | { | |
1601 | ANIMAL = 0, | |
1602 | VEGETABLE = 1, | |
1603 | MINERAL = 2, | |
1604 | BASIC_CATEGORY = 3, | |
1605 | ||
1606 | WHITE = 0, | |
1607 | BLUE = 4, | |
1608 | GREEN = 8, | |
1609 | BLACK = 0xc, | |
1610 | COLOR = 0xc, | |
1611 | ||
1612 | ALIVE = 0x10, | |
1613 | ||
1614 | LARGE = 0x20 | |
1615 | } v; | |
1616 | ||
7ae38352 AC |
1617 | -- |
1618 | ||
c906108c SS |
1619 | If try to modify value in file with "set write off" should give |
1620 | appropriate error not "cannot access memory at address 0x65e0". | |
1621 | ||
7ae38352 | 1622 | -- |
c906108c | 1623 | |
c906108c SS |
1624 | Allow core file without exec file on RS/6000. |
1625 | ||
7ae38352 AC |
1626 | -- |
1627 | ||
c906108c SS |
1628 | Make sure "shell" with no arguments works right on DOS. |
1629 | ||
7ae38352 AC |
1630 | -- |
1631 | ||
c906108c SS |
1632 | Make gdb.ini (as well as .gdbinit) be checked on all platforms, so |
1633 | the same directory can be NFS-mounted on unix or DOS, and work the | |
1634 | same way. | |
1635 | ||
7ae38352 AC |
1636 | -- |
1637 | ||
1638 | [Is this another delete???] | |
c906108c SS |
1639 | |
1640 | Get SECT_OFF_TEXT stuff out of objfile_relocate (might be needed to | |
1641 | get RS/6000 to work right, might not be immediately relevant). | |
1642 | ||
7ae38352 | 1643 | -- |
c906108c SS |
1644 | |
1645 | Work out some kind of way to allow running the inferior to be done as | |
1646 | a sub-execution of, eg. breakpoint command lists. Currently running | |
1647 | the inferior interupts any command list execution. This would require | |
1648 | some rewriting of wait_for_inferior & friends, and hence should | |
1649 | probably be done in concert with the above. | |
1650 | ||
7ae38352 AC |
1651 | -- |
1652 | ||
c906108c SS |
1653 | Add function arguments to gdb user defined functions. |
1654 | ||
7ae38352 AC |
1655 | -- |
1656 | ||
c906108c SS |
1657 | Add convenience variables that refer to exec file, symbol file, |
1658 | selected frame source file, selected frame function, selected frame | |
1659 | line number, etc. | |
1660 | ||
7ae38352 AC |
1661 | -- |
1662 | ||
c906108c SS |
1663 | Modify the handling of symbols grouped through BINCL/EINCL stabs to |
1664 | allocate a partial symtab for each BINCL/EINCL grouping. This will | |
1665 | seriously decrease the size of inter-psymtab dependencies and hence | |
1666 | lessen the amount that needs to be read in when a new source file is | |
1667 | accessed. | |
1668 | ||
7ae38352 | 1669 | -- |
c906108c | 1670 | |
c906108c SS |
1671 | Add a command for searching memory, a la adb. It specifies size, |
1672 | mask, value, start address. ADB searches until it finds it or hits | |
1673 | an error (or is interrupted). | |
1674 | ||
7ae38352 AC |
1675 | -- |
1676 | ||
b83266a0 SS |
1677 | Remove the range and type checking code and documentation, if not |
1678 | going to implement. | |
1679 | ||
c906108c SS |
1680 | # Local Variables: |
1681 | # mode: text | |
1682 | # End: |