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 | |
2a00c9ce 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 | |
bc9e5bbf | 14 | (The names in paren indicate people that posted the original problem.) |
138f88c0 AC |
15 | |
16 | -- | |
17 | ||
bc9e5bbf | 18 | GDB doesn't build under IRIX6.4 |
c906108c | 19 | |
bc9e5bbf AC |
20 | Benjamin Gamsa wrote: |
21 | ||
22 | Has anyone successfully built the latest (from cvs) gdb on IRIX6.4 or | |
23 | later? The first problem I hit is that proc-api.c includes | |
24 | sys/user.h, which no longer exists under IRIX6.4. If I comment out | |
25 | that include, the next problem I hit is that PIOCGETPR and PIOCGETU | |
26 | are no longer defined in IRIX6.4 (presumably related to the | |
27 | disappearance of user.h). | |
138f88c0 AC |
28 | |
29 | -- | |
30 | ||
4fd99b5a AC |
31 | The BFD directory requires bug-fixed AUTOMAKE et.al. |
32 | ||
33 | AUTOMAKE 1.4 incorrectly set the TEXINPUTS environment variable. It | |
34 | contained the full path to texinfo.tex when it should have only | |
35 | contained the directory. The bug has been fixed in the current | |
36 | AUTOMAKE sources. Automake snapshots can be found in: | |
37 | ftp://sourceware.cygnus.com/pub/gdb/snapshots | |
38 | and ftp://sourceware.cygnus.com/pub/binutils | |
39 | ||
40 | -- | |
41 | ||
bc9e5bbf AC |
42 | gdb-cvs fails to build on freebsd-elf |
43 | http://sourceware.cygnus.com/ml/gdb/2000-04/msg00004.html | |
138f88c0 | 44 | |
bc9e5bbf AC |
45 | Either the FreeBSD group need to contribute their local GDB changes |
46 | back to the master sources or someone needs to provides a new | |
47 | (clean-room) implementation. Since the former involves a fairly | |
48 | complicated assignment the latter may be easier. [cagney] | |
138f88c0 AC |
49 | |
50 | -- | |
51 | ||
bc9e5bbf AC |
52 | Generic: lin-thread cannot handle thread exit (Mark Kettenis, Michael |
53 | Snyder) http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00525.html | |
138f88c0 AC |
54 | |
55 | The thread_db assisted debugging code doesn't handle exiting threads | |
56 | properly, at least in combination with glibc 2.1.3 (the framework is | |
57 | there, just not the actual code). There are at least two problems | |
58 | that prevent this from working. | |
59 | ||
bc9e5bbf | 60 | As an additional reference point, the pre thread_db code did not work |
138f88c0 AC |
61 | either. |
62 | ||
63 | -- | |
64 | ||
65 | Java (Anthony Green, David Taylor) | |
66 | ||
bc9e5bbf AC |
67 | Anthony Green has a number of Java patches that did not make it into |
68 | the 5.0 release. | |
138f88c0 AC |
69 | |
70 | Patch: java tests | |
71 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00512.html | |
72 | ||
73 | Patch: java booleans | |
74 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00515.html | |
75 | ||
76 | Patch: handle N_MAIN stab | |
77 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00527.html | |
78 | ||
138f88c0 AC |
79 | -- |
80 | ||
81 | Pascal (Pierre Muller, David Taylor) | |
82 | ||
bc9e5bbf AC |
83 | Pierre Muller has contributed patches for adding Pascal Language |
84 | support to GDB. | |
67edb2c6 AC |
85 | |
86 | 2 pascal language patches inserted in database | |
87 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00521.html | |
88 | ||
bc9e5bbf AC |
89 | Indent -gnu ? |
90 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00496.html | |
138f88c0 AC |
91 | |
92 | -- | |
93 | ||
94 | GNU/Linux/x86 and random thread signals (and Solaris/SPARC but not | |
bc9e5bbf AC |
95 | Solaris/x86). |
96 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00336.html | |
138f88c0 AC |
97 | |
98 | Christopher Blizzard writes: | |
99 | ||
100 | So, I've done some more digging into this and it looks like Jim | |
101 | Kingdon has reported this problem in the past: | |
102 | ||
103 | http://sourceware.cygnus.com/ml/bug-gdb/1999-10/msg00058.html | |
104 | ||
105 | I can reproduce this problem both with and without Tom's patch. Has | |
106 | anyone seen this before? Maybe have a solution for it hanging around? | |
107 | :) | |
108 | ||
109 | There's a test case for this documented at: | |
110 | ||
111 | when debugging threaded applications you get extra SIGTRAPs | |
112 | http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=9565 | |
113 | ||
114 | [There should be a GDB testcase - cagney] | |
115 | ||
116 | -- | |
117 | ||
bc9e5bbf | 118 | Possible regressions with some devel GCCs. |
138f88c0 AC |
119 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00475.html |
120 | ||
bc9e5bbf AC |
121 | gcc-2.95.2 outputs a line note *before* the prologue (and one for the |
122 | closing brace after the epilogue, instead of before it, as it used to | |
123 | be). By disabling the RTL-style prologue generating mechanism | |
124 | (undocumented GCC option -mno-schedule-prologue), you get back the | |
125 | traditional behaviour. | |
126 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00510.html | |
138f88c0 | 127 | |
bc9e5bbf | 128 | This should now be fixed. |
138f88c0 AC |
129 | |
130 | -- | |
131 | ||
bc9e5bbf AC |
132 | RFD: infrun.c: No bpstat_stop_status call after proceed over break? |
133 | (Peter Schauer) | |
138f88c0 AC |
134 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00665.html |
135 | ||
bc9e5bbf AC |
136 | GDB misses watchpoint triggers after proceeding over a breakpoint on |
137 | x86 targets. | |
138 | ||
138f88c0 AC |
139 | -- |
140 | ||
67edb2c6 AC |
141 | x86 linux GDB and SIGALRM (???) |
142 | http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00803.html | |
143 | ||
bc9e5bbf AC |
144 | I know there are problems with single stepping through signal |
145 | handlers. These problems were present in 4.18. They were just masked | |
146 | because 4.18 failed to recognize signal handlers. Fixing it is not | |
147 | easy, and will require changes to handle_inferior_event(), that I | |
148 | prefer not to make before the 5.0 release. | |
67edb2c6 | 149 | |
bc9e5bbf | 150 | Mark |
67edb2c6 AC |
151 | |
152 | -- | |
153 | ||
9d6d78f2 AC |
154 | Revised UDP support (was: Re: [Fwd: [patch] UDP transport support]) |
155 | http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00000.html | |
156 | ||
bc9e5bbf AC |
157 | (Broken) support for GDB's remote protocol across UDP is to be |
158 | included in the follow-on release. | |
138f88c0 | 159 | |
b2f4b24d AC |
160 | -- |
161 | ||
162 | Can't build IRIX -> arm GDB. | |
163 | http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00356.html | |
164 | ||
165 | David Whedon writes: | |
166 | > Now I'm building for an embedded arm target. If there is a way of turning | |
167 | > remote-rdi off, I couldn't find it. It looks like it gets built by default | |
168 | > in gdb/configure.tgt(line 58) Anyway, the build dies in | |
169 | > gdb/rdi-share/unixcomm.c. SERPORT1 et. al. never get defined because we | |
170 | > aren't one of the architectures supported. | |
171 | ||
6bc37a96 AC |
172 | -- |
173 | ||
174 | Problem with weak functions | |
175 | http://sourceware.cygnus.com/ml/gdb/2000-05/msg00060.html | |
176 | ||
177 | Dan Nicolaescu writes: | |
178 | > It seems that gdb-4.95.1 does not display correctly the function when | |
179 | > stoping in weak functions. | |
180 | > | |
181 | > It stops in a function that is defined as weak, not in the function | |
182 | > that is actualy run... | |
b2f4b24d | 183 | |
3fffcb5e AC |
184 | -- |
185 | ||
186 | GDB5 TOT on unixware 7 | |
187 | http://sourceware.cygnus.com/ml/gdb/2000-04/msg00119.html | |
188 | ||
189 | Robert Lipe writes: | |
190 | > I just spun the top of tree of the GDB5 branch on UnixWare 7. As a | |
191 | > practical matter, the current thread support is somewhat more annoying | |
192 | > than when GDB was thread-unaware. | |
193 | ||
138f88c0 AC |
194 | -- |
195 | ||
2a00c9ce AC |
196 | Code cleanups |
197 | ============= | |
bc9e5bbf AC |
198 | |
199 | The following code cleanups are planned for the follow-on release to | |
200 | GDB 5.0. | |
201 | ||
202 | -- | |
203 | ||
204 | ATTRIBUTE_UNUSED | |
205 | ||
206 | The need for this as almost been eliminated. The next version of GCC | |
207 | (assuming cagney gets the relevant patch committed) will be able to | |
208 | supress unused parameter warnings. | |
209 | ||
210 | -- | |
211 | ||
212 | Delete macro TARGET_BYTE_ORDER_SELECTABLE. | |
213 | ||
214 | Patches in the database. | |
215 | ||
216 | -- | |
217 | ||
218 | Updated readline | |
219 | ||
220 | Readline 4.? is out. A merge wouldn't hurt. | |
221 | ||
222 | -- | |
223 | ||
224 | Purge PARAMS | |
225 | ||
226 | Eliminate all uses of PARAMS in GDB's source code. | |
227 | ||
228 | -- | |
229 | ||
230 | Elimination of make_cleanup_func. (Andrew Cagney) | |
231 | ||
232 | make_cleanup_func elimination | |
233 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00791.html | |
234 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00814.html | |
235 | ||
236 | -- | |
237 | ||
bc9e5bbf AC |
238 | Re: Various C++ things |
239 | ||
240 | value_headof/value_from_vtable_info are worthless, and should be removed. | |
241 | The one place in printcmd.c that uses it should use the RTTI functions. | |
242 | ||
243 | RTTI for g++ should be using the typeinfo functions rather than the vtables. | |
244 | The typeinfo functions are always at offset 4 from the beginning of the vtable, | |
245 | and are always right. The vtables will have weird names like E::VB sometimes. | |
246 | The typeinfo function will always be "E type_info function", or somesuch. | |
247 | ||
248 | value_virtual_fn_field needs to be fixed so there are no failures for virtual | |
249 | functions for C++ using g++. | |
250 | ||
251 | Testsuite cases are the major priority right now for C++ support, since i have | |
252 | to make a lot of changes that could potentially break each other. | |
138f88c0 | 253 | |
67edb2c6 AC |
254 | -- |
255 | ||
256 | GDBARCH cleanup (Andrew Cagney) | |
257 | ||
258 | The non-generated parts of gdbarch.{sh,h,c} should be separated out | |
bc9e5bbf | 259 | into arch-utils.[hc]. |
67edb2c6 AC |
260 | |
261 | The ``info architecture'' command should be replaced with a fixed | |
262 | ``set architecture'' (implemented using the command.c enum code). | |
263 | ||
264 | Document that gdbarch_init_ftype could easily fail because it didn't | |
265 | identify an architecture. | |
266 | ||
267 | -- | |
268 | ||
bc9e5bbf AC |
269 | Migrate qfThreadInfo packet -> qThreadInfo. (Andrew Cagney) |
270 | ||
271 | Add support for packet enable/disable commands with these thread | |
272 | packets. General cleanup. | |
273 | ||
274 | [PATCH] Document the ThreadInfo remote protocol queries | |
275 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00832.html | |
276 | ||
277 | [PATCH] "info threads" queries for remote.c | |
278 | http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00831.html | |
279 | ||
280 | -- | |
281 | ||
747d1ccb AC |
282 | Eliminate gdb/tui/Makefile.in. |
283 | Cleanup configury support for optional sub-directories. | |
284 | ||
285 | Check how GCC handles multiple front ends for an example of how things | |
286 | could work. A tentative first step is to rationalize things so that | |
287 | all sub directories are handled in a fashion similar to gdb/mi. | |
288 | ||
289 | -- | |
290 | ||
9debab2f AC |
291 | [PATCH/5] src/intl/Makefile.in:distclean additions |
292 | http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00363.html | |
293 | ||
294 | Do not forget to merge the patch back into the trunk. | |
295 | ||
296 | -- | |
297 | ||
e55e8cee AC |
298 | Re: [RFC] Change configure.in so -W arnings match reality |
299 | http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00350.html | |
300 | ||
301 | Some GCC compilers do not like -Wreturn-type. (Going forward there | |
302 | may be more problems like that). Need to check which of the warning | |
303 | options are valid. Need to probably disable warnings by default. | |
304 | ||
305 | -- | |
306 | ||
2a00c9ce AC |
307 | General Wish List |
308 | ================= | |
309 | ||
310 | -- | |
311 | ||
312 | Register Cache Cleanup (below from Andrew Cagney) | |
313 | ||
314 | I would depict the current register architecture as something like: | |
315 | ||
316 | High GDB --> Low GDB | |
317 | | | | |
318 | \|/ \|/ | |
319 | --- REG NR ----- | |
320 | | | |
321 | register + REGISTER_BYTE(reg_nr) | |
322 | | | |
323 | \|/ | |
324 | ------------------------- | |
325 | | extern register[] | | |
326 | ------------------------- | |
327 | ||
328 | where neither the high (valops.c et.al.) or low gdb (*-tdep.c) are | |
329 | really clear on what mechanisms they should be using to manipulate that | |
330 | buffer. Further, much code assumes, dangerously, that registers are | |
331 | contigious. Having got mips-tdep.c to support multiple ABIs, believe | |
332 | me, that is a bad assumption. Finally, that register cache layout is | |
333 | determined by the current remote/local target and _not_ the less | |
334 | specific target ISA. In fact, in many cases it is determined by the | |
335 | somewhat arbitrary layout of the [gG] packets! | |
336 | ||
337 | ||
338 | How I would like the register file to work is more like: | |
339 | ||
340 | ||
341 | High GDB | |
342 | | | |
343 | \|/ | |
344 | pseudo reg-nr | |
345 | | | |
346 | map pseudo <-> | |
347 | random cache | |
348 | bytes | |
349 | | | |
350 | \|/ | |
351 | ------------ | |
352 | | register | | |
353 | | cache | | |
354 | ------------ | |
355 | /|\ | |
356 | | | |
357 | map random cache | |
358 | bytes to target | |
359 | dependant i-face | |
360 | /|\ | |
361 | | | |
362 | target dependant | |
363 | such as [gG] packet | |
364 | or ptrace buffer | |
365 | ||
366 | The main objectives being: | |
367 | ||
368 | o a clear separation between the low | |
369 | level target and the high level GDB | |
370 | ||
371 | o a mechanism that solves the general | |
372 | problem of register aliases, overlaps | |
373 | etc instead of treating them as optional | |
374 | extras that can be wedged in as an after | |
375 | thought (that is a reasonable description | |
376 | of the current code). | |
377 | ||
378 | Identify then solve the hard case and the | |
379 | rest just falls out. GDB solved the easy | |
380 | case and then tried to ignore the real | |
381 | world :-) | |
382 | ||
383 | o a removal of the assumption that the | |
384 | mapping between the register cache | |
385 | and virtual registers is largely static. | |
386 | If you flip the USR/SSR stack register | |
387 | select bit in the status-register then | |
388 | the corresponding stack registers should | |
389 | reflect the change. | |
390 | ||
391 | o a mechanism that clearly separates the | |
392 | gdb internal register cache from any | |
393 | target (not architecture) dependant | |
394 | specifics such as [gG] packets. | |
395 | ||
396 | Of course, like anything, it sounds good in theory. In reality, it | |
397 | would have to contend with many<->many relationships at both the | |
398 | virt<->cache and cache<->target level. For instance: | |
399 | ||
400 | virt<->cache | |
401 | Modifying an mmx register may involve | |
402 | scattering values across both FP and | |
403 | mmpx specific parts of a buffer | |
404 | ||
405 | cache<->target | |
406 | When writing back a SP it may need to | |
407 | both be written to both SP and USP. | |
408 | ||
409 | ||
410 | Hmm, | |
411 | ||
412 | Rather than let this like the last time it was discussed, just slip, I'm | |
413 | first going to add this e-mail (+ references) to TODO. I'd then like to | |
414 | sketch out a broad strategy I think could get us there. | |
415 | ||
416 | ||
417 | First thing I'd suggest is separating out the ``extern registers[]'' | |
418 | code so that we can at least identify what is using it. At present | |
419 | things are scattered across many files. That way we can at least | |
420 | pretend that there is a cache instead of a global array :-) | |
421 | ||
422 | I'd then suggest someone putting up a proposal for the pseudo-reg / | |
423 | high-level side interface so that code can be adopted to it. For old | |
424 | code, initially a blanket rename of write_register_bytes() to | |
425 | deprecated_write_register_bytes() would help. | |
426 | ||
427 | Following that would, finaly be the corresponding changes to the target. | |
bc9e5bbf AC |
428 | |
429 | -- | |
430 | ||
67edb2c6 AC |
431 | Check that GDB can handle all BFD architectures (Andrew Cagney) |
432 | ||
433 | There should be a test that checks that BFD/GDB are in sync with | |
434 | regard to architecture changes. Something like a test that first | |
435 | queries GDB for all supported architectures and then feeds each back | |
436 | to GDB.. Anyone interested in learning how to write tests? :-) | |
437 | ||
438 | -- | |
439 | ||
2a00c9ce AC |
440 | Add support for Modula3 |
441 | ||
442 | Get DEC/Compaq to contribute their Modula-3 support. | |
443 | ||
444 | -- | |
445 | ||
446 | Legacy Wish List | |
447 | ================ | |
448 | ||
449 | This list is not up to date, and opinions vary about the importance or | |
450 | even desirability of some of the items. If you do fix something, it | |
451 | always pays to check the below. | |
452 | ||
453 | -- | |
c906108c | 454 | |
b83266a0 SS |
455 | Document trace machinery. |
456 | ||
457 | Document overlay machinery. | |
458 | ||
459 | Extend .gdbinit mechanism to specify name on command line, allow for | |
460 | lists of files to load, include function of --tclcommand. | |
461 | ||
462 | @c This does not work (yet if ever). FIXME. | |
463 | @c @item --parse=@var{lang} @dots{} | |
464 | @c Configure the @value{GDBN} expression parser to parse the listed languages. | |
465 | @c @samp{all} configures @value{GDBN} for all supported languages. To get a | |
466 | @c list of all supported languages, omit the argument. Without this | |
467 | @c option, @value{GDBN} is configured to parse all supported languages. | |
468 | ||
c906108c SS |
469 | Add an "info bfd" command that displays supported object formats, |
470 | similarly to objdump -i. | |
471 | ||
472 | START_INFERIOR_TRAPS_EXPECTED need never be defined to 2, since that | |
473 | is its default value. Clean this up. | |
474 | ||
475 | It should be possible to use symbols from shared libraries before we know | |
476 | exactly where the libraries will be loaded. E.g. "b perror" before running | |
477 | the program. This could maybe be done as an extension of the "breakpoint | |
478 | re-evaluation" after new symbols are loaded. | |
479 | ||
480 | Make single_step() insert and remove breakpoints in one operation. | |
481 | ||
482 | Speed up single stepping by avoiding extraneous ptrace calls. | |
483 | ||
484 | Speed up single stepping by not inserting and removing breakpoints | |
485 | each time the inferior starts and stops. | |
486 | ||
487 | Breakpoints should not be inserted and deleted all the time. Only the | |
488 | one(s) there should be removed when we have to step over one. Support | |
489 | breakpoints that don't have to be removed to step over them. | |
490 | ||
491 | Update gdbint.texinfo to include doc on the directory structure and | |
492 | the various tricks of building gdb. | |
493 | ||
494 | Do a tutorial in gdb.texinfo on how to do simple things in gdb. | |
495 | E.g. how to set a breakpoint that just prints something and continues. | |
496 | How to break on aborts. Etc. | |
497 | ||
498 | Provide "voodoo" debugging of core files. This creates a zombie | |
499 | process as a child of the debugger, and loads it up with the data, | |
500 | stack, and regs of the core file. This allows you to call functions | |
501 | in the executable, to manipulate the data in the core file. | |
502 | ||
503 | GDB reopens the source file on every line, as you "next" through it. | |
504 | ||
505 | Referencing the vtbl member of a struct doesn't work. It prints OK | |
506 | if you print the struct, but it gets 0 if you try to deref it. | |
507 | ||
508 | Persistent command history: A feature where you could save off a list | |
509 | of the commands you did, so you can edit it into something that will bring | |
510 | the target to the same place every time you source it. | |
511 | This would also be useful for automated fast watchpointing; if you go | |
512 | past the place where it watchpoints, you just start it over again and | |
513 | do it more carefully. | |
514 | ||
515 | Deal with the SunOS 4.0 and 4.1.1 ptrace bug that loses the registers if | |
516 | the stack is paged out. | |
517 | ||
518 | Finish the C++ exception handling stub routines. Lint points them out | |
519 | as unused statics functions. | |
520 | ||
521 | Perhaps "i source" should take an argument like that of "list". | |
522 | ||
523 | See if core-aout.c's fetch_core_registers can be used on more machines. | |
524 | E.g. MIPS (mips-xdep.c). | |
525 | ||
526 | unpack_double() does not handle IEEE float on the target unless the host | |
527 | is also IEEE. Death on a vax. | |
528 | ||
529 | Set up interface between GDB and INFO so that you can hop into interactive | |
530 | INFO and back out again. When running under Emacs, should use Emacs | |
531 | info, else fork the info program. Installation of GDB should install | |
532 | its texinfo files into the info tree automagically, including the readline | |
533 | texinfo files. | |
534 | ||
535 | "help address" ought to find the "help set print address" entry. | |
536 | ||
537 | Remove the VTBL internal guts from printouts of C++ structs, unless | |
538 | vtblprint is set. | |
539 | ||
540 | Remove "at 0xnnnn" from the "b foo" response, if `print address off' and if | |
541 | it matches the source line indicated. | |
542 | ||
543 | The prompt at end of screen should accept space as well as CR. | |
544 | ||
545 | Check STORE_RETURN_VALUE on all architectures. Check near it in tm-sparc.h | |
546 | for other bogosities. | |
547 | ||
548 | Check for storage leaks in GDB, I'm sure there are a lot! | |
549 | ||
550 | vtblprint of a vtbl should demangle the names it's printing. | |
551 | ||
552 | Backtrace should point out what the currently selected frame is, in | |
553 | its display, perhaps showing "@3 foo (bar, ...)" or ">3 foo (bar, | |
554 | ...)" rather than "#3 foo (bar, ...)". | |
555 | ||
556 | "i program" should work for core files, and display more info, like what | |
557 | actually caused it to die. | |
558 | ||
559 | "x/10i" should shorten the long name, if any, on subsequent lines. | |
560 | ||
561 | Check through the code for FIXME comments and fix them. dbxread.c, | |
562 | blockframe.c, and plenty more. (I count 634 as of 940621 - sts) | |
563 | ||
564 | "next" over a function that longjumps, never stops until next time you happen | |
565 | to get to that spot by accident. E.g. "n" over execute_command which has | |
566 | an error. | |
567 | ||
568 | "set zeroprint off", don't bother printing members of structs which | |
569 | are entirely zero. Useful for those big structs with few useful | |
570 | members. | |
571 | ||
572 | GDB does four ioctl's for every command, probably switching terminal modes | |
573 | to/from inferior or for readline or something. | |
574 | ||
575 | terminal_ours versus terminal_inferior: cache state. Switch should be a noop | |
576 | if the state is the same, too. | |
577 | ||
578 | ptype $i6 = void??! | |
579 | ||
580 | Clean up invalid_float handling so gdb doesn't coredump when it tries to | |
581 | access a NaN. While this might work on SPARC, other machines are not | |
582 | configured right. | |
583 | ||
584 | "b value_at ; commands ; continue ; end" stops EVERY OTHER TIME! | |
585 | Then once you enter a command, it does the command, runs two more | |
586 | times, and then stops again! Bizarre... (This behaviour has been | |
587 | modified, but it is not yet 100% predictable when e.g. the commands | |
588 | call functions in the child, and while there, the child is interrupted | |
589 | with a signal, or hits a breakpoint.) | |
590 | ||
591 | help completion, help history should work. | |
592 | ||
593 | Check that we can handle stack trace through varargs AND alloca in same | |
594 | function, on 29K. | |
595 | ||
596 | wait_for_inferior loops forever if wait() gives it an error. | |
597 | ||
598 | "i frame" shows wrong "arglist at" location, doesn't show where the args | |
599 | should be found, only their actual values. | |
600 | ||
601 | There should be a way for "set" commands to validate the new setting | |
602 | before it takes effect. | |
603 | ||
604 | A mess of floating point opcodes are missing from sparc-opcode.h. | |
605 | Also, a little program should test the table for bits that are | |
606 | overspecified or underspecified. E.g. if the must-be-ones bits | |
607 | and the must-be-zeroes bits leave some fields unexamined, and the format | |
608 | string leaves them unprinted, then point this out. If multiple | |
609 | non-alias patterns match, point this out too. Finally, there should | |
610 | be a sparc-optest.s file that tries each pattern out. This file | |
611 | should end up coming back the same (modulo transformation comments) | |
612 | if fed to "gas" then the .o is fed to gdb for disassembly. | |
613 | ||
614 | Eliminate all the core_file_command's in all the xdep files. | |
615 | Eliminate separate declarations of registers[] everywhere. | |
616 | ||
617 | "ena d" is ambiguous, why? "ena delete" seems to think it is a command! | |
618 | ||
619 | Perhaps move the tdep, xdep, and nat files, into the config | |
620 | subdirectories. If not, at least straighten out their names so that | |
621 | they all start with the machine name. | |
622 | ||
623 | inferior_status should include stop_print_frame. It won't need to be | |
624 | reset in wait_for_inferior after bpstat_stop_status call, then. | |
625 | ||
626 | i line VAR produces "Line number not known for symbol ``var''.". I | |
627 | thought we were stashing that info now! | |
628 | ||
629 | We should be able to write to random files at hex offsets like adb. | |
630 | ||
631 | Make "target xxx" command interruptible. | |
632 | ||
633 | Handle add_file with separate text, data, and bss addresses. Maybe | |
634 | handle separate addresses for each segment in the object file? | |
635 | ||
636 | Handle free_named_symtab to cope with multiply-loaded object files | |
637 | in a dynamic linking environment. Should remember the last copy loaded, | |
638 | but not get too snowed if it finds references to the older copy. | |
639 | ||
640 | Generalize and Standardize the RPC interface to a target program, | |
641 | improve it beyond the "ptrace" interface, and see if it can become a | |
642 | standard for remote debugging. (This is talking about the vxworks | |
643 | interface. Seems unlikely to me that there will be "a standard" for | |
644 | remote debugging anytime soon --kingdon, 8 Nov 1994). | |
645 | ||
646 | Remove all references to: | |
647 | text_offset | |
648 | data_offset | |
649 | text_data_start | |
650 | text_end | |
651 | exec_data_offset | |
652 | ... | |
653 | now that we have BFD. All remaining are in machine dependent files. | |
654 | ||
655 | When quitting with a running program, if a core file was previously | |
656 | examined, you get "Couldn't read float regs from core file"...if | |
657 | indeed it can't. generic_mourn_inferior... | |
658 | ||
659 | Have remote targets give a warning on a signal argument to | |
660 | target_resume. Or better yet, extend the protocols so that it works | |
661 | like it does on the Unix-like systems. | |
662 | ||
663 | Sort help and info output. | |
664 | ||
665 | Re-organize help categories into things that tend to fit on a screen | |
666 | and hang together. | |
667 | ||
668 | renote-nindy.c handles interrupts poorly; it error()s out of badly | |
669 | chosen places, e.g. leaving current_frame zero, which causes core dumps | |
670 | on the next command. | |
671 | ||
672 | Add in commands like ADB's for searching for patterns, etc. We should | |
673 | be able to examine and patch raw unsymboled binaries as well in gdb as | |
674 | we can in adb. (E.g. increase the timeout in /bin/login without source). | |
675 | ||
676 | Those xdep files that call register_addr without defining it are | |
677 | probably simply broken. When reconfiguring this part of gdb, I could | |
678 | only make guesses about how to redo some of those files, and I | |
679 | probably guessed wrong, or left them "for later" when I have a | |
680 | machine that can attempt to build them. | |
681 | ||
682 | When doing "step" or "next", if a few lines of source are skipped between | |
683 | the previous line and the current one, print those lines, not just the | |
684 | last line of a multiline statement. | |
685 | ||
686 | When searching for C++ superclasses in value_cast in valops.c, we must | |
687 | not search the "fields", only the "superclasses". There might be a | |
688 | struct with a field name that matches the superclass name. This can | |
689 | happen when the struct was defined before the superclass (before the | |
690 | name became a typedef). | |
691 | ||
692 | Handling of "&" address-of operator needs some serious overhaul | |
693 | for ANSI C and consistency on arrays and functions. | |
694 | For "float point[15];": | |
695 | ptype &point[4] ==> Attempt to take address of non-lvalue. | |
696 | For "char *malloc();": | |
697 | ptype malloc ==> "char *()"; should be same as | |
698 | ptype &malloc ==> "char *(*)()" | |
699 | call printf ("%x\n", malloc) ==> weird value, should be same as | |
700 | call printf ("%x\n", &malloc) ==> correct value | |
701 | ||
702 | Fix dbxread.c symbol reading in the presence of interrupts. It | |
703 | currently leaves a cleanup to blow away the entire symbol table when a | |
704 | QUIT occurs. (What's wrong with that? -kingdon, 28 Oct 1993). | |
705 | ||
706 | Mipsread.c reads include files depth-first, because the dependencies | |
707 | in the psymtabs are way too inclusive (it seems to me). Figure out what | |
708 | really depends on what, to avoid recursing 20 or 30 times while reading | |
709 | real symtabs. | |
710 | ||
711 | value_add() should be subtracting the lower bound of arrays, if known, | |
712 | and possibly checking against the upper bound for error reporting. | |
713 | ||
714 | mipsread.c symbol table allocation and deallocation should be checked. | |
715 | My suspicion is that it's full of memory leaks. | |
716 | ||
717 | SunOS should have a target_lookup_symbol() for common'd things allocated | |
718 | by the shared library linker ld.so. | |
719 | ||
720 | When listing source lines, check for a preceding \n, to verify that | |
721 | the file hasn't changed out from under us. | |
722 | ||
723 | When listing source lines, eat leading whitespace corresponding to the | |
724 | line-number prefix we print. This avoids long lines wrapping. | |
725 | ||
726 | mipsread.c needs to check for old symtabs and psymtabs for the same | |
727 | files, the way it happens for dbxread.c and coffread.c, for VxWorks | |
728 | incremental symbol table reloading. | |
729 | ||
730 | Get all the remote systems (where the protocol allows it) to be able to | |
731 | stop the remote system when the GDB user types ^C (like remote.c | |
732 | does). For ebmon, use ^Ak. | |
733 | ||
734 | Possible feature: A version of the "disassemble" command which shows | |
735 | both source and assembly code ("set symbol-filename on" is a partial | |
736 | solution). | |
737 | ||
738 | investigate "x/s 0" (right now stops early) (I think maybe GDB is | |
739 | using a 0 address for bad purposes internally). | |
740 | ||
741 | Make "info path" and path_command work again (but independent of the | |
742 | environment either of gdb or that we'll pass to the inferior). | |
743 | ||
744 | Make GDB understand the GCC feature for putting octal constants in | |
745 | enums. Make it so overflow on an enum constant does not error_type | |
746 | the whole type. Allow arbitrarily large enums with type attributes. | |
747 | Put all this stuff in the testsuite. | |
748 | ||
749 | Make TYPE_CODE_ERROR with a non-zero TYPE_LENGTH more useful (print | |
750 | the value in hex; process type attributes). Add this to the | |
751 | testsuite. This way future compilers can add new types and old | |
752 | versions of GDB can do something halfway reasonable. | |
753 | ||
754 | Clean up formatting of "info registers" on MIPS and 88k. See if it | |
755 | is possible to do this generically across all target architectures. | |
756 | ||
757 | GDB gets bfd/corefile.c and gdb/corefile.c confused (this should be easy to | |
758 | repeat even with something more recent than GDB 4.9). | |
759 | ||
760 | Check that unmatched RBRAC doesn't abort(). | |
761 | ||
762 | Fix mdebugread.c:parse_type to do fundamental types right (see | |
763 | rs6000_builtin_type in stabsread.c for what "right" is--the point is | |
764 | that the debug format fixes the sizes of these things and it shouldn't | |
765 | depend on stuff like TARGET_PTR_BIT and so on. For mdebug, there seem | |
766 | to be separate bt* codes for 64 bit and 32 bit things, and GDB should | |
767 | be aware of that). Also use a switch statement for clarity and speed. | |
768 | ||
769 | Investigate adding symbols in target_load--some targets do, some | |
770 | don't. | |
771 | ||
772 | Put dirname in psymtabs and change lookup*symtab to use dirname (so | |
773 | /foo/bar.c works whether compiled by cc /foo/bar.c, or cd /foo; cc | |
774 | bar.c). | |
775 | ||
776 | Merge xcoffread.c and coffread.c. Use breakpoint_re_set instead of | |
777 | fixup_breakpoints. | |
778 | ||
779 | Fix byte order and int size sins in tm-a29k.h | |
780 | (EXTRACT_RETURN_VALUE). Perhaps should reproduce bug and verify fix | |
781 | (or perhaps should just fix it...). | |
782 | ||
783 | Make a watchpoint on a constant expression an error (or warning | |
784 | perhaps) | |
785 | ||
786 | Make a watchpoint which contains a function call an error (it is | |
787 | broken now, making it work is probably not worth the effort). | |
788 | ||
789 | Re-do calls to signal() in remote.c, and inflow.c (set_sigint_trap and | |
790 | so on) to be independent of the debugging target, using target_stop to | |
791 | stop the inferior. Probably the part which is now handled by | |
792 | interrupt_query in remote.c can be done without any new features in | |
793 | the debugging target. | |
794 | ||
795 | New test case based on weird.exp but in which type numbers are not | |
796 | renumbered (thus multiply defining a type). This currently causes an | |
797 | infinite loop on "p v_comb". | |
798 | ||
799 | Nuke baseclass_addr. | |
800 | ||
801 | Nuke USG define. | |
802 | ||
803 | "source file more recent" loses on re-read | |
804 | ||
805 | Fix 386 floating point so that floating point registers are real | |
806 | registers (but code can deal at run-time if they are missing, like | |
807 | mips and 68k). This would clean up "info float" and related stuff. | |
808 | ||
809 | Look at Solaris bug in interrupt.exp. Can get out of syscall with | |
810 | PRSABORT (syscall will return EINTR) but merely doing that leads to a | |
811 | "can't read memory" error. | |
812 | ||
813 | gcc -g -c enummask.c then gdb enummask.o, then "p v". GDB complains | |
814 | about not being able to access memory location 0. | |
815 | ||
816 | -------------------- enummask.c | |
817 | enum mask | |
818 | { | |
819 | ANIMAL = 0, | |
820 | VEGETABLE = 1, | |
821 | MINERAL = 2, | |
822 | BASIC_CATEGORY = 3, | |
823 | ||
824 | WHITE = 0, | |
825 | BLUE = 4, | |
826 | GREEN = 8, | |
827 | BLACK = 0xc, | |
828 | COLOR = 0xc, | |
829 | ||
830 | ALIVE = 0x10, | |
831 | ||
832 | LARGE = 0x20 | |
833 | } v; | |
834 | ||
835 | If try to modify value in file with "set write off" should give | |
836 | appropriate error not "cannot access memory at address 0x65e0". | |
837 | ||
838 | Why do we allow a target to omit standard register names (NO_STD_REGS | |
839 | in tm-z8k.h)? I thought the standard register names were supposed to | |
840 | be just that, standard. | |
841 | ||
c906108c SS |
842 | Allow core file without exec file on RS/6000. |
843 | ||
844 | Make sure "shell" with no arguments works right on DOS. | |
845 | ||
846 | Make gdb.ini (as well as .gdbinit) be checked on all platforms, so | |
847 | the same directory can be NFS-mounted on unix or DOS, and work the | |
848 | same way. | |
849 | ||
850 | cd ~/tmp/<M-?> causes infinite loop (where ~/tmp is a directory). | |
851 | ||
852 | Get SECT_OFF_TEXT stuff out of objfile_relocate (might be needed to | |
853 | get RS/6000 to work right, might not be immediately relevant). | |
854 | ||
855 | Clean up add_toc_to_loadinfo | |
856 | ||
857 | Think about attached processes and sharing terminal. | |
858 | ||
859 | John sez in reference to ignoring errors from tcsegpgrp if attach_flag: | |
860 | set_tty_state should not have any trouble with attached processes. | |
861 | Instead, the tty handling should leave the pgrp of the tty alone when | |
862 | attaching to processes (perhaps pass terminal_init_inferior a flag | |
863 | saying whether we're attaching). | |
864 | ||
865 | PAGE_SIZE redefined warnings on AIX. Probably should be using | |
866 | BFD_PAGE_SIZE throughout BFD. | |
867 | ||
868 | Rewrite proceed, wait_for_inferior, and normal_stop to clean them up. | |
869 | Suggestions: | |
870 | ||
871 | 1) Make each test in wait_for_inferior a seperate subroutine | |
872 | call. | |
873 | 2) Combine wait_for_inferior and normal_stop to clean up | |
874 | communication via global variables. | |
875 | 3) See if you can find some way to clean up the global | |
876 | variables that are used; possibly group them by data flow | |
877 | and information content? | |
878 | ||
879 | Work out some kind of way to allow running the inferior to be done as | |
880 | a sub-execution of, eg. breakpoint command lists. Currently running | |
881 | the inferior interupts any command list execution. This would require | |
882 | some rewriting of wait_for_inferior & friends, and hence should | |
883 | probably be done in concert with the above. | |
884 | ||
885 | Add function arguments to gdb user defined functions. | |
886 | ||
887 | Add convenience variables that refer to exec file, symbol file, | |
888 | selected frame source file, selected frame function, selected frame | |
889 | line number, etc. | |
890 | ||
891 | Add a "suspend" subcommand of the "continue" command to suspend gdb | |
892 | while continuing execution of the subprocess. Useful when you are | |
893 | debugging servers and you want to dodge out and initiate a connection | |
894 | to a server running under gdb. | |
895 | ||
896 | Add stab information to allow reasonable debugging of inline functions | |
897 | (possibly they should show up on a stack backtrace? With a note | |
898 | indicating that they weren't "real"?). | |
899 | ||
900 | Modify the naked "until" command to step until past the current source | |
901 | line, rather than past the current pc value. This is tricky simply | |
902 | because the low level routines have no way of specifying a multi-line | |
903 | step range, and there is no way of saying "don't print stuff when we | |
904 | stop" from above (otherwise could just call step many times). | |
905 | ||
906 | Modify the handling of symbols grouped through BINCL/EINCL stabs to | |
907 | allocate a partial symtab for each BINCL/EINCL grouping. This will | |
908 | seriously decrease the size of inter-psymtab dependencies and hence | |
909 | lessen the amount that needs to be read in when a new source file is | |
910 | accessed. | |
911 | ||
912 | Do an "x/i $pc" after each stepi or nexti. | |
913 | ||
914 | Modify all of the disassemblers to use printf_filtered to get correct | |
915 | more filtering. | |
916 | ||
917 | Modify gdb to work correctly with Pascal. | |
918 | ||
919 | Add a command for searching memory, a la adb. It specifies size, | |
920 | mask, value, start address. ADB searches until it finds it or hits | |
921 | an error (or is interrupted). | |
922 | ||
b83266a0 SS |
923 | Remove the range and type checking code and documentation, if not |
924 | going to implement. | |
925 | ||
c906108c SS |
926 | # Local Variables: |
927 | # mode: text | |
928 | # End: |