OpenCores
URL https://opencores.org/ocsvn/or1k/or1k/trunk

Subversion Repositories or1k

[/] [or1k/] [trunk/] [gdb-5.0/] [gdb/] [TODO] - Blame information for rev 1765

Details | Compare with Previous | View Log

Line No. Rev Author Line
1 104 markom
If you find inaccuracies in this list, please send mail to
2
bug-gdb@prep.ai.mit.edu.  If you would like to work on any of these,
3
you should consider sending mail to the same address, to find out
4
whether anyone else is working on it.
5
 
6
 
7
Known problems in GDB 5.0
8
=========================
9
 
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.
13
 
14
(The names in paren indicate people that posted the original problem.)
15
 
16
--
17
 
18
GDB requires GCC to build under IRIX
19
 
20
IRIX, being more pedantic than GCC reports as errors certain
21
assignments that GCC treats as warnings.
22
 
23
This can be worked around by building GDB with the GCC compiler.
24
 
25
--
26
 
27
The BFD directory requires bug-fixed AUTOMAKE et.al.
28
 
29
AUTOMAKE 1.4 incorrectly set the TEXINPUTS environment variable.  It
30
contained the full path to texinfo.tex when it should have only
31
contained the directory.  The bug has been fixed in the current
32
AUTOMAKE sources.  Automake snapshots can be found in:
33
        ftp://sourceware.cygnus.com/pub/gdb/snapshots
34
and     ftp://sourceware.cygnus.com/pub/binutils
35
 
36
--
37
 
38
gdb-cvs fails to build on freebsd-elf
39
http://sourceware.cygnus.com/ml/gdb/2000-04/msg00004.html
40
 
41
Either the FreeBSD group need to contribute their local GDB changes
42
back to the master sources or someone needs to provides a new
43
(clean-room) implementation.  Since the former involves a fairly
44
complicated assignment the latter may be easier. [cagney]
45
 
46
--
47
 
48
Generic: lin-thread cannot handle thread exit (Mark Kettenis, Michael
49
Snyder) http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00525.html
50
 
51
The thread_db assisted debugging code doesn't handle exiting threads
52
properly, at least in combination with glibc 2.1.3 (the framework is
53
there, just not the actual code).  There are at least two problems
54
that prevent this from working.
55
 
56
As an additional reference point, the pre thread_db code did not work
57
either.
58
 
59
--
60
 
61
Java (Anthony Green, David Taylor)
62
 
63
Anthony Green has a number of Java patches that did not make it into
64
the 5.0 release.
65
 
66
Patch: java tests
67
http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00512.html
68
 
69
Patch: java booleans
70
http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00515.html
71
 
72
Patch: handle N_MAIN stab
73
http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00527.html
74
 
75
--
76
 
77
Pascal (Pierre Muller, David Taylor)
78
 
79
Pierre Muller has contributed patches for adding Pascal Language
80
support to GDB.
81
 
82
2 pascal language patches inserted in database
83
http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00521.html
84
 
85
Indent -gnu ?
86
http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00496.html
87
 
88
--
89
 
90
GNU/Linux/x86 and random thread signals (and Solaris/SPARC but not
91
Solaris/x86).
92
http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00336.html
93
 
94
Christopher Blizzard writes:
95
 
96
So, I've done some more digging into this and it looks like Jim
97
Kingdon has reported this problem in the past:
98
 
99
http://sourceware.cygnus.com/ml/bug-gdb/1999-10/msg00058.html
100
 
101
I can reproduce this problem both with and without Tom's patch.  Has
102
anyone seen this before?  Maybe have a solution for it hanging around?
103
:)
104
 
105
There's a test case for this documented at:
106
 
107
when debugging threaded applications you get extra SIGTRAPs
108
http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=9565
109
 
110
[There should be a GDB testcase - cagney]
111
 
112
--
113
 
114
Possible regressions with some devel GCCs.
115
http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00475.html
116
 
117
gcc-2.95.2 outputs a line note *before* the prologue (and one for the
118
closing brace after the epilogue, instead of before it, as it used to
119
be). By disabling the RTL-style prologue generating mechanism
120
(undocumented GCC option -mno-schedule-prologue), you get back the
121
traditional behaviour.
122
http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00510.html
123
 
124
This should now be fixed.
125
 
126
--
127
 
128
RFD: infrun.c: No bpstat_stop_status call after proceed over break?
129
(Peter Schauer)
130
http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00665.html
131
 
132
GDB misses watchpoint triggers after proceeding over a breakpoint on
133
x86 targets.
134
 
135
--
136
 
137
x86 linux GDB and SIGALRM (???)
138
http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00803.html
139
 
140
I know there are problems with single stepping through signal
141
handlers.  These problems were present in 4.18.  They were just masked
142
because 4.18 failed to recognize signal handlers.  Fixing it is not
143
easy, and will require changes to handle_inferior_event(), that I
144
prefer not to make before the 5.0 release.
145
 
146
Mark
147
 
148
--
149
 
150
Revised UDP support (was: Re: [Fwd: [patch] UDP transport support])
151
http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00000.html
152
 
153
(Broken) support for GDB's remote protocol across UDP is to be
154
included in the follow-on release.
155
 
156
--
157
 
158
Can't build IRIX -> arm GDB.
159
http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00356.html
160
 
161
David Whedon writes:
162
> Now I'm building for an embedded arm target.  If there is a way of turning
163
> remote-rdi off, I couldn't find it. It looks like it gets built by default
164
> in gdb/configure.tgt(line 58) Anyway, the build dies in
165
> gdb/rdi-share/unixcomm.c.  SERPORT1 et. al. never get defined because we
166
> aren't one of the architectures supported.
167
 
168
--
169
 
170
Problem with weak functions
171
http://sourceware.cygnus.com/ml/gdb/2000-05/msg00060.html
172
 
173
Dan Nicolaescu writes:
174
> It seems that gdb-4.95.1  does not display correctly the function when
175
> stoping in weak functions.
176
>
177
> It stops in a function that is defined as weak, not in the function
178
> that is actualy run...
179
 
180
--
181
 
182
GDB5 TOT on unixware 7
183
http://sourceware.cygnus.com/ml/gdb/2000-04/msg00119.html
184
 
185
Robert Lipe writes:
186
> I just spun the top of tree of the GDB5 branch on UnixWare 7.  As a
187
> practical matter, the current thread support is somewhat more annoying
188
> than when GDB was thread-unaware.
189
 
190
--
191
 
192
                ------------------------------------------------
193
 
194
Code cleanups
195
=============
196
 
197
The following code cleanups are planned for the follow-on release to
198
GDB 5.0.
199
 
200
--
201
 
202
ATTRIBUTE_UNUSED
203
 
204
The need for this as almost been eliminated.  The next version of GCC
205
(assuming cagney gets the relevant patch committed) will be able to
206
supress unused parameter warnings.
207
 
208
--
209
 
210
Delete macro TARGET_BYTE_ORDER_SELECTABLE.
211
 
212
Patches in the database.
213
 
214
--
215
 
216
Updated readline
217
 
218
Readline 4.? is out.  A merge wouldn't hurt.
219
 
220
--
221
 
222
Purge PARAMS
223
 
224
Eliminate all uses of PARAMS in GDB's source code.
225
 
226
--
227
 
228
Elimination of make_cleanup_func. (Andrew Cagney)
229
 
230
make_cleanup_func elimination
231
http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00791.html
232
http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00814.html
233
 
234
--
235
 
236
ChangeLog.mi vs ChangeLog-mi (Andrew Cagney)
237
Needs further debate.
238
 
239
Re: [PATCH] Add change-log variables to more MI files
240
http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00811.html
241
 
242
--
243
 
244
Re: Various C++ things
245
 
246
value_headof/value_from_vtable_info are worthless, and should be removed.
247
The one place in printcmd.c that uses it should use the RTTI functions.
248
 
249
RTTI for g++ should be using the typeinfo functions rather than the vtables.
250
The typeinfo functions are always at offset 4 from the beginning of the vtable,
251
and are always right. The vtables will have weird names like E::VB sometimes.
252
The typeinfo function will always be "E type_info function", or somesuch.
253
 
254
value_virtual_fn_field needs to be fixed so there are no failures for virtual
255
functions for C++ using g++.
256
 
257
Testsuite cases are the major priority right now for C++ support, since i have
258
to make a lot of changes that could potentially break each other.
259
 
260
--
261
 
262
GDBARCH cleanup (Andrew Cagney)
263
 
264
The non-generated parts of gdbarch.{sh,h,c} should be separated out
265
into arch-utils.[hc].
266
 
267
The ``info architecture'' command should be replaced with a fixed
268
``set architecture'' (implemented using the command.c enum code).
269
 
270
Document that gdbarch_init_ftype could easily fail because it didn't
271
identify an architecture.
272
 
273
--
274
 
275
Migrate qfThreadInfo packet -> qThreadInfo. (Andrew Cagney)
276
 
277
Add support for packet enable/disable commands with these thread
278
packets.  General cleanup.
279
 
280
[PATCH] Document the ThreadInfo remote protocol queries
281
http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00832.html
282
 
283
[PATCH] "info threads" queries for remote.c
284
http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00831.html
285
 
286
--
287
 
288
Eliminate gdb/tui/Makefile.in.
289
Cleanup configury support for optional sub-directories.
290
 
291
Check how GCC handles multiple front ends for an example of how things
292
could work.  A tentative first step is to rationalize things so that
293
all sub directories are handled in a fashion similar to gdb/mi.
294
 
295
--
296
 
297
General Wish List
298
=================
299
 
300
--
301
 
302
Check that GDB can handle all BFD architectures (Andrew Cagney)
303
 
304
There should be a test that checks that BFD/GDB are in sync with
305
regard to architecture changes.  Something like a test that first
306
queries GDB for all supported architectures and then feeds each back
307
to GDB..  Anyone interested in learning how to write tests?  :-)
308
 
309
--
310
 
311
This list is probably not up to date, and opinions vary about the
312
importance or even desirability of some of the items.
313
 
314
Document trace machinery.
315
 
316
Document overlay machinery.
317
 
318
Extend .gdbinit mechanism to specify name on command line, allow for
319
lists of files to load, include function of --tclcommand.
320
 
321
@c This does not work (yet if ever).  FIXME.
322
@c @item --parse=@var{lang} @dots{}
323
@c Configure the @value{GDBN} expression parser to parse the listed languages.
324
@c @samp{all} configures @value{GDBN} for all supported languages.  To get a
325
@c list of all supported languages, omit the argument.  Without this
326
@c option, @value{GDBN} is configured to parse all supported languages.
327
 
328
Add an "info bfd" command that displays supported object formats,
329
similarly to objdump -i.
330
 
331
START_INFERIOR_TRAPS_EXPECTED need never be defined to 2, since that
332
is its default value.  Clean this up.
333
 
334
It should be possible to use symbols from shared libraries before we know
335
exactly where the libraries will be loaded.  E.g. "b perror" before running
336
the program.  This could maybe be done as an extension of the "breakpoint
337
re-evaluation" after new symbols are loaded.
338
 
339
Make single_step() insert and remove breakpoints in one operation.
340
 
341
Speed up single stepping by avoiding extraneous ptrace calls.
342
 
343
Speed up single stepping by not inserting and removing breakpoints
344
each time the inferior starts and stops.
345
 
346
Breakpoints should not be inserted and deleted all the time.  Only the
347
one(s) there should be removed when we have to step over one.  Support
348
breakpoints that don't have to be removed to step over them.
349
 
350
Update gdbint.texinfo to include doc on the directory structure and
351
the various tricks of building gdb.
352
 
353
Do a tutorial in gdb.texinfo on how to do simple things in gdb.
354
E.g. how to set a breakpoint that just prints something and continues.
355
How to break on aborts.  Etc.
356
 
357
Provide "voodoo" debugging of core files.  This creates a zombie
358
process as a child of the debugger, and loads it up with the data,
359
stack, and regs of the core file.  This allows you to call functions
360
in the executable, to manipulate the data in the core file.
361
 
362
GDB reopens the source file on every line, as you "next" through it.
363
 
364
Referencing the vtbl member of a struct doesn't work.  It prints OK
365
if you print the struct, but it gets 0 if you try to deref it.
366
 
367
Persistent command history: A feature where you could save off a list
368
of the commands you did, so you can edit it into something that will bring
369
the target to the same place every time you source it.
370
This would also be useful for automated fast watchpointing; if you go
371
past the place where it watchpoints, you just start it over again and
372
do it more carefully.
373
 
374
Deal with the SunOS 4.0 and 4.1.1 ptrace bug that loses the registers if
375
the stack is paged out.
376
 
377
Finish the C++ exception handling stub routines.  Lint points them out
378
as unused statics functions.
379
 
380
Perhaps "i source" should take an argument like that of "list".
381
 
382
See if core-aout.c's fetch_core_registers can be used on more machines.
383
E.g. MIPS (mips-xdep.c).
384
 
385
unpack_double() does not handle IEEE float on the target unless the host
386
is also IEEE.  Death on a vax.
387
 
388
Set up interface between GDB and INFO so that you can hop into interactive
389
INFO and back out again.  When running under Emacs, should use Emacs
390
info, else fork the info program.  Installation of GDB should install
391
its texinfo files into the info tree automagically, including the readline
392
texinfo files.
393
 
394
"help address" ought to find the "help set print address" entry.
395
 
396
Remove the VTBL internal guts from printouts of C++ structs, unless
397
vtblprint is set.
398
 
399
Remove "at 0xnnnn" from the "b foo" response, if `print address off' and if
400
it matches the source line indicated.
401
 
402
The prompt at end of screen should accept space as well as CR.
403
 
404
Check STORE_RETURN_VALUE on all architectures.  Check near it in tm-sparc.h
405
for other bogosities.
406
 
407
Check for storage leaks in GDB, I'm sure there are a lot!
408
 
409
vtblprint of a vtbl should demangle the names it's printing.
410
 
411
Backtrace should point out what the currently selected frame is, in
412
its display, perhaps showing "@3 foo (bar, ...)" or ">3 foo (bar,
413
...)" rather than "#3 foo (bar, ...)".
414
 
415
"i program" should work for core files, and display more info, like what
416
actually caused it to die.
417
 
418
"x/10i" should shorten the long name, if any, on subsequent lines.
419
 
420
Check through the code for FIXME comments and fix them.  dbxread.c,
421
blockframe.c, and plenty more. (I count 634 as of 940621 - sts)
422
 
423
"next" over a function that longjumps, never stops until next time you happen
424
to get to that spot by accident.  E.g. "n" over execute_command which has
425
an error.
426
 
427
"set zeroprint off", don't bother printing members of structs which
428
are entirely zero.  Useful for those big structs with few useful
429
members.
430
 
431
GDB does four ioctl's for every command, probably switching terminal modes
432
to/from inferior or for readline or something.
433
 
434
terminal_ours versus terminal_inferior: cache state.  Switch should be a noop
435
if the state is the same, too.
436
 
437
ptype $i6 = void??!
438
 
439
Clean up invalid_float handling so gdb doesn't coredump when it tries to
440
access a NaN.  While this might work on SPARC, other machines are not
441
configured right.
442
 
443
"b value_at ; commands ; continue ; end" stops EVERY OTHER TIME!
444
Then once you enter a command, it does the command, runs two more
445
times, and then stops again!  Bizarre...  (This behaviour has been
446
modified, but it is not yet 100% predictable when e.g. the commands
447
call functions in the child, and while there, the child is interrupted
448
with a signal, or hits a breakpoint.)
449
 
450
help completion, help history should work.
451
 
452
Check that we can handle stack trace through varargs AND alloca in same
453
function, on 29K.
454
 
455
wait_for_inferior loops forever if wait() gives it an error.
456
 
457
"i frame" shows wrong "arglist at" location, doesn't show where the args
458
should be found, only their actual values.
459
 
460
There should be a way for "set" commands to validate the new setting
461
before it takes effect.
462
 
463
A mess of floating point opcodes are missing from sparc-opcode.h.
464
Also, a little program should test the table for bits that are
465
overspecified or underspecified.  E.g. if the must-be-ones bits
466
and the must-be-zeroes bits leave some fields unexamined, and the format
467
string leaves them unprinted, then point this out.  If multiple
468
non-alias patterns match, point this out too.  Finally, there should
469
be a sparc-optest.s file that tries each pattern out.  This file
470
should end up coming back the same (modulo transformation comments)
471
if fed to "gas" then the .o is fed to gdb for disassembly.
472
 
473
Eliminate all the core_file_command's in all the xdep files.
474
Eliminate separate declarations of registers[] everywhere.
475
 
476
"ena d" is ambiguous, why?  "ena delete" seems to think it is a command!
477
 
478
Perhaps move the tdep, xdep, and nat files, into the config
479
subdirectories.  If not, at least straighten out their names so that
480
they all start with the machine name.
481
 
482
inferior_status should include stop_print_frame.  It won't need to be
483
reset in wait_for_inferior after bpstat_stop_status call, then.
484
 
485
i line VAR produces "Line number not known for symbol ``var''.".  I
486
thought we were stashing that info now!
487
 
488
We should be able to write to random files at hex offsets like adb.
489
 
490
Make "target xxx" command interruptible.
491
 
492
Handle add_file with separate text, data, and bss addresses.  Maybe
493
handle separate addresses for each segment in the object file?
494
 
495
Handle free_named_symtab to cope with multiply-loaded object files
496
in a dynamic linking environment.  Should remember the last copy loaded,
497
but not get too snowed if it finds references to the older copy.
498
 
499
Generalize and Standardize the RPC interface to a target program,
500
improve it beyond the "ptrace" interface, and see if it can become a
501
standard for remote debugging.  (This is talking about the vxworks
502
interface.  Seems unlikely to me that there will be "a standard" for
503
remote debugging anytime soon --kingdon, 8 Nov 1994).
504
 
505
Remove all references to:
506
        text_offset
507
        data_offset
508
        text_data_start
509
        text_end
510
        exec_data_offset
511
        ...
512
now that we have BFD.  All remaining are in machine dependent files.
513
 
514
When quitting with a running program, if a core file was previously
515
examined, you get "Couldn't read float regs from core file"...if
516
indeed it can't.  generic_mourn_inferior...
517
 
518
Have remote targets give a warning on a signal argument to
519
target_resume.  Or better yet, extend the protocols so that it works
520
like it does on the Unix-like systems.
521
 
522
Sort help and info output.
523
 
524
Re-organize help categories into things that tend to fit on a screen
525
and hang together.
526
 
527
renote-nindy.c handles interrupts poorly; it error()s out of badly
528
chosen places, e.g. leaving current_frame zero, which causes core dumps
529
on the next command.
530
 
531
Add in commands like ADB's for searching for patterns, etc.  We should
532
be able to examine and patch raw unsymboled binaries as well in gdb as
533
we can in adb.  (E.g. increase the timeout in /bin/login without source).
534
 
535
Those xdep files that call register_addr without defining it are
536
probably simply broken.  When reconfiguring this part of gdb, I could
537
only make guesses about how to redo some of those files, and I
538
probably guessed wrong, or left them "for later" when I have a
539
machine that can attempt to build them.
540
 
541
When doing "step" or "next", if a few lines of source are skipped between
542
the previous line and the current one, print those lines, not just the
543
last line of a multiline statement.
544
 
545
When searching for C++ superclasses in value_cast in valops.c, we must
546
not search the "fields", only the "superclasses".  There might be a
547
struct with a field name that matches the superclass name.  This can
548
happen when the struct was defined before the superclass (before the
549
name became a typedef).
550
 
551
Handling of "&" address-of operator needs some serious overhaul
552
for ANSI C and consistency on arrays and functions.
553
  For "float point[15];":
554
ptype &point[4]   ==> Attempt to take address of non-lvalue.
555
  For "char *malloc();":
556
ptype malloc      ==> "char *()";  should be same as
557
ptype &malloc     ==> "char *(*)()"
558
call printf ("%x\n", malloc)   ==> weird value, should be same as
559
call printf ("%x\n", &malloc)  ==> correct value
560
 
561
Fix dbxread.c symbol reading in the presence of interrupts.  It
562
currently leaves a cleanup to blow away the entire symbol table when a
563
QUIT occurs.  (What's wrong with that? -kingdon, 28 Oct 1993).
564
 
565
Mipsread.c reads include files depth-first, because the dependencies
566
in the psymtabs are way too inclusive (it seems to me).  Figure out what
567
really depends on what, to avoid recursing 20 or 30 times while reading
568
real symtabs.
569
 
570
value_add() should be subtracting the lower bound of arrays, if known,
571
and possibly checking against the upper bound for error reporting.
572
 
573
mipsread.c symbol table allocation and deallocation should be checked.
574
My suspicion is that it's full of memory leaks.
575
 
576
SunOS should have a target_lookup_symbol() for common'd things allocated
577
by the shared library linker ld.so.
578
 
579
When listing source lines, check for a preceding \n, to verify that
580
the file hasn't changed out from under us.
581
 
582
When listing source lines, eat leading whitespace corresponding to the
583
line-number prefix we print.  This avoids long lines wrapping.
584
 
585
mipsread.c needs to check for old symtabs and psymtabs for the same
586
files, the way it happens for dbxread.c and coffread.c, for VxWorks
587
incremental symbol table reloading.
588
 
589
Get all the remote systems (where the protocol allows it) to be able to
590
stop the remote system when the GDB user types ^C (like remote.c
591
does).  For ebmon, use ^Ak.
592
 
593
Possible feature: A version of the "disassemble" command which shows
594
both source and assembly code ("set symbol-filename on" is a partial
595
solution).
596
 
597
investigate "x/s 0" (right now stops early) (I think maybe GDB is
598
using a 0 address for bad purposes internally).
599
 
600
Make "info path" and path_command work again (but independent of the
601
environment either of gdb or that we'll pass to the inferior).
602
 
603
Make GDB understand the GCC feature for putting octal constants in
604
enums.  Make it so overflow on an enum constant does not error_type
605
the whole type.  Allow arbitrarily large enums with type attributes.
606
Put all this stuff in the testsuite.
607
 
608
Make TYPE_CODE_ERROR with a non-zero TYPE_LENGTH more useful (print
609
the value in hex; process type attributes).  Add this to the
610
testsuite.  This way future compilers can add new types and old
611
versions of GDB can do something halfway reasonable.
612
 
613
Clean up formatting of "info registers" on MIPS and 88k.  See if it
614
is possible to do this generically across all target architectures.
615
 
616
GDB gets bfd/corefile.c and gdb/corefile.c confused (this should be easy to
617
repeat even with something more recent than GDB 4.9).
618
 
619
Check that unmatched RBRAC doesn't abort().
620
 
621
Fix mdebugread.c:parse_type to do fundamental types right (see
622
rs6000_builtin_type in stabsread.c for what "right" is--the point is
623
that the debug format fixes the sizes of these things and it shouldn't
624
depend on stuff like TARGET_PTR_BIT and so on.  For mdebug, there seem
625
to be separate bt* codes for 64 bit and 32 bit things, and GDB should
626
be aware of that).  Also use a switch statement for clarity and speed.
627
 
628
Investigate adding symbols in target_load--some targets do, some
629
don't.
630
 
631
Put dirname in psymtabs and change lookup*symtab to use dirname (so
632
/foo/bar.c works whether compiled by cc /foo/bar.c, or cd /foo; cc
633
bar.c).
634
 
635
Merge xcoffread.c and coffread.c.  Use breakpoint_re_set instead of
636
fixup_breakpoints.
637
 
638
Fix byte order and int size sins in tm-a29k.h
639
(EXTRACT_RETURN_VALUE).  Perhaps should reproduce bug and verify fix
640
(or perhaps should just fix it...).
641
 
642
Make a watchpoint on a constant expression an error (or warning
643
perhaps)
644
 
645
Make a watchpoint which contains a function call an error (it is
646
broken now, making it work is probably not worth the effort).
647
 
648
Re-do calls to signal() in remote.c, and inflow.c (set_sigint_trap and
649
so on) to be independent of the debugging target, using target_stop to
650
stop the inferior.  Probably the part which is now handled by
651
interrupt_query in remote.c can be done without any new features in
652
the debugging target.
653
 
654
New test case based on weird.exp but in which type numbers are not
655
renumbered (thus multiply defining a type).  This currently causes an
656
infinite loop on "p v_comb".
657
 
658
Nuke baseclass_addr.
659
 
660
Nuke USG define.
661
 
662
"source file more recent" loses on re-read
663
 
664
Fix 386 floating point so that floating point registers are real
665
registers (but code can deal at run-time if they are missing, like
666
mips and 68k).  This would clean up "info float" and related stuff.
667
 
668
Look at Solaris bug in interrupt.exp.  Can get out of syscall with
669
PRSABORT (syscall will return EINTR) but merely doing that leads to a
670
"can't read memory" error.
671
 
672
gcc -g -c enummask.c then gdb enummask.o, then "p v".  GDB complains
673
about not being able to access memory location 0.
674
 
675
-------------------- enummask.c
676
enum mask
677
{
678
  ANIMAL = 0,
679
  VEGETABLE = 1,
680
  MINERAL = 2,
681
  BASIC_CATEGORY = 3,
682
 
683
  WHITE = 0,
684
  BLUE = 4,
685
  GREEN = 8,
686
  BLACK = 0xc,
687
  COLOR = 0xc,
688
 
689
  ALIVE = 0x10,
690
 
691
  LARGE = 0x20
692
} v;
693
 
694
If try to modify value in file with "set write off" should give
695
appropriate error not "cannot access memory at address 0x65e0".
696
 
697
Why do we allow a target to omit standard register names (NO_STD_REGS
698
in tm-z8k.h)?  I thought the standard register names were supposed to
699
be just that, standard.
700
 
701
Allow core file without exec file on RS/6000.
702
 
703
Make sure "shell" with no arguments works right on DOS.
704
 
705
Make gdb.ini (as well as .gdbinit) be checked on all platforms, so
706
the same directory can be NFS-mounted on unix or DOS, and work the
707
same way.
708
 
709
cd ~/tmp/ causes infinite loop (where ~/tmp is a directory).
710
 
711
Get SECT_OFF_TEXT stuff out of objfile_relocate (might be needed to
712
get RS/6000 to work right, might not be immediately relevant).
713
 
714
Clean up add_toc_to_loadinfo
715
 
716
Think about attached processes and sharing terminal.
717
 
718
John sez in reference to ignoring errors from tcsegpgrp if attach_flag:
719
set_tty_state should not have any trouble with attached processes.
720
Instead, the tty handling should leave the pgrp of the tty alone when
721
attaching to processes (perhaps pass terminal_init_inferior a flag
722
saying whether we're attaching).
723
 
724
PAGE_SIZE redefined warnings on AIX.  Probably should be using
725
BFD_PAGE_SIZE throughout BFD.
726
 
727
Rewrite proceed, wait_for_inferior, and normal_stop to clean them up.
728
Suggestions:
729
 
730
        1) Make each test in wait_for_inferior a seperate subroutine
731
           call.
732
        2) Combine wait_for_inferior and normal_stop to clean up
733
           communication via global variables.
734
        3) See if you can find some way to clean up the global
735
           variables that are used; possibly group them by data flow
736
           and information content?
737
 
738
Work out some kind of way to allow running the inferior to be done as
739
a sub-execution of, eg. breakpoint command lists.  Currently running
740
the inferior interupts any command list execution.  This would require
741
some rewriting of wait_for_inferior & friends, and hence should
742
probably be done in concert with the above.
743
 
744
Add function arguments to gdb user defined functions.
745
 
746
Add convenience variables that refer to exec file, symbol file,
747
selected frame source file, selected frame function, selected frame
748
line number, etc.
749
 
750
Add a "suspend" subcommand of the "continue" command to suspend gdb
751
while continuing  execution of the subprocess.  Useful when you are
752
debugging servers and you want to dodge out and initiate a connection
753
to a server running under gdb.
754
 
755
Add stab information to allow reasonable debugging of inline functions
756
(possibly they should show up on a stack backtrace?  With a note
757
indicating that they weren't "real"?).
758
 
759
Modify the naked "until" command to step until past the current source
760
line, rather than past the current pc value.  This is tricky simply
761
because the low level routines have no way of specifying a multi-line
762
step range, and there is no way of saying "don't print stuff when we
763
stop" from above (otherwise could just call step many times).
764
 
765
Modify the handling of symbols grouped through BINCL/EINCL stabs to
766
allocate a partial symtab for each BINCL/EINCL grouping.  This will
767
seriously decrease the size of inter-psymtab dependencies and hence
768
lessen the amount that needs to be read in when a new source file is
769
accessed.
770
 
771
Do an "x/i $pc" after each stepi or nexti.
772
 
773
Modify all of the disassemblers to use printf_filtered to get correct
774
more filtering.
775
 
776
Modify gdb to work correctly with Pascal.
777
 
778
Add a command for searching memory, a la adb.  It specifies size,
779
mask, value, start address.  ADB searches until it finds it or hits
780
an error (or is interrupted).
781
 
782
Remove the range and type checking code and documentation, if not
783
going to implement.
784
 
785
# Local Variables:
786
# mode: text
787
# End:

powered by: WebSVN 2.1.0

© copyright 1999-2024 OpenCores.org, equivalent to Oliscience, all rights reserved. OpenCores®, registered trademark.