gdb/TODO

   1 If you find inaccuracies in this list, please send mail to
   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.
   5
   6
   7                         GDB 5.1 - Fixes
   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 5.1.
  12
  13 -- 2001-03-08
  14
  15 Update GDB's coding standard documentation.  Known topics:
  16
  17        o     alloca/malloc et.al.
  18
  19        o     typedef and structs
  20
  21        o        ISO-C
  22
  23 and most likely also:
  24
  25        o         include conventions
  26
  27 --
  28
  29 Wow, three bug reports for the same problem in one day!  We should
  30 probably make fixing this a real priority :-).
  31
  32 Anyway, thanks for reporting.
  33
  34 The following patch will fix the problems with setting breakpoints in
  35 dynamically loaded objects:
  36
  37    http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00230.html
  38
  39 This patch isn't checked in yet (ping Michael/JimB), but I hope this
  40 will be in the next GDB release.
  41
  42 There should really be a test in the testsuite for this problem, since
  43 it keeps coming up :-(.  Any volunteers?
  44
  45 Mark
  46
  47 --
  48
  49 x86 linux GDB and SIGALRM (???)
  50 http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00803.html
  51
  52 This problem has been fixed, but a regression test still needs to be
  53 added to the testsuite:
  54 http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00309.html
  55
  56 Mark
  57
  58 [The test has been submitted for approval - cagney]
  59
  60 --
  61
  62 RFD: infrun.c: No bpstat_stop_status call after proceed over break?
  63 http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00665.html
  64
  65 GDB misses watchpoint triggers after proceeding over a breakpoint on
  66 x86 targets.
  67
  68 --
  69
  70 GDB 5.0 doesn't work on Linux/SPARC
  71
  72 There are two parts to this.
  73
  74       o   GDB 5.0 doesn't work on GNU/Linux/SPARC32
  75
  76       o   GDB 5.0 doesn't work on the new target
  77           GNU/Linux/SPARC64
  78
  79 GDB does build on both these targets.
  80
  81 The first problem is the one that should be fixed.
  82
  83 --
  84
  85                 GDB 5.1 - New features
  86                 ======================
  87
  88 The following new features should be included in 5.1.
  89
  90 --
  91
  92 Enable MI by default.  Old code can be deleted after 5.1 is out.
  93
  94 Issues:
  95
  96         o       syntax change where a list would
  97                 look like:
  98                      [ foo=a, foo=b, foo=c ]
  99                 instead of
 100                      { foo=a, foo=b, foo=c }
 101
 102         o       kill off the idea of a reverse
 103                 query.
 104
 105         o       review test cases
 106
 107         o       enable it
 108
 109 --
 110
 111 Pascal (Pierre Muller, David Taylor)
 112
 113 Pierre Muller has contributed patches for adding Pascal Language
 114 support to GDB.
 115
 116 2 pascal language patches inserted in database
 117 http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00521.html
 118
 119 Indent -gnu ?
 120 http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00496.html
 121
 122 [I think this has been merged, need to confirm - cagney]
 123
 124 --
 125
 126 Java (Anthony Green, David Taylor)
 127
 128 Anthony Green has a number of Java patches that did not make it into
 129 the 5.0 release.  The first two are in cvs now, but the third needs
 130 some fixing up before it can go in.
 131
 132 Patch: java tests
 133 http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00512.html
 134
 135 Patch: java booleans
 136 http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00515.html
 137
 138 Patch: handle N_MAIN stab
 139 http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00527.html
 140
 141 -- 2001-03-08
 142
 143 Add CRIS target.
 144
 145 A predicate to this is the multi-arching of SOFTWARE_SINGLE_STEP().  A
 146 patch has been submitted.
 147
 148 --
 149
 150                 GDB 5.1 - Cleanups
 151                 ==================
 152
 153 The following code cleanups will hopefully be applied to GDB 5.1.
 154
 155 -- 2001-03-14
 156
 157 Resolve the build status of all broken targets as identified by the
 158 MAINTAINERS file.
 159
 160         o       arm-* vs NetBSD's lack of ``unix''
 161         o       arm-* vs IRIX (see below)
 162
 163         o       delete m88k?
 164
 165 -- 2001-03-15
 166
 167                         Obsolete some targets.
 168
 169 Possible selection criteria are:
 170
 171         o       uses a deprecated feature
 172
 173         o       doesn't build
 174
 175         o       doesn't have a maintainer
 176
 177 Steps:
 178
 179         o       post proposals to gdb@ (DONE)
 180
 181         o       post announcement to gdb-announce@
 182                 crossed with gdb@ reply-to to gdb@
 183                 (DONE)
 184
 185                         ns32k-*-mach3*
 186                         ns32k-umax-*
 187                         ns32k-utek-sysv*
 188                         tic80-*
 189                         m68*-isi-*
 190                         m68*-sony-*
 191                         a29k-ultra3
 192                         powerpcle-*-solaris*
 193                         powerpcle-*-cygwin*
 194                         w65-*-*
 195                         i[3456]86-*-sunos*
 196
 197         o       clobber the files:
 198
 199                         configure.{in,host,tgt}
 200                         Makefile.in
 201                         *-tdep.c *-nat.c *-xdep.c
 202                         configure/*/*
 203
 204         o       update NEWS
 205
 206 --
 207
 208 Change documentation to GFDL license.
 209
 210 ``It is time to make an effort to start using the GFDL more
 211 thoroughly.  Would all GNU maintainers please change the license to
 212 the GFDL, for all manuals and other major documentation files?
 213
 214 The GFDL and some instructions for using it can be found in
 215 http://www.gnu.org/copyleft/''
 216
 217         RMS
 218
 219 --
 220
 221 Fix copyright notices.
 222
 223 Turns out that ``1998-2000'' isn't considered valid :-(
 224
 225 http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00467.html
 226
 227 --
 228
 229                 GDB 5.1 - Known Problems
 230                 ========================
 231
 232 --
 233
 234 z8k
 235
 236 The z8k has suffered bit rot and is known to not build.  The problem
 237 was occuring in the opcodes directory.
 238
 239 --
 240
 241 Solaris 8 x86 CURSES_H problem
 242 http://sources.redhat.com/ml/gdb/2000-07/msg00038.html
 243
 244 The original problem was worked around with:
 245
 246     2000-06-06  Michael Snyder  <msnyder@cygnus.com>
 247
 248         * configure.in: Enable autoconf to find curses.h on Solaris 2.8.
 249         * configure: Regenerate.
 250
 251 When building both GDB and SID using the same source tree the problem
 252 will still occure. sid/component/configure.in mis-configures
 253 <curses.h> and leaves wrong information in the config cache.
 254
 255 --
 256
 257                 GDB 5.2 - Fixes
 258                 ===============
 259
 260 --
 261
 262 Thread support.  Right now, as soon as a thread finishes and exits,
 263 you're hosed.  This problem is reported once a week or so.
 264
 265 --
 266
 267                 GDB 5.2 - New features
 268                 ======================
 269
 270 --
 271
 272 GCC 3.0 ABI support (but hopefully sooner...).
 273
 274 --
 275
 276 Objective C/C++ support (but hopefully sooner...).
 277
 278 --
 279
 280                 GDB 5.2 - Cleanups
 281                 ==================
 282
 283 The following cleanups have been identified as part of GDB 5.2.
 284
 285 --
 286
 287 Remove old code that does not use ui_out functions and all the related
 288 "ifdef"s.  This also allows the elimination of -DUI_OUT from
 289 Makefile.in and configure.in.
 290
 291 --
 292
 293 Compiler warnings.
 294
 295 Eliminate all warnings for at least one host/target for the flags:
 296 -Wimplicit -Wreturn-type -Wcomment -Wtrigraphs -Wformat -Wparentheses
 297 -Wpointer-arith -Wuninitialized
 298
 299 --
 300
 301 Restructure gdb directory tree so that it avoids any 8.3 and 14
 302 filename problems.
 303
 304 --
 305
 306 Convert GDB build process to AUTOMAKE.
 307
 308 See also sub-directory configure below.
 309
 310 The current convention is (kind of) to use $(<header>_h) in all
 311 dependency lists.  It isn't done in a consistent way.
 312
 313 --
 314
 315                 GDB 5.2 - Known Problems
 316                 ========================
 317
 318 --
 319
 320                 Code Cleanups: General
 321                 ======================
 322
 323 The following are more general cleanups and fixes.  They are not tied
 324 to any specific release.
 325
 326 --
 327
 328 Revised UDP support (was: Re: [Fwd: [patch] UDP transport support])
 329 http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00000.html
 330
 331 (Broken) support for GDB's remote protocol across UDP is to be
 332 included in the follow-on release.
 333
 334 It should be noted that UDP can only work when the [Gg] packet fits in
 335 a single UDP packet.
 336
 337 There is also much debate over the merit of this.
 338
 339 --
 340
 341 Can't build IRIX -> arm GDB.
 342 http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00356.html
 343
 344 David Whedon writes:
 345 > Now I'm building for an embedded arm target.  If there is a way of turning
 346 > remote-rdi off, I couldn't find it. It looks like it gets built by default
 347 > in gdb/configure.tgt(line 58) Anyway, the build dies in
 348 > gdb/rdi-share/unixcomm.c.  SERPORT1 et. al. never get defined because we
 349 > aren't one of the architectures supported.
 350
 351 --
 352
 353 Problem with weak functions
 354 http://sourceware.cygnus.com/ml/gdb/2000-05/msg00060.html
 355
 356 Dan Nicolaescu writes:
 357 > It seems that gdb-4.95.1  does not display correctly the function when
 358 > stoping in weak functions.
 359 >
 360 > It stops in a function that is defined as weak, not in the function
 361 > that is actually run...
 362
 363 --
 364
 365 Follow through `make check' with --enable-shared.
 366
 367 When the srcware tree is configured with --enable-shared, the `expect'
 368 program won't run properly.  Jim Wilson found out gdb has a local hack
 369 to set LD_LIBRARY_PATH, but, AFAIK, no other project has been hacked
 370 similarly.
 371
 372 http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00845.html
 373
 374 --
 375
 376 Delete macro TARGET_BYTE_ORDER_SELECTABLE.
 377
 378 Patches in the database.
 379
 380 --
 381
 382 printcmd.c (print_address_numeric):
 383
 384 NOTE: This assumes that the significant address information is kept in
 385 the least significant bits of ADDR - the upper bits were either zero
 386 or sign extended.  Should ADDRESS_TO_POINTER() or some
 387 ADDRESS_TO_PRINTABLE() be used to do the conversion?
 388
 389 --
 390
 391 The BFD directory requires bug-fixed AUTOMAKE et.al.
 392
 393 AUTOMAKE 1.4 incorrectly set the TEXINPUTS environment variable.  It
 394 contained the full path to texinfo.tex when it should have only
 395 contained the directory.  The bug has been fixed in the current
 396 AUTOMAKE sources.  Automake snapshots can be found in:
 397         ftp://sourceware.cygnus.com/pub/gdb/snapshots
 398 and     ftp://sourceware.cygnus.com/pub/binutils
 399
 400 --
 401
 402 Find something better than DEFAULT_BFD_ARCH, DEFAULT_BFD_VEC to
 403 determine the default isa/byte-order.
 404
 405 --
 406
 407 Rely on BFD_BIG_ENDIAN and BFD_LITTLE_ENDIAN instead of host dependent
 408 BIG_ENDIAN and LITTLE_ENDIAN.
 409
 410 --
 411
 412 Eliminate more compiler warnings.
 413
 414 Of course there also needs to be the usual debate over which warnings
 415 are valid and how to best go about this.
 416
 417 One method: choose a single option; get agreement that it is
 418 reasonable; try it out to see if there isn't anything silly about it
 419 (-Wunused-parameters is an example of that) then incrementally hack
 420 away.
 421
 422 The other method is to enable all warnings and eliminate them from one
 423 file at a time.
 424
 425 --
 426
 427 Elimination of ``(catch_errors_ftype *) func''.
 428
 429 Like make_cleanup_func it isn't portable.
 430 http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00791.html
 431 http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00814.html
 432
 433 --
 434
 435 Nuke #define CONST_PTR.
 436
 437 --
 438
 439 Nuke USG define.
 440
 441 --
 442
 443 [PATCH/5] src/intl/Makefile.in:distclean additions
 444 http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00363.html
 445
 446 Do not forget to merge the patch back into the trunk.
 447
 448 --
 449
 450 Rationalize the host-endian code (grep for HOST_BYTE_ORDER).
 451
 452 At present defs.h includes <endian.h> (which is linux specific) yet
 453 almost nothing depends on it.  Suggest "gdb_endian.h" which can also
 454 handle <machine/endian.h> and only include that where it is really
 455 needed.
 456
 457 --
 458
 459 Replace savestring() with something from libiberty.
 460
 461 An xstrldup()? but that would have different semantics.
 462
 463 --
 464
 465 Rationalize use of floatformat_unknown in GDB sources.
 466
 467 Instead of defaulting to floatformat_unknown, should hosts/targets
 468 specify the value explicitly?
 469
 470 http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00447.html
 471
 472 --
 473
 474 Add a ``name'' member to include/floatformat.h:struct floatformat.
 475 Print that name in gdbarch.c.
 476
 477 --
 478
 479 Sort out the harris mess in include/floatformat.h (it hardwires two
 480 different floating point formats).
 481
 482 --
 483
 484 See of the GDB local floatformat_do_doublest() and libiberty's
 485 floatformat_to_double (which was once GDB's ...) can be merged some
 486 how.
 487
 488 --
 489
 490 Eliminate mmalloc(), mstrsave() et.al. from GDB.
 491
 492 Also eliminate it from defs.h.
 493
 494 --
 495
 496 Eliminate PTR.  ISO-C allows ``void *''.
 497
 498 --
 499
 500 Eliminate abort ().
 501
 502 GDB should never abort.  GDB should either throw ``error ()'' or
 503 ``internal_error ()''.  Better still GDB should naturally unwind with
 504 an error status.
 505
 506 --
 507
 508 GDB probably doesn't build on FreeBSD pre 2.2.x
 509 http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00378.html
 510
 511 Fixes to get FreeBSD working on 2.2.x, 3.x and 4.x caused the code to
 512 suffer bit rot.
 513
 514 --
 515
 516 Deprecate "fg".  Apparently ``fg'' is actually continue.
 517
 518 http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00417.html
 519
 520 --
 521
 522 Deprecate current use of ``floatformat_unknown''.
 523
 524 Require all targets to explicitly provide their float format instead
 525 of defaulting to floatformat unknown.  Doing the latter leads to nasty
 526 bugs.
 527
 528 http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00447.html
 529
 530 --
 531
 532 Rationalize floatformat_to_double() vs floatformat_to_doublest().
 533
 534 Looks like GDB migrated floatformat_to_double() to libiberty but then
 535 turned around and created a ..._to_doublest() the latter containing
 536 several bug fixes.
 537
 538 http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00472.html
 539
 540 --
 541
 542 Move floatformat_ia64_ext to libiberty/include floatformat.[ch].
 543
 544 http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00466.html
 545
 546 --
 547
 548 The ``maintenance deprecate set endian big'' command doesn't notice
 549 that it is deprecating ``set endian'' and not ``set endian big'' (big
 550 is implemented using an enum).  Is anyone going to notice this?
 551
 552 --
 553
 554 When tab expanding something like ``set arch<tab>'' ignore the
 555 deprecated ``set archdebug'' and expand to ``set architecture''.
 556
 557 --
 558
 559 Eliminate ``arm_register_names[j] = (char *) regnames[j]'' and the
 560 like from arm-tdep.c.
 561
 562 --
 563
 564 Fix uses of ->function.cfunc = set_function().
 565
 566 The command.c code calls sfunc() when a set command.  Rather than
 567 change it suggest fixing the callback function so that it is more
 568 useful.  See:
 569
 570 http://sourceware.cygnus.com/ml/gdb-patches/2000-06/msg00062.html
 571
 572 See also ``Fix implementation of  ``target xxx''.'' below.
 573
 574 --
 575
 576 IRIX 3.x support is probably broken.
 577
 578 --
 579
 580 Delete sim/SIM_HAVE_BREAKPOINTS and gdb/SIM_HAS_BREAKPOINTS.
 581 http://sourceware.cygnus.com/ml/gdb-patches/2000-07/msg00042.html
 582
 583 Apart from the d30v, are there any sim/common simulators that make use
 584 of this?
 585
 586 A brief summary of what happened is that sim/common/sim-break.c was
 587 created as a good idea.  It turned out a better idea was to use
 588 SIM_SIGBREAK and have GDB pass back sim_resume (..., SIGBREAK).
 589
 590 --
 591
 592 Move remote_remove_hw_breakpoint, remote_insert_hw_breakpoint,
 593 remote_remove_watchpoint, remote_insert_watchpoint into target vector.
 594
 595 --
 596
 597 Eliminate ``extern'' from C files.
 598
 599 --
 600
 601 Replace ``STREQ()'' et.al. with ``strcmp() == 0'' et.al.
 602
 603 Extreme care is recommeded - perhaps only modify tests that are
 604 exercised by the testsuite (as determined using some type of code
 605 coverage analysis).
 606
 607 --
 608
 609 Replace the file gdb/CONTRIBUTE with a file that is generated from the
 610 gdb/doc/*.texinfo directory.
 611
 612 --
 613
 614                         New Features and Fixes
 615                         ======================
 616
 617 These are harder than cleanups but easier than work involving
 618 fundamental architectural change.
 619
 620 --
 621
 622 Hardware watchpoint problems on x86 OSes, including Linux:
 623
 624 1. Delete/disable hardware watchpoints should free hardware debug
 625 registers.
 626 2. Watch for different values on a viariable with one hardware debug
 627 register.
 628
 629 According to Eli Zaretskii <eliz@delorie.com>:
 630
 631 These are not GDB/ia32 issues per se: the above features are all
 632 implemented in the DJGPP port of GDB and work in v5.0.  Every
 633 x86-based target should be able to lift the relevant parts of
 634 go32-nat.c and use them almost verbatim.  You get debug register
 635 sharing through reference counts, and the ability to watch large
 636 regions (up to 16 bytes) using multiple registers.  (The required
 637 infrastructure in high-level GDB application code, mostly in
 638 breakpoint.c, is also working since v5.0.)
 639
 640 --
 641
 642 Add built-by, build-date, tm, xm, nm and anything else into gdb binary
 643 so that you can see how the GDB was created.
 644
 645 --
 646
 647 Add an "info bfd" command that displays supported object formats,
 648 similarly to objdump -i.
 649
 650 Is there a command already?
 651
 652 --
 653
 654 Fix ``I'm sorry, Dave, I can't do that.'' from symfile.c.
 655
 656 This requires internationalization.
 657
 658 --
 659
 660 Add support for:
 661
 662 (gdb) p fwprintf(stdout,L"%S\n", f)
 663 No symbol "L" in current context.
 664
 665 --
 666
 667 Cleanup configury support for optional sub-directories.
 668
 669 Check how GCC handles multiple front ends for an example of how things
 670 could work.  A tentative first step is to rationalize things so that
 671 all sub directories are handled in a fashion similar to gdb/mi.
 672
 673 See also automake above.
 674
 675 --
 676
 677 Add a transcript mechanism to GDB.
 678
 679 Such a mechanism might log all gdb input and output to a file in a
 680 form that would allow it to be replayed.  It could involve ``gdb
 681 --transcript=FILE'' or it could involve ``(gdb) transcript file''.
 682
 683 --
 684
 685 Can the xdep files be replaced by autoconf?
 686
 687 --
 688
 689 Document trace machinery
 690
 691 --
 692
 693 Document ui-out and ui-file.
 694
 695 http://sourceware.cygnus.com/ml/gdb/2000-04/msg00121.html
 696
 697 --
 698
 699 Update texinfo.tex to latest?
 700
 701 --
 702
 703 Incorporate agentexpr.texi into gdb.texinfo
 704
 705 agentexpr.texi mostly describes the details of the byte code used for
 706 tracepoints, not the internals of the support for this in GDB.  So it
 707 looks like gdb.texinfo is a better place for this information.
 708
 709 http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00566.html
 710
 711 --
 712
 713 Document overlay machinery.
 714
 715 --
 716
 717 ``(gdb) catch signal SIGNAL''
 718
 719 Overlaps with ``handle SIGNAL'' but the implied behavior is different.
 720 You can attach commands to a catch but not a handle.  A handle has a
 721 limited number of hardwired actions.
 722
 723 --
 724
 725 Get the TUI working on all platforms.
 726
 727 --
 728
 729 Add support for ``gdb --- PROGRAM ARGS ...''.
 730 Add support for ``gdb -cmd=...''
 731
 732 Along with many variations.  Check:
 733
 734 ????? for a full discussion.
 735
 736 for a discussion.
 737
 738 --
 739
 740 Implement ``(gdb) !ls''.
 741
 742 Which is very different from ``(gdb) ! ls''.  Implementing the latter
 743 is trivial.
 744
 745 http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00034.html
 746
 747 --
 748
 749 Change the (char *list[]) to (const char (*)[]) so that dynamic lists can
 750 be passed.
 751
 752 --
 753
 754 When tab expanding something like ``set arch<tab>'' ignore the
 755 deprecated ``set archdebug'' and expand to ``set architecture''.
 756
 757 --
 758
 759 Replace the code that uses the host FPU with an emulator of the target
 760 FPU.
 761
 762 --
 763
 764 The "ocd reset" command needs to flush the dcache, which requires breaking
 765 the abstraction layer between the target independent and target code.  One
 766 way to address this is provide a generic "reset" command and target vector.
 767
 768 http://sources.redhat.com/ml/gdb-patches/2000-10/msg00011.html
 769
 770 --
 771
 772                         Thread Support
 773                         ==============
 774
 775 --
 776
 777 Generic: lin-thread cannot handle thread exit (Mark Kettenis, Michael
 778 Snyder) http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00525.html
 779
 780 The thread_db assisted debugging code doesn't handle exiting threads
 781 properly, at least in combination with glibc 2.1.3 (the framework is
 782 there, just not the actual code).  There are at least two problems
 783 that prevent this from working.
 784
 785 As an additional reference point, the pre thread_db code did not work
 786 either.
 787
 788 --
 789
 790 GNU/Linux/x86 and random thread signals (and Solaris/SPARC but not
 791 Solaris/x86).
 792 http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00336.html
 793
 794 Christopher Blizzard writes:
 795
 796 So, I've done some more digging into this and it looks like Jim
 797 Kingdon has reported this problem in the past:
 798
 799 http://sourceware.cygnus.com/ml/bug-gdb/1999-10/msg00058.html
 800
 801 I can reproduce this problem both with and without Tom's patch.  Has
 802 anyone seen this before?  Maybe have a solution for it hanging around?
 803 :)
 804
 805 There's a test case for this documented at:
 806
 807 when debugging threaded applications you get extra SIGTRAPs
 808 http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=9565
 809
 810 [There should be a GDB testcase - cagney]
 811
 812 --
 813
 814 GDB5 TOT on unixware 7
 815 http://sourceware.cygnus.com/ml/gdb/2000-04/msg00119.html
 816
 817 Robert Lipe writes:
 818 > I just spun the top of tree of the GDB5 branch on UnixWare 7.  As a
 819 > practical matter, the current thread support is somewhat more annoying
 820 > than when GDB was thread-unaware.
 821
 822 --
 823
 824 Migrate qfThreadInfo packet -> qThreadInfo. (Andrew Cagney)
 825
 826 Add support for packet enable/disable commands with these thread
 827 packets.  General cleanup.
 828
 829 [PATCH] Document the ThreadInfo remote protocol queries
 830 http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00832.html
 831
 832 [PATCH] "info threads" queries for remote.c
 833 http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00831.html
 834
 835 --
 836
 837                         Language Support
 838                         ================
 839
 840 New languages come onto the scene all the time.
 841
 842 --
 843
 844 Re: Various C++ things
 845
 846 value_headof/value_from_vtable_info are worthless, and should be
 847 removed.  The one place in printcmd.c that uses it should use the RTTI
 848 functions.
 849
 850 RTTI for g++ should be using the typeinfo functions rather than the
 851 vtables.  The typeinfo functions are always at offset 4 from the
 852 beginning of the vtable, and are always right. The vtables will have
 853 weird names like E::VB sometimes.  The typeinfo function will always
 854 be "E type_info function", or somesuch.
 855
 856 value_virtual_fn_field needs to be fixed so there are no failures for
 857 virtual functions for C++ using g++.
 858
 859 Testsuite cases are the major priority right now for C++ support,
 860 since i have to make a lot of changes that could potentially break
 861 each other.
 862
 863 --
 864
 865 Add support for Modula3
 866
 867 Get DEC/Compaq to contribute their Modula-3 support.
 868
 869 --
 870
 871                 Remote Protocol Support
 872                 =======================
 873
 874 --
 875
 876 Remote protocol doco feedback.
 877
 878 Too much feedback to mention needs to be merged in (901660).  Search
 879 for the word ``remote''.
 880
 881
 882 http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00023.html
 883 http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00056.html
 884 http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00382.html
 885
 886 --
 887
 888 GDB doesn't recover gracefully from remote protocol errors.
 889
 890 GDB wasn't checking for NAKs from the remote target.  Instead a NAK is
 891 ignored and a timeout is required before GDB retries.  A pre-cursor to
 892 fixing this this is making GDB's remote protocol packet more robust.
 893
 894 While downloading to a remote protocol target, gdb ignores packet
 895 errors in so far as it will continue to download with chunk N+1 even
 896 if chunk N was not correctly sent.  This causes gdb.base/remote.exp to
 897 take a painfully long time to run.  As a PS that test needs to be
 898 fixed so that it builds on 16 bit machines.
 899
 900 --
 901
 902 Add the cycle step command.
 903
 904 http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00237.html
 905
 906 --
 907
 908 Resolve how to scale things to support very large packets.
 909
 910 --
 911
 912 Resolve how to handle a target that changes things like its endianess
 913 on the fly - should it be returned in the ``T'' packet?
 914
 915 Underlying problem is that the register file is target endian.  If the
 916 target endianess changes gdb doesn't know.
 917
 918 --
 919
 920 Rename read_register{,_pid}() to read_unsigned_register{,_pid}().
 921
 922 --
 923
 924                         Symbol Support
 925                         ==============
 926
 927 If / when GDB starts to support the debugging of multi-processor
 928 (rather than multi-thread) applications the symtab code will need to
 929 be updated a little so that several independent symbol tables are
 930 active at a given time.
 931
 932 The other interesting change is a clarification of the exact meaning
 933 of CORE_ADDR and that has had consequences for a few targets (that
 934 were abusing that data type).
 935
 936 --
 937
 938 Investiagate ways of reducing memory.
 939
 940 --
 941
 942 Investigate ways of improving load time.
 943
 944 --
 945
 946 Get the d10v to use POINTER_TO_ADDRESS and ADDRESS_TO_POINTER.
 947
 948 Consequence of recent symtab clarification.  No marks for figuring out
 949 who maintains the d10v.
 950
 951 --
 952
 953 Get the MIPS to correctly sign extend all address <-> pointer
 954 conversions.
 955
 956 Consequence of recent symtab clarification.  No marks for figuring out
 957 who maintains the MIPS.
 958
 959 --
 960
 961 GDB truncates 64 bit enums.
 962
 963 http://sourceware.cygnus.com/ml/gdb-patches/2000-06/msg00290.html
 964
 965 --
 966
 967                         Testsuite Support
 968                         =================
 969
 970 There are never to many testcases.
 971
 972 --
 973
 974 Better thread testsuite.
 975
 976 --
 977
 978 Better C++ testsuite.
 979
 980 --
 981
 982 Look at adding a GDB specific testsuite directory so that white box
 983 tests of key internals can be added (eg ui_file).
 984
 985 --
 986
 987 Separate out tests that involve the floating point (FP).
 988
 989 (Something for people brining up new targets).  FP and non-fp tests
 990 are combined.  I think there should be set of basic tests that
 991 exercise pure integer support and then a more expanded set that
 992 exercise FP and FP/integer interactions.
 993
 994 As an example, the MIPS, for n32 as problems with passing FP's and
 995 structs.  Since most inferior call tests include FP it is difficult to
 996 determine of the integer tests are ok.
 997
 998 --
 999
1000                 Architectural Changes: General
1001                 ==============================
1002
1003 These are harder than simple cleanups / fixes and, consequently
1004 involve more work.  Typically an Architectural Change will be broken
1005 down into a more digestible set of cleanups and fixes.
1006
1007 --
1008
1009 Cleanup software single step.
1010
1011 At present many targets implement software single step by directly
1012 blatting memory (see rs6000-tdep.c).  Those targets should register
1013 the applicable breakpoints using the breakpoint framework.  Perhaphs a
1014 new internal breakpoint class ``step'' is needed.
1015
1016 --
1017
1018 Replace READ_FP() with FRAME_HANDLE().
1019
1020 READ_FP() is a hangover from the days of the vax when the ABI really
1021 did have a frame pointer register.  Modern architectures typically
1022 construct a virtual frame-handle from the stack pointer and various
1023 other bits of string.
1024
1025 Unfortunately GDB still treats this synthetic FP register as though it
1026 is real.  That in turn really confuses users (arm and ``print $fp'' VS
1027 ``info registers fp'').  The synthetic FP should be separated out of
1028 the true register set presented to the user.
1029
1030 --
1031
1032 Register Cache Cleanup (below from Andrew Cagney)
1033
1034 I would depict the current register architecture as something like:
1035
1036         High GDB  -->   Low GDB
1037             |              |
1038            \|/            \|/
1039             --- REG NR -----
1040                    |
1041     register + REGISTER_BYTE(reg_nr)
1042                    |
1043                   \|/
1044         -------------------------
1045         |  extern register[]    |
1046         -------------------------
1047
1048 where neither the high (valops.c et.al.) or low gdb (*-tdep.c) are
1049 really clear on what mechanisms they should be using to manipulate that
1050 buffer.   Further, much code assumes, dangerously, that registers are
1051 contigious.  Having got mips-tdep.c to support multiple ABIs, believe
1052 me, that is a bad assumption.  Finally, that register cache layout is
1053 determined by the current remote/local target and _not_ the less
1054 specific target ISA.  In fact, in many cases it is determined by the
1055 somewhat arbitrary layout of the [gG] packets!
1056
1057
1058 How I would like the register file to work is more like:
1059
1060
1061           High GDB
1062              |
1063             \|/
1064        pseudo reg-nr
1065              |
1066         map pseudo <->
1067         random cache
1068         bytes
1069              |
1070             \|/
1071         ------------
1072         | register |
1073         |  cache   |
1074         ------------
1075             /|\
1076              |
1077         map random cache
1078         bytes to target
1079         dependent i-face
1080             /|\
1081              |
1082        target dependent
1083         such as [gG] packet
1084         or ptrace buffer
1085
1086 The main objectives being:
1087
1088         o       a clear separation between the low
1089                 level target and the high level GDB
1090
1091         o       a mechanism that solves the general
1092                 problem of register aliases, overlaps
1093                 etc instead of treating them as optional
1094                 extras that can be wedged in as an after
1095                 thought (that is a reasonable description
1096                 of the current code).
1097
1098                 Identify then solve the hard case and the
1099                 rest just falls out.  GDB solved the easy
1100                 case and then tried to ignore the real
1101                 world :-)
1102
1103         o       a removal of the assumption that the
1104                 mapping between the register cache
1105                 and virtual registers is largely static.
1106                 If you flip the USR/SSR stack register
1107                 select bit in the status-register then
1108                 the corresponding stack registers should
1109                 reflect the change.
1110
1111         o       a mechanism that clearly separates the
1112                 gdb internal register cache from any
1113                 target (not architecture) dependent
1114                 specifics such as [gG] packets.
1115
1116 Of course, like anything, it sounds good in theory.  In reality, it
1117 would have to contend with many<->many relationships at both the
1118 virt<->cache and cache<->target level.  For instance:
1119
1120         virt<->cache
1121                 Modifying an mmx register may involve
1122                 scattering values across both FP and
1123                 mmpx specific parts of a buffer
1124
1125         cache<->target
1126                 When writing back a SP it may need to
1127                 both be written to both SP and USP.
1128
1129
1130 Hmm,
1131
1132 Rather than let this like the last time it was discussed, just slip, I'm
1133 first going to add this e-mail (+ references) to TODO.  I'd then like to
1134 sketch out a broad strategy I think could get us there.
1135
1136
1137 First thing I'd suggest is separating out the ``extern registers[]''
1138 code so that we can at least identify what is using it.  At present
1139 things are scattered across many files.  That way we can at least
1140 pretend that there is a cache instead of a global array :-)
1141
1142 I'd then suggest someone putting up a proposal for the pseudo-reg /
1143 high-level side interface so that code can be adopted to it.  For old
1144 code, initially a blanket rename of write_register_bytes() to
1145 deprecated_write_register_bytes() would help.
1146
1147 Following that would, finaly be the corresponding changes to the target.
1148
1149 --
1150
1151 Check that GDB can handle all BFD architectures (Andrew Cagney)
1152
1153 There should be a test that checks that BFD/GDB are in sync with
1154 regard to architecture changes.  Something like a test that first
1155 queries GDB for all supported architectures and then feeds each back
1156 to GDB..  Anyone interested in learning how to write tests?  :-)
1157
1158 --
1159
1160                 Architectural Change: Multi-arch et al.
1161                 =======================================
1162
1163 The long term objective is to remove all assumptions that there is a
1164 single target with a single address space with a single instruction
1165 set architecture and single application binary interface.
1166
1167 This is an ongoing effort.  The first milestone is to enable
1168 ``multi-arch'' where by all architectural decisions are made at
1169 runtime.
1170
1171 It should be noted that ``gdbarch'' is really ``gdbabi'' and
1172 ``gdbisa''.  Once things are multi-arched breaking that down correctly
1173 will become much easier.
1174
1175 --
1176
1177 GDBARCH cleanup (Andrew Cagney)
1178
1179 The non-generated parts of gdbarch.{sh,h,c} should be separated out
1180 into arch-utils.[hc].
1181
1182 Document that gdbarch_init_ftype could easily fail because it didn't
1183 identify an architecture.
1184
1185 --
1186
1187 Fix BELIEVE_PPC_PROMOTION.  Change it to BELIEVE_PPC_PROMOTION_P?
1188
1189 At present there is still #ifdef BELIEVE_PPC_PROMOTION code in the
1190 symtab file.
1191
1192 --
1193
1194 Fix target_signal_from_host() etc.
1195
1196 The name is wrong for starters.  ``target_signal'' should probably be
1197 ``gdb_signal''.  ``from_host'' should be ``from_target_signal''.
1198 After that it needs to be multi-arched and made independent of any
1199 host signal numbering.
1200
1201 --
1202
1203 Update ALPHA so that it uses ``struct frame_extra_info'' instead of
1204 EXTRA_FRAME_INFO.
1205
1206 This is a barrier to replacing mips_extra_func_info with something
1207 that works with multi-arch.
1208
1209 --
1210
1211 Multi-arch  mips_extra_func_info.
1212
1213 This first needs the alpha to be updated so that it uses ``struct
1214 frame_extra_info''.
1215
1216 --
1217
1218 Rationalize TARGET_SINGLE_FORMAT and TARGET_SINGLE_BIT et al.
1219
1220 Surely one of them is redundant.
1221
1222 --
1223
1224 Convert ALL architectures to MULTI-ARCH.
1225
1226 --
1227
1228 Select the initial multi-arch ISA / ABI based on --target or similar.
1229
1230 At present the default is based on what ever is first in the BFD
1231 archures table.  It should be determined based on the ``--target=...''
1232 name.
1233
1234 --
1235
1236 Make MIPS pure multi-arch.
1237
1238 It is only at the multi-arch enabled stage.
1239
1240 --
1241
1242 Truly multi-arch.
1243
1244 Enable the code to recognize --enable-targets=.... like BINUTILS does.
1245
1246 Can the tm.h and nm.h files be eliminated by multi-arch.
1247
1248 --
1249
1250         Architectural Change: MI, LIBGDB and scripting languages
1251         ========================================================
1252
1253 See also architectural changes related to the event loop.  LIBGDB
1254 can't be finished until there is a generic event loop being used by
1255 all targets.
1256
1257 The long term objective is it to be possible to integrate GDB into
1258 scripting languages.
1259
1260 --
1261
1262 Implement generic ``(gdb) commmand > file''
1263
1264 Once everything is going through ui_file it should be come fairly
1265 easy.
1266
1267 http://sourceware.cygnus.com/ml/gdb/2000-04/msg00104.html
1268
1269 --
1270
1271 Replace gdb_stdtarg with gdb_targout (and possibly gdb_targerr).
1272
1273 gdb_stdtarg is easily confused with gdb_stdarg.
1274
1275 --
1276
1277 Extra ui_file methods - dump.
1278
1279 Very useful for whitebox testing.
1280
1281 --
1282
1283 Eliminate error_begin().
1284
1285 With ui_file, there is no need for the statefull error_begin ()
1286 function.
1287
1288 --
1289
1290 Send normal output to gdb_stdout.
1291 Send error messages to gdb_stderror.
1292 Send debug and log output log gdb_stdlog.
1293
1294 GDB still contains many cases where (f)printf or printf_filtered () is
1295 used when it should be sending the messages to gdb_stderror or
1296 gdb_stdlog.  The thought of #defining printf to something has crossed
1297 peoples minds ;-)
1298
1299 --
1300
1301 Re-do GDB's output pager.
1302
1303 GDB's output pager still relies on people correctly using *_filtered
1304 for gdb_stdout and *_unfiltered for gdb_stdlog / gdb_stderr.
1305 Hopefully, with all normal output going to gdb_stdout, the pager can
1306 just look at the ui_file that the output is on and then use that to
1307 decide what to do about paging.  Sounds good in theory.
1308
1309 --
1310
1311 Check/cleanup MI documentation.
1312
1313 The list of commands specified in the documentation needs to be
1314 checked against the mi-cmds.c table in a mechanical way (so that they
1315 two can be kept up-to-date).
1316
1317 --
1318
1319 Convert MI into libgdb
1320
1321 MI provides a text interface into what should be many of the libgdb
1322 functions.  The implementation of those functions should be separated
1323 into the MI interface and the functions proper.  Those functions being
1324 moved to gdb/lib say.
1325
1326 --
1327
1328 Create libgdb.h
1329
1330 The first part can already be found in defs.h.
1331
1332 --
1333
1334 MI's input does not use buffering.
1335
1336 At present the MI interface reads raw characters of from an unbuffered
1337 FD.  This is to avoid several nasty buffer/race conditions.  That code
1338 should be changed so that it registers its self with the event loop
1339 (on the input FD) and then push commands up to MI as they arrive.
1340
1341 The serial code already does this.
1342
1343 --
1344
1345 Make MI interface accessible from existing CLI.
1346
1347 --
1348
1349 Add a breakpoint-edit command to MI.
1350
1351 It would be similar to MI's breakpoint create but would apply to an
1352 existing breakpoint.  It saves the need to delete/create breakpoints
1353 when ever they are changed.
1354
1355 --
1356
1357 Add directory path to MI breakpoint.
1358
1359 That way the GUI's task of finding the file within which the
1360 breakpoint was set is simplified.
1361
1362 --
1363
1364 Add a mechanism to reject certain expression classes to MI
1365
1366 There are situtations where you don't want GDB's expression
1367 parser/evaluator to perform inferior function calls or variable
1368 assignments.  A way of restricting the expression parser so that such
1369 operations are not accepted would be very helpful.
1370
1371 --
1372
1373 Remove sideffects from libgdb breakpoint create function.
1374
1375 The user can use the CLI to create a breakpoint with partial
1376 information - no file (gdb would use the file from the last
1377 breakpoint).
1378
1379 The libgdb interface currently affects that environment which can lead
1380 to confusion when a user is setting breakpoints via both the MI and
1381 the CLI.
1382
1383 This is also a good example of how getting the CLI ``right'' will be
1384 hard.
1385
1386 --
1387
1388 Move gdb_lasterr to ui_out?
1389
1390 The way GDB throws errors and records them needs a re-think.  ui_out
1391 handles the correct output well.  It doesn't resolve what to do with
1392 output / error-messages when things go wrong.
1393
1394 --
1395
1396 do_setshow_command contains a 1024 byte buffer.
1397
1398 The function assumes that there will never be any more than 1024 bytes
1399 of enum.  It should use mem_file.
1400
1401 --
1402
1403 Should struct cmd_list_element . completer take the command as an
1404 argument?
1405
1406 --
1407
1408 Should the bulk of top.c:line_completion_function() be moved to
1409 command.[hc]?  complete_on_cmdlist() and complete_on_enums() could
1410 then be made private.
1411
1412 --
1413
1414 top.c (execute_command): Should a command being valid when the target
1415 is running be made an attribute (predicate) to the command rather than
1416 an explicit set of tests.
1417
1418 --
1419
1420 top.c (execute_command): Should the bulk of this function be moved
1421 into command.[hc] so that top.c doesn't grub around in the command
1422 internals?
1423
1424 --
1425
1426                 Architectural Change: Async
1427                 ===========================
1428
1429 While GDB uses an event loop when prompting the user for input.  That
1430 event loop is not exploited by targets when they allow the target
1431 program to continue.  Typically targets still block in (target_wait())
1432 until the program again halts.
1433
1434 The closest a target comes to supporting full asynchronous mode are
1435 the remote targets ``async'' and ``extended-async''.
1436
1437 --
1438
1439 Asynchronous expression evaluator
1440
1441 Inferior function calls hang GDB.
1442
1443 --
1444
1445 Fix implementation of  ``target xxx''.
1446
1447 At present when the user specifies ``target xxxx'', the CLI maps that
1448 directly onto a target open method.  It is then assumed that the
1449 target open method should do all sorts of complicated things as this
1450 is the only chance it has.  Check how the various remote targets
1451 duplicate the target operations.  Check also how the various targets
1452 behave differently for purely arbitrary reasons.
1453
1454 What should happen is that ``target xxxx'' should call a generic
1455 ``target'' function and that should then co-ordinate the opening of
1456 ``xxxx''.  This becomes especially important when you're trying to
1457 open an asynchronous target that may need to perform background tasks
1458 as part of the ``attach'' phase.
1459
1460 Unfortunately, due to limitations in the old/creaking command.h
1461 interface, that isn't possible.  The function being called isn't told
1462 of the ``xxx'' or any other context information.
1463
1464 Consequently a precursor to fixing ``target xxxx'' is to clean up the
1465 CLI code so that it passes to the callback function (attatched to a
1466 command) useful information such as the actual command and a context
1467 for that command.  Other changes such as making ``struct command''
1468 opaque may also help.
1469
1470 See also:
1471 http://sourceware.cygnus.com/ml/gdb-patches/2000-06/msg00062.html
1472
1473 --
1474
1475 Make "target xxx" command interruptible.
1476
1477 As things become async this becomes possible.  A target would start
1478 the connect and then return control to the event loop.  A cntrl-c
1479 would notify the target that the operation is to be abandoned and the
1480 target code could respond.
1481
1482 --
1483
1484 Add a "suspend" subcommand of the "continue" command to suspend gdb
1485 while continuing  execution of the subprocess.  Useful when you are
1486 debugging servers and you want to dodge out and initiate a connection
1487 to a server running under gdb.
1488
1489 [hey async!!]
1490
1491 --
1492
1493                         TODO FAQ
1494                         ========
1495
1496 Frequently requested but not approved requests.
1497
1498 --
1499
1500 Eliminate unused argument warnings using ATTRIBUTE_UNUSED.
1501
1502 The benefits on this one are thought to be marginal - GDBs design
1503 means that unused parameters are very common.  GCC 3.0 will also
1504 include the option -Wno-unused-parameter which means that ``-Wall
1505 -Wno-unused-parameters -Werror'' can be specified.
1506
1507 --
1508
1509
1510
1511                         Legacy Wish List
1512                         ================
1513
1514 This list is not up to date, and opinions vary about the importance or
1515 even desirability of some of the items.  If you do fix something, it
1516 always pays to check the below.
1517
1518 --
1519
1520 @c This does not work (yet if ever).  FIXME.
1521 @c @item --parse=@var{lang} @dots{}
1522 @c Configure the @value{GDBN} expression parser to parse the listed languages.
1523 @c @samp{all} configures @value{GDBN} for all supported languages.  To get a
1524 @c list of all supported languages, omit the argument.  Without this
1525 @c option, @value{GDBN} is configured to parse all supported languages.
1526
1527 --
1528
1529 START_INFERIOR_TRAPS_EXPECTED need never be defined to 2, since that
1530 is its default value.  Clean this up.
1531
1532 --
1533
1534 It should be possible to use symbols from shared libraries before we know
1535 exactly where the libraries will be loaded.  E.g. "b perror" before running
1536 the program.  This could maybe be done as an extension of the "breakpoint
1537 re-evaluation" after new symbols are loaded.
1538
1539 --
1540
1541 Make single_step() insert and remove breakpoints in one operation.
1542
1543 [If this is talking about having single_step() insert the breakpoints,
1544 run the target then pull the breakpoints then it is wrong.  The
1545 function has to return as control has to eventually be passed back to
1546 the main event loop.]
1547
1548 --
1549
1550 Speed up single stepping by avoiding extraneous ptrace calls.
1551
1552 --
1553
1554 Speed up single stepping by not inserting and removing breakpoints
1555 each time the inferior starts and stops.
1556
1557 Breakpoints should not be inserted and deleted all the time.  Only the
1558 one(s) there should be removed when we have to step over one.  Support
1559 breakpoints that don't have to be removed to step over them.
1560
1561 [this has resulted in numerous debates.  The issue isn't clear cut]
1562
1563 --
1564
1565 Provide "voodoo" debugging of core files.  This creates a zombie
1566 process as a child of the debugger, and loads it up with the data,
1567 stack, and regs of the core file.  This allows you to call functions
1568 in the executable, to manipulate the data in the core file.
1569
1570 [you wish]
1571
1572 --
1573
1574 GDB reopens the source file on every line, as you "next" through it.
1575
1576 [still true? I've a memory of this being fixed]
1577
1578 --
1579
1580 Perhaps "i source" should take an argument like that of "list".
1581
1582 --
1583
1584 Remove "at 0xnnnn" from the "b foo" response, if `print address off' and if
1585 it matches the source line indicated.
1586
1587 --
1588
1589 The prompt at end of screen should accept space as well as CR.
1590
1591 --
1592
1593 Backtrace should point out what the currently selected frame is, in
1594 its display, perhaps showing "@3 foo (bar, ...)" or ">3 foo (bar,
1595 ...)" rather than "#3 foo (bar, ...)".
1596
1597 --
1598
1599 "i program" should work for core files, and display more info, like what
1600 actually caused it to die.
1601
1602 --
1603
1604 "x/10i" should shorten the long name, if any, on subsequent lines.
1605
1606 --
1607
1608 "next" over a function that longjumps, never stops until next time you happen
1609 to get to that spot by accident.  E.g. "n" over execute_command which has
1610 an error.
1611
1612 --
1613
1614 "set zeroprint off", don't bother printing members of structs which
1615 are entirely zero.  Useful for those big structs with few useful
1616 members.
1617
1618 --
1619
1620 GDB does four ioctl's for every command, probably switching terminal modes
1621 to/from inferior or for readline or something.
1622
1623 --
1624
1625 terminal_ours versus terminal_inferior: cache state.  Switch should be a noop
1626 if the state is the same, too.
1627
1628 --
1629
1630 "i frame" shows wrong "arglist at" location, doesn't show where the args
1631 should be found, only their actual values.
1632
1633 --
1634
1635 There should be a way for "set" commands to validate the new setting
1636 before it takes effect.
1637
1638 --
1639
1640 "ena d" is ambiguous, why?  "ena delete" seems to think it is a command!
1641
1642 --
1643
1644 i line VAR produces "Line number not known for symbol ``var''.".  I
1645 thought we were stashing that info now!
1646
1647 --
1648
1649 We should be able to write to random files at hex offsets like adb.
1650
1651 --
1652
1653 [elena - delete this]
1654
1655 Handle add_file with separate text, data, and bss addresses.  Maybe
1656 handle separate addresses for each segment in the object file?
1657
1658 --
1659
1660 [Jimb/Elena delete this one]
1661
1662 Handle free_named_symtab to cope with multiply-loaded object files
1663 in a dynamic linking environment.  Should remember the last copy loaded,
1664 but not get too snowed if it finds references to the older copy.
1665
1666 --
1667
1668 [elena delete this also]
1669
1670 Remove all references to:
1671         text_offset
1672         data_offset
1673         text_data_start
1674         text_end
1675         exec_data_offset
1676         ...
1677 now that we have BFD.  All remaining are in machine dependent files.
1678
1679 --
1680
1681 Re-organize help categories into things that tend to fit on a screen
1682 and hang together.
1683
1684 --
1685
1686 Add in commands like ADB's for searching for patterns, etc.  We should
1687 be able to examine and patch raw unsymboled binaries as well in gdb as
1688 we can in adb.  (E.g. increase the timeout in /bin/login without source).
1689
1690 [actually, add ADB interface :-]
1691
1692 --
1693
1694 When doing "step" or "next", if a few lines of source are skipped between
1695 the previous line and the current one, print those lines, not just the
1696 last line of a multiline statement.
1697
1698 --
1699
1700 Handling of "&" address-of operator needs some serious overhaul
1701 for ANSI C and consistency on arrays and functions.
1702   For "float point[15];":
1703 ptype &point[4]   ==> Attempt to take address of non-lvalue.
1704   For "char *malloc();":
1705 ptype malloc      ==> "char *()";  should be same as
1706 ptype &malloc     ==> "char *(*)()"
1707 call printf ("%x\n", malloc)   ==> weird value, should be same as
1708 call printf ("%x\n", &malloc)  ==> correct value
1709
1710 --
1711
1712 Fix dbxread.c symbol reading in the presence of interrupts.  It
1713 currently leaves a cleanup to blow away the entire symbol table when a
1714 QUIT occurs.  (What's wrong with that? -kingdon, 28 Oct 1993).
1715
1716 [I suspect that the grype was that, on a slow system, you might want
1717 to cntrl-c and get just half the symbols and then load the rest later
1718 - scary to be honest]
1719
1720 --
1721
1722 Mipsread.c reads include files depth-first, because the dependencies
1723 in the psymtabs are way too inclusive (it seems to me).  Figure out what
1724 really depends on what, to avoid recursing 20 or 30 times while reading
1725 real symtabs.
1726
1727 --
1728
1729 value_add() should be subtracting the lower bound of arrays, if known,
1730 and possibly checking against the upper bound for error reporting.
1731
1732 --
1733
1734 When listing source lines, check for a preceding \n, to verify that
1735 the file hasn't changed out from under us.
1736
1737 [fixed by some other means I think.  That hack wouldn't actually work
1738 reliably - the file might move such that another \n appears. ]
1739
1740 --
1741
1742 Get all the remote systems (where the protocol allows it) to be able to
1743 stop the remote system when the GDB user types ^C (like remote.c
1744 does).  For ebmon, use ^Ak.
1745
1746 --
1747
1748 Possible feature: A version of the "disassemble" command which shows
1749 both source and assembly code ("set symbol-filename on" is a partial
1750 solution).
1751
1752 [has this been done?  It was certainly done for MI and GDBtk]
1753
1754 --
1755
1756 investigate "x/s 0" (right now stops early) (I think maybe GDB is
1757 using a 0 address for bad purposes internally).
1758
1759 --
1760
1761 Make "info path" and path_command work again (but independent of the
1762 environment either of gdb or that we'll pass to the inferior).
1763
1764 --
1765
1766 Make GDB understand the GCC feature for putting octal constants in
1767 enums.  Make it so overflow on an enum constant does not error_type
1768 the whole type.  Allow arbitrarily large enums with type attributes.
1769 Put all this stuff in the testsuite.
1770
1771 --
1772
1773 Make TYPE_CODE_ERROR with a non-zero TYPE_LENGTH more useful (print
1774 the value in hex; process type attributes).  Add this to the
1775 testsuite.  This way future compilers can add new types and old
1776 versions of GDB can do something halfway reasonable.
1777
1778 --
1779
1780 Fix mdebugread.c:parse_type to do fundamental types right (see
1781 rs6000_builtin_type in stabsread.c for what "right" is--the point is
1782 that the debug format fixes the sizes of these things and it shouldn't
1783 depend on stuff like TARGET_PTR_BIT and so on.  For mdebug, there seem
1784 to be separate bt* codes for 64 bit and 32 bit things, and GDB should
1785 be aware of that).  Also use a switch statement for clarity and speed.
1786
1787 --
1788
1789 Investigate adding symbols in target_load--some targets do, some
1790 don't.
1791
1792 --
1793
1794 Put dirname in psymtabs and change lookup*symtab to use dirname (so
1795 /foo/bar.c works whether compiled by cc /foo/bar.c, or cd /foo; cc
1796 bar.c).
1797
1798 --
1799
1800 Merge xcoffread.c and coffread.c.  Use breakpoint_re_set instead of
1801 fixup_breakpoints.
1802
1803 --
1804
1805 Make a watchpoint which contains a function call an error (it is
1806 broken now, making it work is probably not worth the effort).
1807
1808 --
1809
1810 New test case based on weird.exp but in which type numbers are not
1811 renumbered (thus multiply defining a type).  This currently causes an
1812 infinite loop on "p v_comb".
1813
1814 --
1815
1816 [Hey! Hint Hint Delete Delete!!!]
1817
1818 Fix 386 floating point so that floating point registers are real
1819 registers (but code can deal at run-time if they are missing, like
1820 mips and 68k).  This would clean up "info float" and related stuff.
1821
1822 --
1823
1824 gcc -g -c enummask.c then gdb enummask.o, then "p v".  GDB complains
1825 about not being able to access memory location 0.
1826
1827 -------------------- enummask.c
1828 enum mask
1829 {
1830   ANIMAL = 0,
1831   VEGETABLE = 1,
1832   MINERAL = 2,
1833   BASIC_CATEGORY = 3,
1834
1835   WHITE = 0,
1836   BLUE = 4,
1837   GREEN = 8,
1838   BLACK = 0xc,
1839   COLOR = 0xc,
1840
1841   ALIVE = 0x10,
1842
1843   LARGE = 0x20
1844 } v;
1845
1846 --
1847
1848 If try to modify value in file with "set write off" should give
1849 appropriate error not "cannot access memory at address 0x65e0".
1850
1851 --
1852
1853 Allow core file without exec file on RS/6000.
1854
1855 --
1856
1857 Make sure "shell" with no arguments works right on DOS.
1858
1859 --
1860
1861 Make gdb.ini (as well as .gdbinit) be checked on all platforms, so
1862 the same directory can be NFS-mounted on unix or DOS, and work the
1863 same way.
1864
1865 --
1866
1867 [Is this another delete???]
1868
1869 Get SECT_OFF_TEXT stuff out of objfile_relocate (might be needed to
1870 get RS/6000 to work right, might not be immediately relevant).
1871
1872 --
1873
1874 Work out some kind of way to allow running the inferior to be done as
1875 a sub-execution of, eg. breakpoint command lists.  Currently running
1876 the inferior interupts any command list execution.  This would require
1877 some rewriting of wait_for_inferior & friends, and hence should
1878 probably be done in concert with the above.
1879
1880 --
1881
1882 Add function arguments to gdb user defined functions.
1883
1884 --
1885
1886 Add convenience variables that refer to exec file, symbol file,
1887 selected frame source file, selected frame function, selected frame
1888 line number, etc.
1889
1890 --
1891
1892 Modify the handling of symbols grouped through BINCL/EINCL stabs to
1893 allocate a partial symtab for each BINCL/EINCL grouping.  This will
1894 seriously decrease the size of inter-psymtab dependencies and hence
1895 lessen the amount that needs to be read in when a new source file is
1896 accessed.
1897
1898 --
1899
1900 Add a command for searching memory, a la adb.  It specifies size,
1901 mask, value, start address.  ADB searches until it finds it or hits
1902 an error (or is interrupted).
1903
1904 --
1905
1906 Remove the range and type checking code and documentation, if not
1907 going to implement.
1908
1909 # Local Variables:
1910 # mode: text
1911 # End: