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 --
  14
  15 Wow, three bug reports for the same problem in one day!  We should
  16 probably make fixing this a real priority :-).
  17
  18 Anyway, thanks for reporting.
  19
  20 The following patch will fix the problems with setting breakpoints in
  21 dynamically loaded objects:
  22
  23    http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00230.html
  24
  25 This patch isn't checked in yet (ping Michael/JimB), but I hope this
  26 will be in the next GDB release.
  27
  28 There should really be a test in the testsuite for this problem, since
  29 it keeps coming up :-(.  Any volunteers?
  30
  31 Mark
  32
  33 --
  34
  35 x86 linux GDB and SIGALRM (???)
  36 http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00803.html
  37
  38 This problem has been fixed, but a regression test still needs to be
  39 added to the testsuite:
  40 http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00309.html
  41
  42 Mark
  43
  44 [The test has been submitted for approval - cagney]
  45
  46 --
  47
  48                 GDB 5.1 - New features
  49                 ======================
  50
  51 The following new features should be included in 5.1.
  52
  53 --
  54
  55 Pascal (Pierre Muller, David Taylor)
  56
  57 Pierre Muller has contributed patches for adding Pascal Language
  58 support to GDB.
  59
  60 2 pascal language patches inserted in database
  61 http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00521.html
  62
  63 Indent -gnu ?
  64 http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00496.html
  65
  66 [I think this has been merged, need to confirm - cagney]
  67
  68 --
  69
  70 Java (Anthony Green, David Taylor)
  71
  72 Anthony Green has a number of Java patches that did not make it into
  73 the 5.0 release.  The first two are in cvs now, but the third needs
  74 some fixing up before it can go in.
  75
  76 Patch: java tests
  77 http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00512.html
  78
  79 Patch: java booleans
  80 http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00515.html
  81
  82 Patch: handle N_MAIN stab
  83 http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00527.html
  84
  85 -- 2001-03-08
  86
  87 Add CRIS target.
  88
  89 A predicate to this is the multi-arching of SOFTWARE_SINGLE_STEP().  A
  90 patch has been submitted.
  91
  92 --
  93
  94                 GDB 5.1 - Cleanups
  95                 ==================
  96
  97 The following code cleanups will hopefully be applied to GDB 5.1.
  98
  99 --
 100
 101                 GDB 5.1 - Known Problems
 102                 ========================
 103
 104 --
 105
 106 z8k
 107
 108 The z8k has suffered bit rot and is known to not build.  The problem
 109 was occuring in the opcodes directory.
 110
 111 --
 112
 113 m88k
 114
 115 The m88k has suffered bit rot and is known to not build.
 116
 117 --
 118
 119 The BFD directory requires bug-fixed AUTOMAKE et.al.
 120
 121 AUTOMAKE 1.4 incorrectly set the TEXINPUTS environment variable.  It
 122 contained the full path to texinfo.tex when it should have only
 123 contained the directory.  The bug has been fixed in the current
 124 AUTOMAKE sources.  Automake snapshots can be found in:
 125         ftp://sourceware.cygnus.com/pub/gdb/snapshots
 126 and     ftp://sourceware.cygnus.com/pub/binutils
 127
 128 --
 129
 130 Solaris 8 x86 CURSES_H problem
 131 http://sources.redhat.com/ml/gdb/2000-07/msg00038.html
 132
 133 The original problem was worked around with:
 134
 135     2000-06-06  Michael Snyder  <msnyder@cygnus.com>
 136
 137         * configure.in: Enable autoconf to find curses.h on Solaris 2.8.
 138         * configure: Regenerate.
 139
 140 When building both GDB and SID using the same source tree the problem
 141 will still occure. sid/component/configure.in mis-configures
 142 <curses.h> and leaves wrong information in the config cache.
 143
 144 --
 145
 146                 GDB 5.2 - Fixes
 147                 ===============
 148
 149 --
 150
 151                 GDB 5.2 - New features
 152                 ======================
 153
 154 --
 155
 156 GCC 3.0 ABI support (but hopefully sooner...).
 157
 158 --
 159
 160 Objective C/C++ support (but hopefully sooner...).
 161
 162 --
 163
 164 Import of readline 4.2
 165
 166 --
 167
 168                 GDB 5.2 - Cleanups
 169                 ==================
 170
 171 The following cleanups have been identified as part of GDB 5.2.
 172
 173 --
 174
 175 Remove old code that does not use ui_out functions and all the related
 176 "ifdef"s.  This also allows the elimination of -DUI_OUT from
 177 Makefile.in and configure.in.
 178
 179 --
 180
 181 Compiler warnings.
 182
 183 Eliminate warnings for all targets on at least one host for one of the
 184 -W flags.  Flags up for debate include: -Wswitch -Wcomment -trigraphs
 185 -Wtrigraphs -Wunused-function -Wunused-label -Wunused-variable
 186 -Wunused-value -Wchar-subscripts -Wtraditional -Wshadow -Wcast-qual
 187 -Wcast-align -Wwrite-strings -Wconversion -Wstrict-prototypes
 188 -Wmissing-prototypes -Wmissing-declarations -Wredundant-decls
 189 -Woverloaded-virtual -Winline
 190
 191 --
 192
 193 Deprecate, if not delete, the following:
 194
 195         register[]
 196         register_valid[]
 197         register_buffer()
 198         REGISTER_BYTE()
 199                 Replaced by, on the target side
 200                   supply_register()
 201                 and on core-gdb side:
 202                   {read,write}_register_gen()
 203                 Remote.c will need to use something
 204                 other than REGISTER_BYTE() and
 205                 REGISTER_RAW_SIZE() when unpacking
 206                 [gG] packets.
 207
 208         STORE_PSEUDO_REGISTER
 209         FETCH_PSEUDO_REGISTER
 210                 Now handed by the methods
 211                   gdbarch_{read,write}_register()
 212                 which sits between core GDB and
 213                 the register cache.
 214
 215         REGISTER_CONVERTIBLE
 216         REGISTER_CONVERT_TO_RAW
 217         REGISTER_CONVERT_TO_VIRTUAL
 218                 I think these three are redundant.
 219                 gdbarch_register_{read,write} can
 220                 do any conversion it likes.
 221
 222         REGISTER_VIRTUAL_SIZE
 223         MAX_REGISTER_VIRTUAL_SIZE
 224         REGISTER_VIRTUAL_TYPE
 225                 I think these can be replaced by
 226                 the pair:
 227                   FRAME_REGISTER_TYPE(frame, regnum)
 228                   REGISTER_TYPE(regnum)
 229
 230         DO_REGISTERS_INFO
 231                 Replace with
 232                  FRAME_REGISTER_INFO (frame, ...)
 233
 234         REGISTER_SIM_REGNO()
 235                 If nothing else rename this so that
 236                 how it relates to rawreg and the
 237                 regnum is clear.
 238
 239         REGISTER_BYTES
 240                 The size of the cache can be computed
 241                 on the fly.
 242
 243         IS_TRAPPED_INTERNALVAR
 244                 The pseudo registers should eventually make
 245                 this redundant.
 246
 247 --
 248
 249 Obsolete the targets:
 250
 251 arm*-wince-pe
 252 mips*-*-pe
 253 sh*-*-pe
 254
 255 --
 256
 257 Obsolete the protocols:
 258
 259 RDB?
 260
 261 ``As of version 5.3, WindRiver has removed the RDB server (RDB
 262 protocol support is built into gdb).''  -- Till.
 263
 264 --
 265
 266 Restructure gdb directory tree so that it avoids any 8.3 and 14
 267 filename problems.
 268
 269 --
 270
 271 Convert GDB build process to AUTOMAKE.
 272
 273 See also sub-directory configure below.
 274
 275 The current convention is (kind of) to use $(<header>_h) in all
 276 dependency lists.  It isn't done in a consistent way.
 277
 278 --
 279
 280                 GDB 5.2 - Known Problems
 281                 ========================
 282
 283 --
 284
 285                 Code Cleanups: General
 286                 ======================
 287
 288 The following are more general cleanups and fixes.  They are not tied
 289 to any specific release.
 290
 291
 292                         New Features and Fixes
 293                         ======================
 294
 295 These are harder than cleanups but easier than work involving
 296 fundamental architectural change.
 297
 298 --
 299
 300                         Language Support
 301                         ================
 302
 303 New languages come onto the scene all the time.
 304
 305 --
 306
 307 Re: Various C++ things
 308
 309 value_headof/value_from_vtable_info are worthless, and should be
 310 removed.  The one place in printcmd.c that uses it should use the RTTI
 311 functions.
 312
 313 RTTI for g++ should be using the typeinfo functions rather than the
 314 vtables.  The typeinfo functions are always at offset 4 from the
 315 beginning of the vtable, and are always right. The vtables will have
 316 weird names like E::VB sometimes.  The typeinfo function will always
 317 be "E type_info function", or somesuch.
 318
 319 value_virtual_fn_field needs to be fixed so there are no failures for
 320 virtual functions for C++ using g++.
 321
 322 Testsuite cases are the major priority right now for C++ support,
 323 since i have to make a lot of changes that could potentially break
 324 each other.
 325
 326 --
 327
 328
 329                         Symbol Support
 330                         ==============
 331
 332 --
 333
 334 Investiagate ways of reducing memory.
 335
 336 --
 337
 338 Investigate ways of improving load time.
 339
 340 --
 341
 342                         Testsuite Support
 343                         =================
 344
 345 There are never to many testcases.
 346
 347 --
 348
 349 Better thread testsuite.
 350
 351 --
 352
 353 Better C++ testsuite.
 354
 355 --
 356
 357                 Architectural Changes: General
 358                 ==============================
 359
 360 These are harder than simple cleanups / fixes and, consequently
 361 involve more work.  Typically an Architectural Change will be broken
 362 down into a more digestible set of cleanups and fixes.
 363
 364 --
 365
 366                 Architectural Change: Multi-arch et al.
 367                 =======================================
 368
 369 The long term objective is to remove all assumptions that there is a
 370 single target with a single address space with a single instruction
 371 set architecture and single application binary interface.
 372
 373 This is an ongoing effort.  The first milestone is to enable
 374 ``multi-arch'' where by all architectural decisions are made at
 375 runtime.
 376
 377 It should be noted that ``gdbarch'' is really ``gdbabi'' and
 378 ``gdbisa''.  Once things are multi-arched breaking that down correctly
 379 will become much easier.
 380
 381 --
 382
 383         Architectural Change: MI, LIBGDB and scripting languages
 384         ========================================================
 385
 386 See also architectural changes related to the event loop.  LIBGDB
 387 can't be finished until there is a generic event loop being used by
 388 all targets.
 389
 390 The long term objective is it to be possible to integrate GDB into
 391 scripting languages.
 392
 393 --
 394
 395                 Architectural Change: Async
 396                 ===========================
 397
 398 While GDB uses an event loop when prompting the user for input.  That
 399 event loop is not exploited by targets when they allow the target
 400 program to continue.  Typically targets still block in (target_wait())
 401 until the program again halts.
 402
 403 The closest a target comes to supporting full asynchronous mode are
 404 the remote targets ``async'' and ``extended-async''.
 405
 406 --
 407
 408 # Local Variables:
 409 # mode: text
 410 # End: