gdb/TODO

   1
   2         gdb bug list
   3         John Gilmore, gnu@cygnus.com
   4
   5 This bug list is probably not up to date or accurate, but it reflects
   6 some known bugs in gdb, if you are into bug-hunting.
   7
   8
   9 It should be possible to use symbols from shared libraries before we know
  10 exactly where the libraries will be loaded.  E.g. "b perror" before running
  11 the program.  This could maybe be done as an extension of the "breakpoint
  12 re-evaluation" after new symbols are loaded.
  13
  14 Make single_step() insert and remove breakpoints in one operation.
  15
  16 Speed up single stepping by avoiding extraneous ptrace calls.
  17
  18 Speed up single stepping by not inserting and removing breakpoints
  19 each time the inferior starts and stops.
  20
  21 Breakpoints should not be inserted and deleted all the time.  Only the
  22 one(s) there should be removed when we have to step over one.  Support
  23 breakpoints that don't have to be removed to step over them.
  24
  25 Speed up watchpoints by using debug registers, page table diddling (on
  26 SunOS4, can call mprotect() in the inferior; on other machines can do
  27 something simpler), etc.
  28
  29 Update gdbint.texinfo to include doc on the directory structure and
  30 the various tricks of building gdb.
  31
  32 Do a tutorial in gdb.texinfo on how to do simple things in gdb.
  33 E.g. how to set a breakpoint that just prints something and continues.
  34 How to break on aborts.  Etc.
  35
  36 Provide "voodoo" debugging of core files.  This creates a zombie
  37 process as a child of the debugger, and loads it up with the data,
  38 stack, and regs of the core file.  This allows you to call functions
  39 in the executable, to manipulate the data in the core file.
  40
  41 GDB reopens the source file on every line, as you "next" through it.
  42
  43 Referencing the vtbl member of a struct doesn't work.  It prints OK
  44 if you print the struct, but it gets 0 if you try to deref it.
  45
  46 Persistent command history: A feature where you could save off a list
  47 of the commands you did, so you can edit it into something that will bring
  48 the target to the same place every time you source it.
  49 This would also be useful for automated fast watchpointing; if you go
  50 past the place where it watchpoints, you just start it over again and
  51 do it more carefully.
  52
  53 Deal with the SunOS 4.0 and 4.1.1 ptrace bug that loses the registers if
  54 the stack is paged out.
  55
  56 Finish the C++ exception handling stub routines.  Lint points them out
  57 as unused statics functions.
  58
  59 Perhaps "i source" should take an argument like that of "list".
  60
  61 See if coredep.c's fetch_core_registers can be used on more machines.
  62 E.g. MIPS (mips-xdep.c).
  63
  64 unpack_double() does not handle IEEE float on the target unless the host
  65 is also IEEE.  Death on a vax.
  66
  67 Set up interface between GDB and INFO so that you can hop into interactive
  68 INFO and back out again.  When running under Emacs, should use Emacs
  69 info, else fork the info program.  Installation of GDB should install
  70 its texinfo files into the info tree automagically, including the readline
  71 texinfo files.
  72
  73 "help address" ought to find the "help set print address" entry.
  74
  75 Remove the VTBL internal guts from printouts of C++ structs, unless
  76 vtblprint is set.
  77
  78 Remove "at 0xnnnn" from the "b foo" response, if `print address off' and if
  79 it matches the source line indicated.
  80
  81 The prompt at end of screen should accept space as well as CR.
  82
  83 "List" should put you into a pseudo-"more" where you can hit space
  84 to get more, forever to eof.
  85
  86 Check STORE_RETURN_VALUE on all architectures.  Check near it in tm-sparc.h
  87 for other bogosities.
  88
  89 Check for storage leaks in GDB, I'm sure there are a lot!
  90
  91 vtblprint of a vtbl should demangle the names it's printing.
  92
  93 Backtrace should point out what the currently selected frame is, in its
  94 display, perhaps showing "@3 foo (bar, ...)" rather than "#3 foo (bar, ...)".
  95
  96 "i program" should work for core files, and display more info, like what
  97 actually caused it to die.
  98
  99 Hitting ^Z to an inferior doesn't work right, it takes several continues
 100 to make it actually go.
 101
 102 "x/10i" should shorten the long name, if any, on subsequent lines.
 103
 104 Check through the code for FIXME comments and fix them.  dbxread.c,
 105 blockframe.c, and plenty more.
 106
 107 "next" over a function that longjumps, never stops until next time you happen
 108 to get to that spot by accident.  E.g. "n" over execute_command which has
 109 an error.
 110
 111 Watchpoints seem not entirely reliable, though they haven't failed me recently.
 112
 113 "set zeroprint off", don't bother printing members of structs which are entirely
 114 zero.  Useful for those big structs with few useful members.
 115
 116 GDB does four ioctl's for every command, probably switching terminal modes
 117 to/from inferior or for readline or something.
 118
 119 terminal_ours versus terminal_inferior: cache state.  Switch should be a noop
 120 if the state is the same, too.
 121
 122 ptype $i6 = void??!
 123
 124 Clean up invalid_float handling so gdb doesn't coredump when it tries to
 125 access a NaN.  While this might work on SPARC, other machines are not
 126 configured right.
 127
 128 "b value_at ; commands ; continue ; end" stops EVERY OTHER TIME!
 129 Then once you enter a command, it does the command, runs two more
 130 times, and then stops again!  Bizarre...  (This behaviour has been
 131 modified, but it is not yet 100% predictable when e.g. the commands
 132 call functions in the child, and while there, the child is interrupted
 133 with a signal, or hits a breakpoint.)
 134
 135 help completion, help history should work.
 136
 137 Check that we can handle stack trace through varargs AND alloca in same
 138 function, on 29K.
 139
 140 wait_for_inferior loops forever if wait() gives it an error.
 141
 142 "i frame" shows wrong "arglist at" location, doesn't show where the args
 143 should be found, only their actual values.
 144
 145 There should be a way for "set" commands to validate the new setting
 146 before it takes effect.
 147
 148 The "display" command should become the "always" command, e.g.
 149         "always print XXX"
 150         "always p/xxx XXX"
 151         "always echo foo"
 152         "always call XXX"
 153         "always x/i $pc", etc.
 154
 155 A mess of floating point opcodes are missing from sparc-opcode.h.
 156 Also, a little program should test the table for bits that are
 157 overspecified or underspecified.  E.g. if the must-be-ones bits
 158 and the must-be-zeroes bits leave some fields unexamined, and the format
 159 string leaves them unprinted, then point this out.  If multiple
 160 non-alias patterns match, point this out too.  Finally, there should
 161 be a sparc-optest.s file that tries each pattern out.  This file
 162 should end up coming back the same (modulo transformation comments)
 163 if fed to "gas" then the .o is fed to gdb for disassembly.
 164
 165 Eliminate all the core_file_command's in all the xdep files.
 166 Eliminate separate declarations of registers[] everywhere.
 167
 168 "ena d" is ambiguous, why?  "ena delete" seems to think it is a command!
 169
 170 Line numbers are off in some spots.  In proceed() at 1st "oneproc = 1",
 171 it seems to run that statement, but it doesn't actually.
 172
 173 Perhaps move the tdep, xdep, and nat files, into the config
 174 subdirectories.  If not, at least straighten out their names so that
 175 they all start with the machine name.
 176
 177 inferior_status should include stop_print_frame.  It won't need to be
 178 reset in wait_for_inferior after bpstat_stop_status call, then.
 179
 180 i line VAR produces "Line number not known for symbol ``var''.".  I
 181 thought we were stashing that info now!
 182
 183 Make sure we can handle executables with no symbol info, e.g. /bin/csh.
 184
 185 We should be able to write to random files at hex offsets like adb.
 186
 187 Make "target xxx" command interruptible.
 188
 189 Handle add_file with separate text, data, and bss addresses.  Maybe
 190 handle separate addresses for each segment in the object file?
 191
 192 Handle free_named_symtab to cope with multiply-loaded object files
 193 in a dynamic linking environment.  Should remember the last copy loaded,
 194 but not get too snowed if it finds references to the older copy.
 195
 196 The original BFD core dump reading routine would itself coredump when fed
 197 a garbage file as a core file.  Does the current one?
 198
 199 Generalize and Standardize the RPC interface to a target program,
 200 improve it beyond the "ptrace" interface, and see if it can become a
 201 standard for remote debugging.
 202
 203 Remove all references to:
 204         text_offset
 205         data_offset
 206         text_data_start
 207         text_end
 208         exec_data_offset
 209         ...
 210 now that we have BFD.  All remaining are in machine dependent files.
 211
 212 When quitting with a running program, if a core file was previously
 213 examined, you get "Couldn't read float regs from core file"...if
 214 indeed it can't.  generic_mourn_inferior...
 215
 216 Check signal argument to remote proceed's and error if set.
 217
 218 Sort help and info output.
 219
 220 Re-organize help categories into things that tend to fit on a screen
 221 and hang together.
 222
 223 renote-nindy.c handles interrupts poorly; it error()s out of badly
 224 chosen places, e.g. leaving current_frame zero, which causes core dumps
 225 on the next command.
 226
 227 Add in commands like ADB's for searching for patterns, etc.  We should
 228 be able to examine and patch raw unsymboled binaries as well in gdb as
 229 we can in adb.  (E.g. increase the timeout in /bin/login without source).
 230
 231 Those xdep files that call register_addr without defining it are
 232 probably simply broken.  When reconfiguring this part of gdb, I could
 233 only make guesses about how to redo some of those files, and I
 234 probably guessed wrong, or left them "for later" when I have a
 235 machine that can attempt to build them.
 236
 237 Use the complain() mechanism for handling all the error() calls in dbxread.c,
 238 and in similar situations in coffread.c and mipsread.c.
 239
 240 When doing "step" or "next", if a few lines of source are skipped between
 241 the previous line and the current one, print those lines, not just the
 242 last line of a multiline statement.
 243
 244 When searching for C++ superclasses in value_cast in valops.c, we must
 245 not search the "fields", only the "superclasses".  There might be a
 246 struct with a field name that matches the superclass name.  This can
 247 happen when the struct was defined before the superclass (before the
 248 name became a typedef).
 249
 250 Handling of "&" address-of operator needs some serious overhaul
 251 for ANSI C and consistency on arrays and functions.
 252   For "float point[15];":
 253 ptype &point[4]   ==> Attempt to take address of non-lvalue.
 254   For "char *malloc();":
 255 ptype malloc      ==> "char *()";  should be same as
 256 ptype &malloc     ==> "char *(*)()"
 257 call printf ("%x\n", malloc)   ==> wierd value, should be same as
 258 call printf ("%x\n", &malloc)  ==> correct value
 259
 260 Fix dbxread.c symbol reading in the presence of interrupts.  It currently
 261 leaves a cleanup to blow away the entire symbol table when a QUIT occurs.
 262
 263 Mipsread.c reads include files depth-first, because the dependencies
 264 in the psymtabs are way too inclusive (it seems to me).  Figure out what
 265 really depends on what, to avoid recursing 20 or 30 times while reading
 266 real symtabs.
 267
 268 value_add() should be subtracting the lower bound of arrays, if known,
 269 and possibly checking against the upper bound for error reporting.
 270
 271 mipsread.c symbol table allocation and deallocation should be checked.
 272 My suspicion is that it's full of memory leaks.
 273
 274 SunOS should have a target_lookup_symbol() for common'd things allocated
 275 by the shared library linker ld.so.
 276
 277 When listing source lines, check for a preceding \n, to verify that
 278 the file hasn't changed out from under us.
 279
 280 When listing source lines, eat leading whitespace corresponding to the
 281 line-number prefix we print.  This avoids long lines wrapping.
 282
 283 mipsread.c needs to check for old symtabs and psymtabs for the same
 284 files, the way it happens for dbxread.c and coffread.c, for VxWorks
 285 incremental symbol table reloading.
 286
 287 When attached to a non-child process, ^C or other signals are not
 288 propagated to the child.  Do this in the GDB signal handler, using
 289 target_kill().  AMD version: ^C should do ^Ak to stop ebmon.
 290