1 2 Suggested projects for aspiring or current GDB hackers 3 ====================================================== 4 5 (You should probably chat with kingdon@ai.mit.edu to make sure that 6 no one else is doing the project you chose). 7 8Add watchpoints (break if a memory location changes). This would 9usually have to involve constant single stepping, but occasionally 10there is operating system support which gdb should be able to cleanly 11use (e.g. on the 80386, there are 4 debug registers. By ptracing an 12address into them, you can get a trap on writes or on reads and 13writes). 14 15Rewrite proceed, wait_for_inferior, and normal_stop to clean them up. 16Suggestions: 17 18 1) Make each test in wait_for_inferior a seperate subroutine 19 call. 20 2) Combine wait_for_inferior and normal_stop to clean up 21 communication via global variables. 22 3) See if you can find some way to clean up the global 23 variables that are used; possibly group them by data flow 24 and information content? 25 26Work out some kind of way to allow running the inferior to be done as 27a sub-execution of, eg. breakpoint command lists. Currently running 28the inferior interupts any command list execution. This would require 29some rewriting of wait_for_inferior & friends, and hence should 30probably be done in concert with the above. 31 32Add function arguments to gdb user defined functions. 33 34Add convenience variables that refer to exec file, symbol file, 35selected frame source file, selected frame function, selected frame 36line number, etc. 37 38Add a "suspend" subcommand of the "continue" command to suspend gdb 39while continuing execution of the subprocess. Useful when you are 40debugging servers and you want to dodge out and initiate a connection 41to a server running under gdb. 42 43Make "handle" understand symbolic signal names. 44 45Work out and implement a reasonably general mechanism for multi-threaded 46processies. There are parts of one implemented in convex-dep.c, if 47you want an example. 48 49A standalone version of gdb on the i386 exists. Anyone who wants to 50do some serious working cleaning it up and making it a general 51standalone gdb should contact pace@wheaties.ai.mit.edu. 52 53Add stab information to allow reasonable debugging of inline functions 54(possibly they should show up on a stack backtrace? With a note 55indicating that they weren't "real"?). 56 57Implement support for specifying arbitrary locations of stack frames 58(in practice, this usually requires specification of both the top and 59bottom of the stack frame (fp and sp), since you *must* retrieve the 60pc that was saved in the innermost frame). 61 62Modify the naked "until" command to step until past the current source 63line, rather than past the current pc value. This is tricky simply 64because the low level routines have no way of specifying a multi-line 65step range, and there is no way of saying "don't print stuff when we 66stop" from above (otherwise could just call step many times). 67 68Modify the handling of symbols grouped through BINCL/EINCL stabs to 69allocate a partial symtab for each BINCL/EINCL grouping. This will 70seriously decrease the size of inter-psymtab dependencies and hence 71lessen the amount that needs to be read in when a new source file is 72accessed. 73 74Work out some method of saving breakpoints across the reloading of an 75executable. Probably this should be by saving the commands by which 76the breakpoints were set and re-executing them (as text locations may 77change). 78 79Do an "x/i $pc" after each stepi or nexti. 80 81Modify all of the disassemblers to use printf_filtered to get correct 82more filtering. 83 84Modify gdb to work correctly with Pascal. 85 86Rewrite macros that handle frame chaining and frameless functions. 87They should be able to tell the difference between start, main, and a 88frameless function called from main. 89 90Work out what information would need to be included in an executable 91by the compiler to allow gdb to debug functions which do not have a 92frame pointer. Modify gdb and gcc to do this. 93 94When `attached' to a program (via either OS support or remote 95debugging), gdb should arrange to catch signals which the terminal 96might send, as it is unlikely that the program will be able to notice 97them. SIGINT and SIGTSTP are obvious examples. 98 99Enhance the gdb manual with extra examples where needed. 100 101Arrange for list_command not to use decode_line_1 and thus not require 102symbols to be read in simply to read a source file. 103 104Problem in xgdb; the readline library needs the terminal in CBREAK 105mode for command line editing, but this makes it difficult to dispatch 106on button presses. Possible solution: use a define to replace getc in 107readline.c with a routine that does button dispatches. You should 108probably see XGDB-README before you fiddle with XGDB. Also, someone 109is implementing a new xgdb; it may not be worth while fiddling with 110the old one. 111 112# Local Variables: 113# mode: text 114# End: 115