Re: cli/1815: GDB can crash with overridden "down" command

[Skip Montanaro <skip at pobox dot com>]

The following reply was made to PR cli/1815; it has been noted by GNATS.

From: Skip Montanaro <skip@pobox.com>
To: gdb-gnats@sources.redhat.com
Cc:  
Subject: Re: cli/1815: GDB can crash with overridden "down" command
Date: Sun, 21 Nov 2004 21:43:10 -0600

     This version of GDB has substantial changes.  You're going to have to
     report this bug to Apple instead.
 
 Daniel,
 
 Thanks for the response.  I submitted a bug to Apple as you suggested and
 referred to this ticket.  I also gave it a whirl with gdb 6.2.1 on
 Solaris/Intel.  It worked fine if I executed my redefined "up" or "down"
 commands.  However, if I executed "do" (which is normally allowed as an
 abbreviation for "down", right?), it segfaulted immediately:
 
     (gdb) down
     /home/titan/skipm/tmp/tstlineno.py (11): h
     (gdb) 
     #8  0x080abccb in call_function (pp_stack=0x804694c, oparg=1)
         at ../Python/ceval.c:3458
     (gdb) 
     #7  0x080abee1 in fast_function (func=0x81c74fc, pp_stack=0x804694c, n=1, 
         na=1, nk=0) at ../Python/ceval.c:3518
     (gdb) do
 
     Inferior GDB segmentation violation (core dumped)
 
 Unfortunately, I don't have a gdb compiled with debug symbols at the
 moment.  Here's the start of that traceback:
 
     (gdb) bt 10
     #0  0x656d6172 in ?? ()
     #1  0x080bb7dc in cmd_func ()
     #2  0x0808986c in execute_command ()
     #3  0x081217a6 in command_handler ()
     #4  0x08121fc6 in command_line_handler ()
     #5  0x0820a10b in rl_callback_read_char ()
     #6  0x08120ee1 in rl_callback_read_char_wrapper ()
     #7  0x0812166f in stdin_event_handler ()
     #8  0x081205d8 in handle_file_event ()
     #9  0x0811fe72 in process_event ()
 
 When I get a chance I will try to build gdb with debug symbols so I can
 generate a more useful traceback.
 
 Skip