This is the mail archive of the gdb-patches@sources.redhat.com mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: [RFA] deleting breakpoints inside of 'commands' [Repost]



I thought of a problem with this patch: It could change a command list
(adding additional commands) if a breakpoint is hit while processing
another breakpoint.  Bugger.  Maybe I could add a field to struct
command_line for managing orphans?


On Thu, 20 Sep 2001, Don Howard wrote:

>
>
> Here is a patch that I hope will satisfy all sides.  The patch below
> changes free_command_lines() to check if we are currently
> 'executing_breakpoint_commands'.  If so, the command list is appended to
> a list of 'orphaned_command_lines'.  When the execution of the command
> list completes, the orphaned_command_lines are then deleted.
>
> Should I make the global variables static and add accessor/mutator
> functions?  What is the policy on globals?
>
> Comments?
>
>
>
> 2001-09-20  Don Howard  <dhoward@redhat.com>
>
> 	* cli/cli-script.c (free_command_lines): Avoid deleting
> 	command_line lists while executing that list.
> 	* breakpoint.c: (executing_breakpoint_commands): Make this
> 	global visible to other compilation units.
> 	(orphaned_breakpoint_commands) New global.
> 	(bpstat_do_actions): Free orphaned command_line structures once
> 	done executing them.
>
>
> Index: gdb/breakpoint.c
> ===================================================================
> RCS file: /cvs/src/src/gdb/breakpoint.c,v
> retrieving revision 1.53
> diff -p -u -w -r1.53 breakpoint.c
> --- breakpoint.c	2001/09/18 05:00:48	1.53
> +++ breakpoint.c	2001/09/20 22:18:17
> @@ -221,8 +221,11 @@ extern int addressprint;	/* Print machin
>  static int internal_breakpoint_number = -1;
>
>  /* Are we executing breakpoint commands?  */
> -static int executing_breakpoint_commands;
> +int executing_breakpoint_commands;
>
> +/* List of breakpoint commands to be cleaned up after we are done executing them. */
> +struct command_line * orphaned_breakpoint_commands;
> +
>  /* Walk the following statement or block through all breakpoints.
>     ALL_BREAKPOINTS_SAFE does so even if the statment deletes the current
>     breakpoint.  */
> @@ -1845,6 +1848,7 @@ top:
>      }
>
>    executing_breakpoint_commands = 0;
> +  free_command_lines (&orphaned_breakpoint_commands);
>    discard_cleanups (old_chain);
>  }
>
> Index: gdb/cli/cli-script.c
> ===================================================================
> RCS file: /cvs/src/src/gdb/cli/cli-script.c,v
> retrieving revision 1.7
> diff -p -u -w -r1.7 cli-script.c
> --- cli-script.c	2001/06/17 15:16:12	1.7
> +++ cli-script.c	2001/09/20 22:18:17
> @@ -1014,7 +1014,25 @@ free_command_lines (struct command_line
>    register struct command_line *next;
>    struct command_line **blist;
>    int i;
> +  extern int executing_breakpoint_commands;
> +  extern struct command_line * orphaned_breakpoint_commands;
>
> +  /* Avoid deleting command_line lists while executing that list. */
> +  if (executing_breakpoint_commands)
> +    {
> +      struct command_line **b = NULL;
> +
> +      /* Find the end of the orphaned_breakpoint_commands list */
> +      for (b = &orphaned_breakpoint_commands; *b; b = &((*b)->next))
> +	/* Nothing */
> +	;
> +
> +      /* Append the passed-in list */
> +      *b = *lptr;
> +      *lptr = NULL;
> +      return;
> +    }
> +
>    while (l)
>      {
>        if (l->body_count > 0)
>
>
>

-- 
-Don
dhoward@redhat.com
gdb engineering


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]