Re: [RFC/RFA] New 'to' command

Re: [RFC/RFA] New 'to' command

[Michael Elizabeth Chastain]

I'm pretty much in Dan J's camp, up to the point where he doesn't
think of checking in 'help'.  Although 'to' without an argument might
possibly provide some useful text.

Michael C

Re: [RFC/RFA] New 'to' command

[Elena Zannoni]

Michael Elizabeth Chastain writes:
 > I'm pretty much in Dan J's camp, up to the point where he doesn't
 > think of checking in 'help'.  Although 'to' without an argument might
 > possibly provide some useful text.
 > 
 > Michael C

Aehm..., right not it doesn't....  fixed it:

(gdb) b main
Breakpoint 1 at 0x804852e: file /home/ezannoni/sources/src/gdb/testsuite/gdb.base/break.c, line 75.
(gdb) r
Starting program: /home/ezannoni/sources/native/gdb/testsuite/gdb.base/break 

Breakpoint 1, main (argc=1, argv=0xbffff0f4, envp=0xbffff0fc)
    at /home/ezannoni/sources/src/gdb/testsuite/gdb.base/break.c:75
75	    if (argc == 12345) {  /* an unlikely value < 2^16, in case uninited */
(gdb) to
Argument required (a location).
(gdb)


Updated diff attached:

2003-01-12  Elena Zannoni  <ezannoni@redhat.com>

	* breakpoint.c (until_break_command): Add new argument.  Use it to
	decide whether to stop only at the current frame or not.
	* breakpoint.h (until_break_command): Update prototype.
	* infcmd.c (until_command): Add new argument to until_break_command
	call.
	(to_command): New function.
	(_initialize_infcmd): Update help string for 'until' command.
	Add new 'to' command.


Index: breakpoint.c
===================================================================
RCS file: /cvs/uberbaum/gdb/breakpoint.c,v
retrieving revision 1.105
diff -u -p -r1.105 breakpoint.c
--- breakpoint.c	4 Jan 2003 23:07:24 -0000	1.105
+++ breakpoint.c	12 Jan 2003 19:08:07 -0000
@@ -5576,7 +5576,7 @@ until_break_command_continuation (struct
 
 /* ARGSUSED */
 void
-until_break_command (char *arg, int from_tty)
+until_break_command (char *arg, int from_tty, int anywhere)
 {
   struct symtabs_and_lines sals;
   struct symtab_and_line sal;
@@ -5609,9 +5609,16 @@ until_break_command (char *arg, int from
 
   resolve_sal_pc (&sal);
 
-  breakpoint = 
-    set_momentary_breakpoint (sal,get_frame_id (deprecated_selected_frame),
-			      bp_until);
+  if (anywhere)
+    /* If the user told us to continue until a specified location,
+       we don't specify a frame at which we need to stop.  */
+    breakpoint = set_momentary_breakpoint (sal, null_frame_id, bp_until);
+  else
+    /* Otherwise, specify the current frame, because we want to stop only
+       at the very same frame.  */
+    breakpoint = set_momentary_breakpoint (sal,
+					   get_frame_id (deprecated_selected_frame),
+					   bp_until);
 
   if (!event_loop_p || !target_can_async_p ())
     old_chain = make_cleanup_delete_breakpoint (breakpoint);
@@ -5639,8 +5646,8 @@ until_break_command (char *arg, int from
       add_continuation (until_break_command_continuation, arg1);
     }
 
-  /* Keep within the current frame */
-
+  /* Keep within the current frame, or in frames called by the current
+     one.  */
   if (prev_frame)
     {
       sal = find_pc_line (get_frame_pc (prev_frame), 0);

Index: breakpoint.h
===================================================================
RCS file: /cvs/uberbaum/gdb/breakpoint.h,v
retrieving revision 1.16
diff -u -p -r1.16 breakpoint.h
--- breakpoint.h	11 Dec 2002 22:34:47 -0000	1.16
+++ breakpoint.h	12 Jan 2003 19:08:50 -0000
@@ -534,7 +534,7 @@ extern int deprecated_frame_in_dummy (st
 
 extern int breakpoint_thread_match (CORE_ADDR, ptid_t);
 
-extern void until_break_command (char *, int);
+extern void until_break_command (char *, int, int);
 
 extern void breakpoint_re_set (void);

Index: infcmd.c
===================================================================
RCS file: /cvs/uberbaum/gdb/infcmd.c,v
retrieving revision 1.69
diff -u -p -r1.69 infcmd.c
--- infcmd.c	13 Dec 2002 16:26:02 -0000	1.69
+++ infcmd.c	12 Jan 2003 21:07:14 -0000
@@ -1140,10 +1140,41 @@ until_command (char *arg, int from_tty)
     }
 
   if (arg)
-    until_break_command (arg, from_tty);
+    until_break_command (arg, from_tty, 0);
   else
     until_next_command (from_tty);
 }
+
+static void
+to_command (char *arg, int from_tty)
+{
+  int async_exec = 0;
+
+  if (!target_has_execution)
+    error ("The program is not running.");
+
+  if (arg == NULL)
+    error_no_arg ("a location");
+
+  /* Find out whether we must run in the background. */
+  if (arg != NULL)
+    async_exec = strip_bg_char (&arg);
+
+  /* If we must run in the background, but the target can't do it,
+     error out. */
+  if (event_loop_p && async_exec && !target_can_async_p ())
+    error ("Asynchronous execution not supported on this target.");
+
+  /* If we are not asked to run in the bg, then prepare to run in the
+     foreground, synchronously. */
+  if (event_loop_p && !async_exec && target_can_async_p ())
+    {
+      /* Simulate synchronous execution */
+      async_disable_stdin ();
+    }
+
+  until_break_command (arg, from_tty, 1);
+}
 
 
 /* Print the result of a function at the end of a 'finish' command. */
@@ -2130,10 +2161,14 @@ Argument N means do this N times (or til
 
   c = add_com ("until", class_run, until_command,
 	       "Execute until the program reaches a source line greater than the current\n\
-or a specified line or address or function (same args as break command).\n\
-Execution will also stop upon exit from the current stack frame.");
+or a specified location (same args as break command) within the current frame.");
   set_cmd_completer (c, location_completer);
   add_com_alias ("u", "until", class_run, 1);
+
+  c = add_com ("to", class_run, to_command,
+	       "Continue the program up to the given location (same args as break command).\n\
+Execution will also stop upon exit from the current stack frame.");
+  set_cmd_completer (c, location_completer);
 
   c = add_com ("jump", class_run, jump_command,
 	       "Continue program being debugged at specified line or address.\n\

Re: [RFC/RFA] New 'to' command

[Michael Elizabeth Chastain]

'advance' sounds fine to me.

Michael C

Re: [RFC/RFA] New 'to' command

[Elena Zannoni]

Michael Elizabeth Chastain writes:
 > 'advance' sounds fine to me.
 > 
 > Michael C

all right!

Elena

Re: [RFC/RFA] New 'to' command

[Fernando Nasser]

Elena Zannoni wrote:> Michael Elizabeth Chastain writes:
>  > 'advance' sounds fine to me.
>  > 
>  > Michael C
> 
> all right!
> 
> Elena
> 

That was quick.  I was afraid I had started another long long thread :-)

Regards to all,
Fernando

-- 
Fernando Nasser
Red Hat - Toronto                       E-Mail:  fnasser@redhat.com
2323 Yonge Street, Suite #300
Toronto, Ontario   M4P 2C9

Re: [RFC/RFA] New 'to' command

[Elena Zannoni]

Fernando Nasser writes:
 > Elena Zannoni wrote:> Michael Elizabeth Chastain writes:
 > >  > 'advance' sounds fine to me.
 > >  > 
 > >  > Michael C
 > > 
 > > all right!
 > > 
 > > Elena
 > > 
 > 
 > That was quick.  I was afraid I had started another long long thread :-)
 > 
 > Regards to all,
 > Fernando
 > 
 > -- 
 > Fernando Nasser
 > Red Hat - Toronto                       E-Mail:  fnasser@redhat.com
 > 2323 Yonge Street, Suite #300
 > Toronto, Ontario   M4P 2C9

This is what I am checking in.

2003-01-14  Elena Zannoni  <ezannoni@redhat.com>
 
 	* breakpoint.c (until_break_command): Add new argument.  Use it to
 	decide whether to stop only at the current frame or not.
 	* breakpoint.h (until_break_command): Update prototype.
 	* infcmd.c (until_command): Add new argument to until_break_command
 	call.
 	(advance_command): New function.
 	(_initialize_infcmd): Update help string for 'until' command.
 	Add new 'advance' command.
 


Index: breakpoint.c
===================================================================
RCS file: /cvs/uberbaum/gdb/breakpoint.c,v
retrieving revision 1.105
diff -u -p -r1.105 breakpoint.c
--- breakpoint.c	4 Jan 2003 23:07:24 -0000	1.105
+++ breakpoint.c	15 Jan 2003 00:02:46 -0000
@@ -5576,7 +5576,7 @@ until_break_command_continuation (struct
 
 /* ARGSUSED */
 void
-until_break_command (char *arg, int from_tty)
+until_break_command (char *arg, int from_tty, int anywhere)
 {
   struct symtabs_and_lines sals;
   struct symtab_and_line sal;
@@ -5609,9 +5609,16 @@ until_break_command (char *arg, int from
 
   resolve_sal_pc (&sal);
 
-  breakpoint = 
-    set_momentary_breakpoint (sal,get_frame_id (deprecated_selected_frame),
-			      bp_until);
+  if (anywhere)
+    /* If the user told us to continue until a specified location,
+       we don't specify a frame at which we need to stop.  */
+    breakpoint = set_momentary_breakpoint (sal, null_frame_id, bp_until);
+  else
+    /* Otherwise, specify the current frame, because we want to stop only
+       at the very same frame.  */
+    breakpoint = set_momentary_breakpoint (sal,
+					   get_frame_id (deprecated_selected_frame),
+					   bp_until);
 
   if (!event_loop_p || !target_can_async_p ())
     old_chain = make_cleanup_delete_breakpoint (breakpoint);
@@ -5639,8 +5646,8 @@ until_break_command (char *arg, int from
       add_continuation (until_break_command_continuation, arg1);
     }
 
-  /* Keep within the current frame */
-
+  /* Keep within the current frame, or in frames called by the current
+     one.  */
   if (prev_frame)
     {
       sal = find_pc_line (get_frame_pc (prev_frame), 0);
@@ -5659,7 +5666,7 @@ until_break_command (char *arg, int from
   if (!event_loop_p || !target_can_async_p ())
     do_cleanups (old_chain);
 }
-
+
 #if 0
 /* These aren't used; I don't konw what they were for.  */
 /* Set a breakpoint at the catch clause for NAME.  */

Index: infcmd.c
===================================================================
RCS file: /cvs/uberbaum/gdb/infcmd.c,v
retrieving revision 1.69
diff -u -p -r1.69 infcmd.c
--- infcmd.c	13 Dec 2002 16:26:02 -0000	1.69
+++ infcmd.c	15 Jan 2003 00:04:10 -0000
@@ -1140,10 +1140,41 @@ until_command (char *arg, int from_tty)
     }
 
   if (arg)
-    until_break_command (arg, from_tty);
+    until_break_command (arg, from_tty, 0);
   else
     until_next_command (from_tty);
 }
+
+static void
+advance_command (char *arg, int from_tty)
+{
+  int async_exec = 0;
+
+  if (!target_has_execution)
+    error ("The program is not running.");
+
+  if (arg == NULL)
+    error_no_arg ("a location");
+
+  /* Find out whether we must run in the background.  */
+  if (arg != NULL)
+    async_exec = strip_bg_char (&arg);
+
+  /* If we must run in the background, but the target can't do it,
+     error out.  */
+  if (event_loop_p && async_exec && !target_can_async_p ())
+    error ("Asynchronous execution not supported on this target.");
+
+  /* If we are not asked to run in the bg, then prepare to run in the
+     foreground, synchronously.  */
+  if (event_loop_p && !async_exec && target_can_async_p ())
+    {
+      /* Simulate synchronous execution.  */
+      async_disable_stdin ();
+    }
+
+  until_break_command (arg, from_tty, 1);
+}
 \f
 
 /* Print the result of a function at the end of a 'finish' command. */
@@ -2130,10 +2161,14 @@ Argument N means do this N times (or til
 
   c = add_com ("until", class_run, until_command,
 	       "Execute until the program reaches a source line greater than the current\n\
-or a specified line or address or function (same args as break command).\n\
-Execution will also stop upon exit from the current stack frame.");
+or a specified location (same args as break command) within the current frame.");
   set_cmd_completer (c, location_completer);
   add_com_alias ("u", "until", class_run, 1);
+
+  c = add_com ("advance", class_run, advance_command,
+	       "Continue the program up to the given location (same form as args for break command).\n\
+Execution will also stop upon exit from the current stack frame.");
+  set_cmd_completer (c, location_completer);
 
   c = add_com ("jump", class_run, jump_command,
 	       "Continue program being debugged at specified line or address.\n\

Index: breakpoint.h
===================================================================
RCS file: /cvs/uberbaum/gdb/breakpoint.h,v
retrieving revision 1.16
diff -u -p -r1.16 breakpoint.h
--- breakpoint.h	11 Dec 2002 22:34:47 -0000	1.16
+++ breakpoint.h	15 Jan 2003 00:05:35 -0000
@@ -534,7 +534,7 @@ extern int deprecated_frame_in_dummy (st
 
 extern int breakpoint_thread_match (CORE_ADDR, ptid_t);
 
-extern void until_break_command (char *, int);
+extern void until_break_command (char *, int, int);
 
 extern void breakpoint_re_set (void);

Re: [RFC/RFA] New 'to' command

[Eli Zaretskii]

> Date: Tue, 14 Jan 2003 17:11:01 -0500
> From: Fernando Nasser <fnasser@redhat.com>
> 
> That was quick.  I was afraid I had started another long long thread :-)

Come on now, we all know that there's nothing like one of Fernando's
good ideas to solve a problem and end a prolonged discussion ;-)

Re: [RFC/RFA] New 'to' command

[Michael Snyder]

Fernando Nasser wrote:
> 
> Elena Zannoni wrote:> Michael Elizabeth Chastain writes:
> >  > 'advance' sounds fine to me.
> >  >
> >  > Michael C
> >
> > all right!
> >
> > Elena
> >
> 
> That was quick.  I was afraid I had started another long long thread :-)

Sometimes it takes a non-native-english-speaker.   ;-)

Re: [RFC/RFA] New 'to' command

[Fernando Nasser]

Michael Snyder wrote:> Fernando Nasser wrote:
> 
>>Elena Zannoni wrote:> Michael Elizabeth Chastain writes:
>>
>>> > 'advance' sounds fine to me.
>>> >
>>> > Michael C
>>>
>>>all right!
>>>
>>>Elena
>>>
>>
>>That was quick.  I was afraid I had started another long long thread :-)
> 
> 
> Sometimes it takes a non-native-english-speaker.   ;-)
> 

Yeah, specially when the word has a Latin root  ;-)


-- 
Fernando Nasser
Red Hat - Toronto                       E-Mail:  fnasser@redhat.com
2323 Yonge Street, Suite #300
Toronto, Ontario   M4P 2C9

[RFC/RFA] New 'to' command

[Elena Zannoni]

Following up from the long long long thread:
http://sources.redhat.com/ml/gdb-patches/2002-12/msg00584.html

Here is a new command called 'to', which takes a location (any
location) specified like for the break command, and simply continues
to it, with the restriction that the current frame is not exited.

I have left the current 'until' command alone, except for a modification
of the help string.

If this is agreed upon, I'll submit doco changes and testsuite.

2003-01-12  Elena Zannoni  <ezannoni@redhat.com>

	* breakpoint.c (until_break_command): Add new argument.  Use it to
	decide whether to stop only at the current frame or not.
	* breakpoint.h (until_break_command): Update prototype.
	* infcmd.c (until_command): Add new argument to until_break_command
	call.
	(to_command): New function.
	(_initialize_infcmd): Update help string for 'until' command.
	Add new 'to' command.


Index: breakpoint.c
===================================================================
RCS file: /cvs/uberbaum/gdb/breakpoint.c,v
retrieving revision 1.105
diff -u -p -r1.105 breakpoint.c
--- breakpoint.c	4 Jan 2003 23:07:24 -0000	1.105
+++ breakpoint.c	12 Jan 2003 19:08:07 -0000
@@ -5576,7 +5576,7 @@ until_break_command_continuation (struct
 
 /* ARGSUSED */
 void
-until_break_command (char *arg, int from_tty)
+until_break_command (char *arg, int from_tty, int anywhere)
 {
   struct symtabs_and_lines sals;
   struct symtab_and_line sal;
@@ -5609,9 +5609,16 @@ until_break_command (char *arg, int from
 
   resolve_sal_pc (&sal);
 
-  breakpoint = 
-    set_momentary_breakpoint (sal,get_frame_id (deprecated_selected_frame),
-			      bp_until);
+  if (anywhere)
+    /* If the user told us to continue until a specified location,
+       we don't specify a frame at which we need to stop.  */
+    breakpoint = set_momentary_breakpoint (sal, null_frame_id, bp_until);
+  else
+    /* Otherwise, specify the current frame, because we want to stop only
+       at the very same frame.  */
+    breakpoint = set_momentary_breakpoint (sal,
+					   get_frame_id (deprecated_selected_frame),
+					   bp_until);
 
   if (!event_loop_p || !target_can_async_p ())
     old_chain = make_cleanup_delete_breakpoint (breakpoint);
@@ -5639,8 +5646,8 @@ until_break_command (char *arg, int from
       add_continuation (until_break_command_continuation, arg1);
     }
 
-  /* Keep within the current frame */
-
+  /* Keep within the current frame, or in frames called by the current
+     one.  */
   if (prev_frame)
     {
       sal = find_pc_line (get_frame_pc (prev_frame), 0);

Index: breakpoint.h
===================================================================
RCS file: /cvs/uberbaum/gdb/breakpoint.h,v
retrieving revision 1.16
diff -u -p -r1.16 breakpoint.h
--- breakpoint.h	11 Dec 2002 22:34:47 -0000	1.16
+++ breakpoint.h	12 Jan 2003 19:08:50 -0000
@@ -534,7 +534,7 @@ extern int deprecated_frame_in_dummy (st
 
 extern int breakpoint_thread_match (CORE_ADDR, ptid_t);
 
-extern void until_break_command (char *, int);
+extern void until_break_command (char *, int, int);
 
 extern void breakpoint_re_set (void);

Index: infcmd.c
===================================================================
RCS file: /cvs/uberbaum/gdb/infcmd.c,v
retrieving revision 1.69
diff -u -p -r1.69 infcmd.c
--- infcmd.c	13 Dec 2002 16:26:02 -0000	1.69
+++ infcmd.c	12 Jan 2003 19:09:05 -0000
@@ -1140,10 +1140,38 @@ until_command (char *arg, int from_tty)
     }
 
   if (arg)
-    until_break_command (arg, from_tty);
+    until_break_command (arg, from_tty, 0);
   else
     until_next_command (from_tty);
 }
+
+static void
+to_command (char *arg, int from_tty)
+{
+  int async_exec = 0;
+
+  if (!target_has_execution)
+    error ("The program is not running.");
+
+  /* Find out whether we must run in the background. */
+  if (arg != NULL)
+    async_exec = strip_bg_char (&arg);
+
+  /* If we must run in the background, but the target can't do it,
+     error out. */
+  if (event_loop_p && async_exec && !target_can_async_p ())
+    error ("Asynchronous execution not supported on this target.");
+
+  /* If we are not asked to run in the bg, then prepare to run in the
+     foreground, synchronously. */
+  if (event_loop_p && !async_exec && target_can_async_p ())
+    {
+      /* Simulate synchronous execution */
+      async_disable_stdin ();
+    }
+
+  until_break_command (arg, from_tty, 1);
+}
 
 
 /* Print the result of a function at the end of a 'finish' command. */
@@ -2130,10 +2158,14 @@ Argument N means do this N times (or til
 
   c = add_com ("until", class_run, until_command,
 	       "Execute until the program reaches a source line greater than the current\n\
-or a specified line or address or function (same args as break command).\n\
-Execution will also stop upon exit from the current stack frame.");
+or a specified location (same args as break command) within the current frame.");
   set_cmd_completer (c, location_completer);
   add_com_alias ("u", "until", class_run, 1);
+
+  c = add_com ("to", class_run, to_command,
+	       "Continue the program up to the given location (same args as break command).\n\
+Execution will also stop upon exit from the current stack frame.");
+  set_cmd_completer (c, location_completer);
 
   c = add_com ("jump", class_run, jump_command,
 	       "Continue program being debugged at specified line or address.\n\

Re: [RFC/RFA] New 'to' command

[Daniel Jacobowitz]

On Sun, Jan 12, 2003 at 02:20:40PM -0500, Elena Zannoni wrote:
> 
> Following up from the long long long thread:
> http://sources.redhat.com/ml/gdb-patches/2002-12/msg00584.html
> 
> Here is a new command called 'to', which takes a location (any
> location) specified like for the break command, and simply continues
> to it, with the restriction that the current frame is not exited.
> 
> I have left the current 'until' command alone, except for a modification
> of the help string.
> 
> If this is agreed upon, I'll submit doco changes and testsuite.

Well, I like it just because it's nice to see us moving forwards... and
"to" is as good a name as any, I guess.  I'm worried that it doesn't
pass the obviousness test:

  - Hypothesize a forgetful Dan.  This is easy; I can provide one any
time you need one.
  - He remembers a long thread about until and to
  - But he's forgotten which one does which!
  - And he didn't think of checking in "help"!
  - So, how does he figure out which does which?

I think that the names of two commands should suggest logically
different behaviors, or we're just setting up more confusion.  I don't
see how given "until 900" and "to 900" the user could figure out which
wants the current frame.

That said, I don't mind this solution.  I'll get used to it; I suspect
anyone else who wants to use it can too.  Let's see if you satisfy
everyone else :)

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: [RFC/RFA] New 'to' command

[Elena Zannoni]

Daniel Jacobowitz writes:
 > On Sun, Jan 12, 2003 at 02:20:40PM -0500, Elena Zannoni wrote:
 > > 
 > > Following up from the long long long thread:
 > > http://sources.redhat.com/ml/gdb-patches/2002-12/msg00584.html
 > > 
 > > Here is a new command called 'to', which takes a location (any
 > > location) specified like for the break command, and simply continues
 > > to it, with the restriction that the current frame is not exited.
 > > 
 > > I have left the current 'until' command alone, except for a modification
 > > of the help string.
 > > 
 > > If this is agreed upon, I'll submit doco changes and testsuite.
 > 
 > Well, I like it just because it's nice to see us moving forwards... and
 > "to" is as good a name as any, I guess.  I'm worried that it doesn't
 > pass the obviousness test:
 > 
 >   - Hypothesize a forgetful Dan.  This is easy; I can provide one any
 > time you need one.
 >   - He remembers a long thread about until and to
 >   - But he's forgotten which one does which!
 >   - And he didn't think of checking in "help"!
 >   - So, how does he figure out which does which?
 > 
 > I think that the names of two commands should suggest logically
 > different behaviors, or we're just setting up more confusion.  I don't
 > see how given "until 900" and "to 900" the user could figure out which
 > wants the current frame.
 > 

I am not attached to either name, I just couldn't come up with better
ones.  My main rationale was to leave 'until' untouched. 

 > That said, I don't mind this solution.  I'll get used to it; I suspect
 > anyone else who wants to use it can too.  Let's see if you satisfy
 > everyone else :)
 > 

Let's hope...

Elena

 > -- 
 > Daniel Jacobowitz
 > MontaVista Software                         Debian GNU/Linux Developer

Re: [RFC/RFA] New 'to' command

[Michael Snyder]

Elena Zannoni wrote:
> 
> Daniel Jacobowitz writes:
>  > On Sun, Jan 12, 2003 at 02:20:40PM -0500, Elena Zannoni wrote:
>  > >
>  > > Following up from the long long long thread:
>  > > http://sources.redhat.com/ml/gdb-patches/2002-12/msg00584.html
>  > >
>  > > Here is a new command called 'to', which takes a location (any
>  > > location) specified like for the break command, and simply continues
>  > > to it, with the restriction that the current frame is not exited.
>  > >
>  > > I have left the current 'until' command alone, except for a modification
>  > > of the help string.
>  > >
>  > > If this is agreed upon, I'll submit doco changes and testsuite.
>  >
>  > Well, I like it just because it's nice to see us moving forwards... and
>  > "to" is as good a name as any, I guess.  I'm worried that it doesn't
>  > pass the obviousness test:
>  >
>  >   - Hypothesize a forgetful Dan.  This is easy; I can provide one any
>  > time you need one.
>  >   - He remembers a long thread about until and to
>  >   - But he's forgotten which one does which!
>  >   - And he didn't think of checking in "help"!
>  >   - So, how does he figure out which does which?
>  >
>  > I think that the names of two commands should suggest logically
>  > different behaviors, or we're just setting up more confusion.  I don't
>  > see how given "until 900" and "to 900" the user could figure out which
>  > wants the current frame.
>  >
> 
> I am not attached to either name, I just couldn't come up with better
> ones.  My main rationale was to leave 'until' untouched.
> 
>  > That said, I don't mind this solution.  I'll get used to it; I suspect
>  > anyone else who wants to use it can too.  Let's see if you satisfy
>  > everyone else :)
>  >
> 
> Let's hope...

I have no complaints.  Only -- wasn't "until" broken?
Is it still broken?
Should we make "until" and "to" mutually exclusive, ie. 
should "until" reject locations outside the current frame?

Re: [RFC/RFA] New 'to' command

[Fernando Nasser]

Sorry to get back to the name thing. But I believe the discussion concentrated 
more on the semantics of the two commands and how until would be implemented -- 
nobody was specifically fond of the names that came up in any case.

What about "advance-to"?  You can add "to" as an alias, I don't mind (although I 
don't thing we need/should as typing "adv" will suffice)..

I believe it captures the idea of what the user is trying to do in this case.

Regards to all,
Fernando



Elena Zannoni wrote:> Following up from the long long long thread:
> http://sources.redhat.com/ml/gdb-patches/2002-12/msg00584.html
> 
> Here is a new command called 'to', which takes a location (any
> location) specified like for the break command, and simply continues
> to it, with the restriction that the current frame is not exited.
> 
> I have left the current 'until' command alone, except for a modification
> of the help string.
> 
> If this is agreed upon, I'll submit doco changes and testsuite.
> 
> 2003-01-12  Elena Zannoni  <ezannoni@redhat.com>
> 
> 	* breakpoint.c (until_break_command): Add new argument.  Use it to
> 	decide whether to stop only at the current frame or not.
> 	* breakpoint.h (until_break_command): Update prototype.
> 	* infcmd.c (until_command): Add new argument to until_break_command
> 	call.
> 	(to_command): New function.
> 	(_initialize_infcmd): Update help string for 'until' command.
> 	Add new 'to' command.
> 
> 
> Index: breakpoint.c
> ===================================================================
> RCS file: /cvs/uberbaum/gdb/breakpoint.c,v
> retrieving revision 1.105
> diff -u -p -r1.105 breakpoint.c
> --- breakpoint.c	4 Jan 2003 23:07:24 -0000	1.105
> +++ breakpoint.c	12 Jan 2003 19:08:07 -0000
> @@ -5576,7 +5576,7 @@ until_break_command_continuation (struct
>  
>  /* ARGSUSED */
>  void
> -until_break_command (char *arg, int from_tty)
> +until_break_command (char *arg, int from_tty, int anywhere)
>  {
>    struct symtabs_and_lines sals;
>    struct symtab_and_line sal;
> @@ -5609,9 +5609,16 @@ until_break_command (char *arg, int from
>  
>    resolve_sal_pc (&sal);
>  
> -  breakpoint = 
> -    set_momentary_breakpoint (sal,get_frame_id (deprecated_selected_frame),
> -			      bp_until);
> +  if (anywhere)
> +    /* If the user told us to continue until a specified location,
> +       we don't specify a frame at which we need to stop.  */
> +    breakpoint = set_momentary_breakpoint (sal, null_frame_id, bp_until);
> +  else
> +    /* Otherwise, specify the current frame, because we want to stop only
> +       at the very same frame.  */
> +    breakpoint = set_momentary_breakpoint (sal,
> +					   get_frame_id (deprecated_selected_frame),
> +					   bp_until);
>  
>    if (!event_loop_p || !target_can_async_p ())
>      old_chain = make_cleanup_delete_breakpoint (breakpoint);
> @@ -5639,8 +5646,8 @@ until_break_command (char *arg, int from
>        add_continuation (until_break_command_continuation, arg1);
>      }
>  
> -  /* Keep within the current frame */
> -
> +  /* Keep within the current frame, or in frames called by the current
> +     one.  */
>    if (prev_frame)
>      {
>        sal = find_pc_line (get_frame_pc (prev_frame), 0);
> 
> Index: breakpoint.h
> ===================================================================
> RCS file: /cvs/uberbaum/gdb/breakpoint.h,v
> retrieving revision 1.16
> diff -u -p -r1.16 breakpoint.h
> --- breakpoint.h	11 Dec 2002 22:34:47 -0000	1.16
> +++ breakpoint.h	12 Jan 2003 19:08:50 -0000
> @@ -534,7 +534,7 @@ extern int deprecated_frame_in_dummy (st
>  
>  extern int breakpoint_thread_match (CORE_ADDR, ptid_t);
>  
> -extern void until_break_command (char *, int);
> +extern void until_break_command (char *, int, int);
>  
>  extern void breakpoint_re_set (void);
> 
> Index: infcmd.c
> ===================================================================
> RCS file: /cvs/uberbaum/gdb/infcmd.c,v
> retrieving revision 1.69
> diff -u -p -r1.69 infcmd.c
> --- infcmd.c	13 Dec 2002 16:26:02 -0000	1.69
> +++ infcmd.c	12 Jan 2003 19:09:05 -0000
> @@ -1140,10 +1140,38 @@ until_command (char *arg, int from_tty)
>      }
>  
>    if (arg)
> -    until_break_command (arg, from_tty);
> +    until_break_command (arg, from_tty, 0);
>    else
>      until_next_command (from_tty);
>  }
> +
> +static void
> +to_command (char *arg, int from_tty)
> +{
> +  int async_exec = 0;
> +
> +  if (!target_has_execution)
> +    error ("The program is not running.");
> +
> +  /* Find out whether we must run in the background. */
> +  if (arg != NULL)
> +    async_exec = strip_bg_char (&arg);
> +
> +  /* If we must run in the background, but the target can't do it,
> +     error out. */
> +  if (event_loop_p && async_exec && !target_can_async_p ())
> +    error ("Asynchronous execution not supported on this target.");
> +
> +  /* If we are not asked to run in the bg, then prepare to run in the
> +     foreground, synchronously. */
> +  if (event_loop_p && !async_exec && target_can_async_p ())
> +    {
> +      /* Simulate synchronous execution */
> +      async_disable_stdin ();
> +    }
> +
> +  until_break_command (arg, from_tty, 1);
> +}
>  
>  
>  /* Print the result of a function at the end of a 'finish' command. */
> @@ -2130,10 +2158,14 @@ Argument N means do this N times (or til
>  
>    c = add_com ("until", class_run, until_command,
>  	       "Execute until the program reaches a source line greater than the current\n\
> -or a specified line or address or function (same args as break command).\n\
> -Execution will also stop upon exit from the current stack frame.");
> +or a specified location (same args as break command) within the current frame.");
>    set_cmd_completer (c, location_completer);
>    add_com_alias ("u", "until", class_run, 1);
> +
> +  c = add_com ("to", class_run, to_command,
> +	       "Continue the program up to the given location (same args as break command).\n\
> +Execution will also stop upon exit from the current stack frame.");
> +  set_cmd_completer (c, location_completer);
>  
>    c = add_com ("jump", class_run, jump_command,
>  	       "Continue program being debugged at specified line or address.\n\
> 



-- 
Fernando Nasser
Red Hat - Toronto                       E-Mail:  fnasser@redhat.com
2323 Yonge Street, Suite #300
Toronto, Ontario   M4P 2C9

Re: [RFC/RFA] New 'to' command

[Elena Zannoni]

Fernando Nasser writes:
 > Sorry to get back to the name thing. But I believe the discussion concentrated 
 > more on the semantics of the two commands and how until would be implemented -- 
 > nobody was specifically fond of the names that came up in any case.
 > 
 > What about "advance-to"?  You can add "to" as an alias, I don't mind (although I 
 > don't thing we need/should as typing "adv" will suffice)..
 > 
 > I believe it captures the idea of what the user is trying to do in this case.

I like it. 
going once, going twice, sold ?

Elena


 > 
 > Regards to all,
 > Fernando
 > 
 > 
 > 
 > Elena Zannoni wrote:> Following up from the long long long thread:
 > > http://sources.redhat.com/ml/gdb-patches/2002-12/msg00584.html
 > > 
 > > Here is a new command called 'to', which takes a location (any
 > > location) specified like for the break command, and simply continues
 > > to it, with the restriction that the current frame is not exited.
 > > 
 > > I have left the current 'until' command alone, except for a modification
 > > of the help string.
 > > 
 > > If this is agreed upon, I'll submit doco changes and testsuite.
 > > 
 > > 2003-01-12  Elena Zannoni  <ezannoni@redhat.com>
 > > 
 > > 	* breakpoint.c (until_break_command): Add new argument.  Use it to
 > > 	decide whether to stop only at the current frame or not.
 > > 	* breakpoint.h (until_break_command): Update prototype.
 > > 	* infcmd.c (until_command): Add new argument to until_break_command
 > > 	call.
 > > 	(to_command): New function.
 > > 	(_initialize_infcmd): Update help string for 'until' command.
 > > 	Add new 'to' command.
 > > 
 > > 
 > > Index: breakpoint.c
 > > ===================================================================
 > > RCS file: /cvs/uberbaum/gdb/breakpoint.c,v
 > > retrieving revision 1.105
 > > diff -u -p -r1.105 breakpoint.c
 > > --- breakpoint.c	4 Jan 2003 23:07:24 -0000	1.105
 > > +++ breakpoint.c	12 Jan 2003 19:08:07 -0000
 > > @@ -5576,7 +5576,7 @@ until_break_command_continuation (struct
 > >  
 > >  /* ARGSUSED */
 > >  void
 > > -until_break_command (char *arg, int from_tty)
 > > +until_break_command (char *arg, int from_tty, int anywhere)
 > >  {
 > >    struct symtabs_and_lines sals;
 > >    struct symtab_and_line sal;
 > > @@ -5609,9 +5609,16 @@ until_break_command (char *arg, int from
 > >  
 > >    resolve_sal_pc (&sal);
 > >  
 > > -  breakpoint = 
 > > -    set_momentary_breakpoint (sal,get_frame_id (deprecated_selected_frame),
 > > -			      bp_until);
 > > +  if (anywhere)
 > > +    /* If the user told us to continue until a specified location,
 > > +       we don't specify a frame at which we need to stop.  */
 > > +    breakpoint = set_momentary_breakpoint (sal, null_frame_id, bp_until);
 > > +  else
 > > +    /* Otherwise, specify the current frame, because we want to stop only
 > > +       at the very same frame.  */
 > > +    breakpoint = set_momentary_breakpoint (sal,
 > > +					   get_frame_id (deprecated_selected_frame),
 > > +					   bp_until);
 > >  
 > >    if (!event_loop_p || !target_can_async_p ())
 > >      old_chain = make_cleanup_delete_breakpoint (breakpoint);
 > > @@ -5639,8 +5646,8 @@ until_break_command (char *arg, int from
 > >        add_continuation (until_break_command_continuation, arg1);
 > >      }
 > >  
 > > -  /* Keep within the current frame */
 > > -
 > > +  /* Keep within the current frame, or in frames called by the current
 > > +     one.  */
 > >    if (prev_frame)
 > >      {
 > >        sal = find_pc_line (get_frame_pc (prev_frame), 0);
 > > 
 > > Index: breakpoint.h
 > > ===================================================================
 > > RCS file: /cvs/uberbaum/gdb/breakpoint.h,v
 > > retrieving revision 1.16
 > > diff -u -p -r1.16 breakpoint.h
 > > --- breakpoint.h	11 Dec 2002 22:34:47 -0000	1.16
 > > +++ breakpoint.h	12 Jan 2003 19:08:50 -0000
 > > @@ -534,7 +534,7 @@ extern int deprecated_frame_in_dummy (st
 > >  
 > >  extern int breakpoint_thread_match (CORE_ADDR, ptid_t);
 > >  
 > > -extern void until_break_command (char *, int);
 > > +extern void until_break_command (char *, int, int);
 > >  
 > >  extern void breakpoint_re_set (void);
 > > 
 > > Index: infcmd.c
 > > ===================================================================
 > > RCS file: /cvs/uberbaum/gdb/infcmd.c,v
 > > retrieving revision 1.69
 > > diff -u -p -r1.69 infcmd.c
 > > --- infcmd.c	13 Dec 2002 16:26:02 -0000	1.69
 > > +++ infcmd.c	12 Jan 2003 19:09:05 -0000
 > > @@ -1140,10 +1140,38 @@ until_command (char *arg, int from_tty)
 > >      }
 > >  
 > >    if (arg)
 > > -    until_break_command (arg, from_tty);
 > > +    until_break_command (arg, from_tty, 0);
 > >    else
 > >      until_next_command (from_tty);
 > >  }
 > > +
 > > +static void
 > > +to_command (char *arg, int from_tty)
 > > +{
 > > +  int async_exec = 0;
 > > +
 > > +  if (!target_has_execution)
 > > +    error ("The program is not running.");
 > > +
 > > +  /* Find out whether we must run in the background. */
 > > +  if (arg != NULL)
 > > +    async_exec = strip_bg_char (&arg);
 > > +
 > > +  /* If we must run in the background, but the target can't do it,
 > > +     error out. */
 > > +  if (event_loop_p && async_exec && !target_can_async_p ())
 > > +    error ("Asynchronous execution not supported on this target.");
 > > +
 > > +  /* If we are not asked to run in the bg, then prepare to run in the
 > > +     foreground, synchronously. */
 > > +  if (event_loop_p && !async_exec && target_can_async_p ())
 > > +    {
 > > +      /* Simulate synchronous execution */
 > > +      async_disable_stdin ();
 > > +    }
 > > +
 > > +  until_break_command (arg, from_tty, 1);
 > > +}
 > >  
 > >  
 > >  /* Print the result of a function at the end of a 'finish' command. */
 > > @@ -2130,10 +2158,14 @@ Argument N means do this N times (or til
 > >  
 > >    c = add_com ("until", class_run, until_command,
 > >  	       "Execute until the program reaches a source line greater than the current\n\
 > > -or a specified line or address or function (same args as break command).\n\
 > > -Execution will also stop upon exit from the current stack frame.");
 > > +or a specified location (same args as break command) within the current frame.");
 > >    set_cmd_completer (c, location_completer);
 > >    add_com_alias ("u", "until", class_run, 1);
 > > +
 > > +  c = add_com ("to", class_run, to_command,
 > > +	       "Continue the program up to the given location (same args as break command).\n\
 > > +Execution will also stop upon exit from the current stack frame.");
 > > +  set_cmd_completer (c, location_completer);
 > >  
 > >    c = add_com ("jump", class_run, jump_command,
 > >  	       "Continue program being debugged at specified line or address.\n\
 > > 
 > 
 > 
 > 
 > -- 
 > Fernando Nasser
 > Red Hat - Toronto                       E-Mail:  fnasser@redhat.com
 > 2323 Yonge Street, Suite #300
 > Toronto, Ontario   M4P 2C9

Re: [RFC/RFA] New 'to' command

[Daniel Jacobowitz]

On Tue, Jan 14, 2003 at 04:09:51PM -0500, Elena Zannoni wrote:
> Fernando Nasser writes:
>  > Sorry to get back to the name thing. But I believe the discussion concentrated 
>  > more on the semantics of the two commands and how until would be implemented -- 
>  > nobody was specifically fond of the names that came up in any case.
>  > 
>  > What about "advance-to"?  You can add "to" as an alias, I don't mind (although I 
>  > don't thing we need/should as typing "adv" will suffice)..
>  > 
>  > I believe it captures the idea of what the user is trying to do in this case.
> 
> I like it. 
> going once, going twice, sold ?

Until stays in this stack frame, advance-to goes to anything as long as
the frame doesn't exit?  Sold to me.  Thanks for the name, Fernando!


-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: [RFC/RFA] New 'to' command

[Elena Zannoni]

Daniel Jacobowitz writes:
 > On Tue, Jan 14, 2003 at 04:09:51PM -0500, Elena Zannoni wrote:
 > > Fernando Nasser writes:
 > >  > Sorry to get back to the name thing. But I believe the discussion concentrated 
 > >  > more on the semantics of the two commands and how until would be implemented -- 
 > >  > nobody was specifically fond of the names that came up in any case.
 > >  > 
 > >  > What about "advance-to"?  You can add "to" as an alias, I don't mind (although I 
 > >  > don't thing we need/should as typing "adv" will suffice)..
 > >  > 
 > >  > I believe it captures the idea of what the user is trying to do in this case.
 > > 
 > > I like it. 
 > > going once, going twice, sold ?
 > 
 > Until stays in this stack frame, advance-to goes to anything as long as
 > the frame doesn't exit?  Sold to me.  Thanks for the name, Fernando!
 > 

yes, until is untouched. Same as it was. Advance is for the new command.

Ok. Phew!

Elena


 > 
 > -- 
 > Daniel Jacobowitz
 > MontaVista Software                         Debian GNU/Linux Developer