[patch] PR python/12438

[patch] PR python/12438

[Phil Muldoon]

This patch addresses bug PR python/12438.  Mark - and other users - have
asked that maint set python print-stack is off by default.  This makes
sense to me, as while stack-traces are useful, they can impact
readability of GDB output.  Python developers can always turn it back on
again.

Additionally, I have deprecated maint set/show python print-stack and
promoted the setting to the more user visible set/show python
print-backtrace.  Finally a note on the name change.  I wanted to keep
print-stack, but it appears when you deprecate a command it deprecates
it among all command lists.  Deprecating it and adding it back after
just un-deprecates the maint settings.  I could not find a way of just
deprecating maint set/show python print-stack, and leaving the user
visible set python print-stack alone.  If someone knows a way,
I'll happily change it.

Cheers,

Phil

--

2011-06-30  Phil Muldoon  <pmuldoon@redhat.com>

        PR python/12438
	* python/python.c: Set gdbpy_should_print_stack default to off.
	(set_python): Deprecate maint set python print-stack to
	class_deprecate.
	(_initialize_python): Deprecate maint set/show python print-stack.
	Add new prefix command, python.  Add new setting, print-backtrace.

2011-06-30  Phil Muldoon  <pmuldoon@redhat.com>

        PR python/12438
	* gdb.python/python.exp: Add maint set/show python print-stack
          deprecated tests.  Add set/show python print-backtrace tests.

2011-06-30  Phil Muldoon  <pmuldoon@redhat.com>

        PR python/12438
	* gdb.texinfo (Python Commands): Add deprecate note to maint
          set/show python print-stack.  Document set/show python
          print-backtrace.

--

diff --git a/gdb/doc/gdb.texinfo b/gdb/doc/gdb.texinfo
index 3a705c2..61419b0 100644
--- a/gdb/doc/gdb.texinfo
+++ b/gdb/doc/gdb.texinfo
@@ -20791,11 +20791,21 @@ End with a line saying just "end".
 
 @kindex maint set python print-stack
 @item maint set python print-stack
-By default, @value{GDBN} will print a stack trace when an error occurs
-in a Python script.  This can be controlled using @code{maint set
-python print-stack}: if @code{on}, the default, then Python stack
-printing is enabled; if @code{off}, then Python stack printing is
-disabled.
+This command is now deprecated.  Instead use @code{set python print
+backtrace} (@pxref{set python print-backtrace}).  By default,
+@value{GDBN} will not print a stack trace when an error occurs in a
+Python script.  This can be controlled using @code{maint set python
+print-stack}: if @code{on}, then Python stack printing is enabled; if
+@code{off}, the default, then Python stack printing is disabled.
+
+@kindex set python print-backtrace
+@item set python print-backtrace
+@anchor{set python print-backtrace}
+By default, @value{GDBN} will not print a backtrace when an error
+occurs in a Python script.  This can be controlled using @code{set
+python print-backtrace}: if @code{on}, then Python backtrace printing
+is enabled; if @code{off}, the default, then Python backtrace printing
+is disabled.
 @end table
 
 It is also possible to execute a Python script from the @value{GDBN}
diff --git a/gdb/python/python.c b/gdb/python/python.c
index ddfe9ba..5c2ff40 100644
--- a/gdb/python/python.c
+++ b/gdb/python/python.c
@@ -36,7 +36,7 @@
 
 /* True if we should print the stack when catching a Python error,
    false otherwise.  */
-static int gdbpy_should_print_stack = 1;
+static int gdbpy_should_print_stack = 0;
 
 #ifdef HAVE_PYTHON
 
@@ -944,7 +944,10 @@ struct cmd_list_element *show_python_list;
 static void
 set_python (char *args, int from_tty)
 {
-  help_list (set_python_list, "maintenance set python ", -1, gdb_stdout);
+  help_list (set_python_list, "maintenance set python ",
+	     class_deprecated, gdb_stdout);
+  help_list (set_python_list, "set python ", all_commands,
+	     gdb_stdout);
 }
 
 /* Function for use by 'maint show python' prefix command.  */
@@ -963,6 +966,9 @@ extern initialize_file_ftype _initialize_python;
 void
 _initialize_python (void)
 {
+  char *cmd_name;
+  struct cmd_list_element *cmd;
+
   add_com ("python", class_obscure, python_command,
 #ifdef HAVE_PYTHON
 	   _("\
@@ -1002,6 +1008,35 @@ Enables or disables printing of Python stack traces."),
 			   &set_python_list,
 			   &show_python_list);
 
+  /* Deprecate maint set/show python print-stack in favour of
+     non-maintenance alternatives.  */
+  cmd_name = "python print-stack";
+  cmd = lookup_cmd (&cmd_name, maintenance_set_cmdlist, "", -1, 0);
+  deprecate_cmd (cmd, "set python print-backtrace");
+  cmd_name = "python print-stack"; /* Reset name.  */
+  cmd = lookup_cmd (&cmd_name, maintenance_show_cmdlist, "", -1, 0);
+  deprecate_cmd (cmd, "show python print-backtrace");
+
+  /* Add set/show python print-stack.  */
+  add_prefix_cmd ("python", no_class, show_python,
+		  _("Prefix command for python preference settings."),
+		  &show_python_list, "show python ", 0,
+		  &showlist);
+
+  add_prefix_cmd ("python", no_class, set_python,
+		  _("Prefix command for python preference settings."),
+		  &set_python_list, "set python ", 0,
+		  &setlist);
+
+  add_setshow_boolean_cmd ("print-backtrace", no_class,
+			   &gdbpy_should_print_stack, _("\
+Enable or disable printing of Python backtraces on error."), _("\
+Show whether Python backtraces will be printed on error."), _("\
+Enables or disables printing of Python backtraces."),
+			   NULL, NULL,
+			   &set_python_list,
+			   &show_python_list);
+
 #ifdef HAVE_PYTHON
 #ifdef WITH_PYTHON_PATH
   /* Work around problem where python gets confused about where it is,
diff --git a/gdb/testsuite/gdb.python/python.exp b/gdb/testsuite/gdb.python/python.exp
index a68dd24..a36c622 100644
--- a/gdb/testsuite/gdb.python/python.exp
+++ b/gdb/testsuite/gdb.python/python.exp
@@ -178,3 +178,19 @@ gdb_test "python gdb.write(\"Foo\\n\")" "Foo" "Test default write"
 gdb_test "python gdb.write(\"Error stream\\n\", stream=gdb.STDERR)" "Error stream" "Test stderr write"
 gdb_test "python gdb.write(\"Normal stream\\n\", stream=gdb.STDOUT)" "Normal stream" "Test stdout write"
 gdb_test "python gdb.write(\"Log stream\\n\", stream=gdb.STDLOG)" "Log stream" "Test stdlog write"
+
+# Deprecate maint set/show python print-stack
+gdb_test "maint show python print-stack" \
+    "Warning: command 'maintenance show python print-stack' is deprecated.*Use 'show python print-backtrace'.*" \
+    "Test deprecation maint show warning."
+gdb_test "set python print-stack off" \
+    "Warning: command 'set python print-stack' is deprecated.*Use 'set python print-backtrace'.*" \
+    "Test deprecation maint set warning."
+gdb_test "show python print-backtrace" \
+    "Whether Python backtraces will be printed on error is off.*" \
+    "Test print-backtrace show setting. Default off."
+gdb_py_test_silent_cmd "set python print-backtrace on" \
+    "Test print-backtrace set setting" 1
+gdb_test "show python print-backtrace" \
+    "Whether Python backtraces will be printed on error is on.*" \
+    "Test print-backtrace show setting. Check On."

Re: [patch] PR python/12438

[Tom Tromey]

>>>>> "Phil" == Phil Muldoon <pmuldoon@redhat.com> writes:

Phil> Additionally, I have deprecated maint set/show python print-stack and
Phil> promoted the setting to the more user visible set/show python
Phil> print-backtrace.  Finally a note on the name change.  I wanted to keep
Phil> print-stack, but it appears when you deprecate a command it deprecates
Phil> it among all command lists.

I don't see how this could be the case.  deprecate_cmd just deprecates a
single cmd_list_element.  Could you try without the name change, and
then see what the underlying problem is instead?

Tom

Re: [patch] PR python/12438

[Phil Muldoon]

Tom Tromey <tromey@redhat.com> writes:

>>>>>> "Phil" == Phil Muldoon <pmuldoon@redhat.com> writes:
>
> Phil> Additionally, I have deprecated maint set/show python print-stack and
> Phil> promoted the setting to the more user visible set/show python
> Phil> print-backtrace.  Finally a note on the name change.  I wanted to keep
> Phil> print-stack, but it appears when you deprecate a command it deprecates
> Phil> it among all command lists.
>
> I don't see how this could be the case.  deprecate_cmd just deprecates a
> single cmd_list_element.  Could you try without the name change, and
> then see what the underlying problem is instead?

I was adding the new commands to the old list.  So you were quite
right.  Sorry about that ;)

Regenerated patch:

diff --git a/gdb/doc/gdb.texinfo b/gdb/doc/gdb.texinfo
index 129a46b..9ff3c2c 100644
--- a/gdb/doc/gdb.texinfo
+++ b/gdb/doc/gdb.texinfo
@@ -20801,10 +20801,16 @@ End with a line saying just "end".
 
 @kindex maint set python print-stack
 @item maint set python print-stack
-By default, @value{GDBN} will print a stack trace when an error occurs
-in a Python script.  This can be controlled using @code{maint set
-python print-stack}: if @code{on}, the default, then Python stack
-printing is enabled; if @code{off}, then Python stack printing is
+This command is now deprecated.  Instead use @code{set python
+print-stack} (@pxref{set python print-stack}).
+
+@kindex set python print-stack
+@item set python print-stack
+@anchor{set python print-stack}
+By default, @value{GDBN} will not print a stack trace when an error
+occurs in a Python script.  This can be controlled using @code{set
+python print-stack}: if @code{on}, then Python stack printing is
+enabled; if @code{off}, the default, then Python stack printing is
 disabled.
 @end table
 
diff --git a/gdb/python/python.c b/gdb/python/python.c
index ddfe9ba..0cca4c5 100644
--- a/gdb/python/python.c
+++ b/gdb/python/python.c
@@ -36,7 +36,7 @@
 
 /* True if we should print the stack when catching a Python error,
    false otherwise.  */
-static int gdbpy_should_print_stack = 1;
+static int gdbpy_should_print_stack = 0;
 
 #ifdef HAVE_PYTHON
 
@@ -936,23 +936,46 @@ gdbpy_breakpoint_has_py_cond (struct breakpoint_object *bp_obj)
 
 /* Lists for 'maint set python' commands.  */
 
-struct cmd_list_element *set_python_list;
-struct cmd_list_element *show_python_list;
+struct cmd_list_element *maint_set_python_list;
+struct cmd_list_element *maint_show_python_list;
+
+/* Lists for 'set python' commands.  */
+
+struct cmd_list_element *user_set_python_list;
+struct cmd_list_element *user_show_python_list;
 
 /* Function for use by 'maint set python' prefix command.  */
 
 static void
-set_python (char *args, int from_tty)
+maint_set_python (char *args, int from_tty)
 {
-  help_list (set_python_list, "maintenance set python ", -1, gdb_stdout);
+  help_list (maint_set_python_list, "maintenance set python ",
+	     class_deprecated, gdb_stdout);
 }
 
 /* Function for use by 'maint show python' prefix command.  */
 
 static void
-show_python (char *args, int from_tty)
+maint_show_python (char *args, int from_tty)
+{
+  cmd_show_list (maint_show_python_list, from_tty, "");
+}
+
+/* Function for use by 'set python' prefix command.  */
+
+static void
+user_set_python (char *args, int from_tty)
+{
+  help_list (user_set_python_list, "set python ", all_commands,
+	     gdb_stdout);
+}
+
+/* Function for use by 'show python' prefix command.  */
+
+static void
+user_show_python (char *args, int from_tty)
 {
-  cmd_show_list (show_python_list, from_tty, "");
+  cmd_show_list (user_show_python_list, from_tty, "");
 }
 
 /* Initialize the Python code.  */
@@ -963,6 +986,9 @@ extern initialize_file_ftype _initialize_python;
 void
 _initialize_python (void)
 {
+  char *cmd_name;
+  struct cmd_list_element *cmd;
+
   add_com ("python", class_obscure, python_command,
 #ifdef HAVE_PYTHON
 	   _("\
@@ -984,13 +1010,13 @@ This command is only a placeholder.")
 #endif /* HAVE_PYTHON */
 	   );
 
-  add_prefix_cmd ("python", no_class, show_python,
+  add_prefix_cmd ("python", no_class, maint_show_python,
 		  _("Prefix command for python maintenance settings."),
-		  &show_python_list, "maintenance show python ", 0,
+		  &maint_show_python_list, "maintenance show python ", 0,
 		  &maintenance_show_cmdlist);
-  add_prefix_cmd ("python", no_class, set_python,
+  add_prefix_cmd ("python", no_class, maint_set_python,
 		  _("Prefix command for python maintenance settings."),
-		  &set_python_list, "maintenance set python ", 0,
+		  &maint_set_python_list, "maintenance set python ", 0,
 		  &maintenance_set_cmdlist);
 
   add_setshow_boolean_cmd ("print-stack", class_maintenance,
@@ -999,8 +1025,37 @@ Enable or disable printing of Python stack dump on error."), _("\
 Show whether Python stack will be printed on error."), _("\
 Enables or disables printing of Python stack traces."),
 			   NULL, NULL,
-			   &set_python_list,
-			   &show_python_list);
+			   &maint_set_python_list,
+			   &maint_show_python_list);
+
+  /* Deprecate maint set/show python print-stack in favour of
+     non-maintenance alternatives.  */
+  cmd_name = "print-stack";
+  cmd = lookup_cmd (&cmd_name, maint_set_python_list, "", -1, 0);
+  deprecate_cmd (cmd, "set python print-stack");
+  cmd_name = "print-stack"; /* Reset name.  */
+  cmd = lookup_cmd (&cmd_name, maint_show_python_list, "", -1, 0);
+  deprecate_cmd (cmd, "show python print-stack");
+
+  /* Add set/show python print-stack.  */
+  add_prefix_cmd ("python", no_class, user_show_python,
+		  _("Prefix command for python preference settings."),
+		  &user_show_python_list, "show python ", 0,
+		  &showlist);
+
+  add_prefix_cmd ("python", no_class, user_set_python,
+		  _("Prefix command for python preference settings."),
+		  &user_set_python_list, "set python ", 0,
+		  &setlist);
+
+  add_setshow_boolean_cmd ("print-stack", no_class,
+			   &gdbpy_should_print_stack, _("\
+Enable or disable printing of Python stack dump on error."), _("\
+Show whether Python stack will be printed on error."), _("\
+Enables or disables printing of Python stack traces."),
+			   NULL, NULL,
+			   &user_set_python_list,
+			   &user_show_python_list);
 
 #ifdef HAVE_PYTHON
 #ifdef WITH_PYTHON_PATH
diff --git a/gdb/testsuite/gdb.python/py-function.exp b/gdb/testsuite/gdb.python/py-function.exp
index ffadb5b..d579435 100644
--- a/gdb/testsuite/gdb.python/py-function.exp
+++ b/gdb/testsuite/gdb.python/py-function.exp
@@ -93,6 +93,7 @@ gdb_py_test_multiple "Test Normal Error" \
   "NormalError ()" "" \
   "end" ""
 
+gdb_test_no_output "set python print-stack on"
 gdb_test "print \$normalerror()" "Traceback.*File.*line 5.*in invoke.*RuntimeError.*This is a Normal Error.*" \
     "Test a Runtime error.  There should be a stack trace."
 
diff --git a/gdb/testsuite/gdb.python/py-prettyprint.exp b/gdb/testsuite/gdb.python/py-prettyprint.exp
index 1fd8c6f..911725f 100644
--- a/gdb/testsuite/gdb.python/py-prettyprint.exp
+++ b/gdb/testsuite/gdb.python/py-prettyprint.exp
@@ -97,6 +97,7 @@ proc run_lang_tests {lang} {
     gdb_test_no_output "python pp_ls_encoding = 'UTF-8'"
     gdb_test "print estring2" "\"embedded \", <incomplete sequence \\\\302>"
 
+    gdb_test_no_output "set python print-stack on"
     gdb_test "print hint_error" "Exception: hint failed\r\nhint_error_val"
 
     gdb_test "print c" " = container \"container\" with 2 elements = {$nl *.0. = 23,$nl *.1. = 72$nl}"
diff --git a/gdb/testsuite/gdb.python/python.exp b/gdb/testsuite/gdb.python/python.exp
index e0deb63..8906bbc 100644
--- a/gdb/testsuite/gdb.python/python.exp
+++ b/gdb/testsuite/gdb.python/python.exp
@@ -178,3 +178,18 @@ gdb_test "python gdb.write(\"Foo\\n\")" "Foo" "Test default write"
 gdb_test "python gdb.write(\"Error stream\\n\", stream=gdb.STDERR)" "Error stream" "Test stderr write"
 gdb_test "python gdb.write(\"Normal stream\\n\", stream=gdb.STDOUT)" "Normal stream" "Test stdout write"
 gdb_test "python gdb.write(\"Log stream\\n\", stream=gdb.STDLOG)" "Log stream" "Test stdlog write"
+
+# Deprecate maint set/show python print-stack
+gdb_test "maint show python print-stack" \
+    "Warning: command 'maintenance show python print-stack' is deprecated.*Use 'show python print-stack'.*" \
+    "Test deprecation maint show warning."
+gdb_test "maint set python print-stack off" \
+    "Warning: command 'maintenance set python print-stack' is deprecated.*Use 'set python print-stack'.*" \
+    "Test deprecation maint set warning."
+gdb_test "show python print-stack" \
+    "Whether Python stack will be printed on error is off.*" \
+    "Test print-backtrace show setting. Default off."
+gdb_py_test_silent_cmd "set python print-stack on" \
+    "Test print-backtrace set setting" 1
+gdb_test "show python print-stack" \
+    "Whether Python stack will be printed on error is on.*" \

Re: [patch] PR python/12438

[Eli Zaretskii]

> From: Phil Muldoon <pmuldoon@redhat.com>
> Cc: gdb-patches@sourceware.org
> Date: Thu, 07 Jul 2011 17:11:36 +0100
> 
> --- a/gdb/doc/gdb.texinfo
> +++ b/gdb/doc/gdb.texinfo
> @@ -20801,10 +20801,16 @@ End with a line saying just "end".
>  
>  @kindex maint set python print-stack
>  @item maint set python print-stack
> -By default, @value{GDBN} will print a stack trace when an error occurs
> -in a Python script.  This can be controlled using @code{maint set
> -python print-stack}: if @code{on}, the default, then Python stack
> -printing is enabled; if @code{off}, then Python stack printing is
> +This command is now deprecated.  Instead use @code{set python
> +print-stack} (@pxref{set python print-stack}).
> +
> +@kindex set python print-stack
> +@item set python print-stack
> +@anchor{set python print-stack}
> +By default, @value{GDBN} will not print a stack trace when an error
> +occurs in a Python script.  This can be controlled using @code{set
> +python print-stack}: if @code{on}, then Python stack printing is
> +enabled; if @code{off}, the default, then Python stack printing is
>  disabled.
>  @end table

Thanks.  This part is okay, but please lose the @pxref and the anchor,
because the referenced part is just one line away.  Instead, say
something like "..., described below."

Re: [patch] PR python/12438

[Tom Tromey]

>>>>> "Phil" == Phil Muldoon <pmuldoon@redhat.com> writes:

Phil> I was adding the new commands to the old list.  So you were quite
Phil> right.  Sorry about that ;)

No problem.

I think this change should have a NEWS entry -- both to document that
the default changed and to mention the new commands.

Phil> +struct cmd_list_element *maint_set_python_list;
Phil> +struct cmd_list_element *maint_show_python_list;
Phil> +
Phil> +/* Lists for 'set python' commands.  */
Phil> +
Phil> +struct cmd_list_element *user_set_python_list;
Phil> +struct cmd_list_element *user_show_python_list;
 
These can all be static.

Code bits ok with that change.

Tom

Re: [patch] PR python/12438

[Phil Muldoon]

Tom Tromey <tromey@redhat.com> writes:

>>>>>> "Phil" == Phil Muldoon <pmuldoon@redhat.com> writes:
>
> Phil> I was adding the new commands to the old list.  So you were quite
> Phil> right.  Sorry about that ;)
>
> No problem.
>
> I think this change should have a NEWS entry -- both to document that
> the default changed and to mention the new commands.
>
> Phil> +struct cmd_list_element *maint_set_python_list;
> Phil> +struct cmd_list_element *maint_show_python_list;
> Phil> +
> Phil> +/* Lists for 'set python' commands.  */
> Phil> +
> Phil> +struct cmd_list_element *user_set_python_list;
> Phil> +struct cmd_list_element *user_show_python_list;
>  
> These can all be static.
>
> Code bits ok with that change.

So committed with doc and code changes.

Cheers

Phil

Re: [patch] PR python/12438

[Kevin Pouget]

On Thu, Jun 30, 2011 at 11:37 AM, Phil Muldoon <pmuldoon@redhat.com> wrote:
>
> This patch addresses bug PR python/12438.  Mark - and other users - have
> asked that maint set python print-stack is off by default.  This makes
> sense to me, as while stack-traces are useful, they can impact
> readability of GDB output.  Python developers can always turn it back on
> again.

Hello,

I have a question regarding this patch, which was committed at the end of June,
wouldn't GDB let the user know, one way or an other, that there was
something wrong happened ?

when you run a command, and see to "warning/error" message, you expect
that everything went right, but that's not true when this flag is set,
is it?

I don't know if Python allows to do it, but i think it would be nice
to see something like:

> NameError: global name 'comp' is not define

which is the last line of a python stacktrace


Thanks,

Kevin

Re: [patch] PR python/12438

[Phil Muldoon]

Kevin Pouget <kevin.pouget@gmail.com> writes:

> On Thu, Jun 30, 2011 at 11:37 AM, Phil Muldoon <pmuldoon@redhat.com> wrote:
>>
>> This patch addresses bug PR python/12438.  Mark - and other users - have
>> asked that maint set python print-stack is off by default.  This makes
>> sense to me, as while stack-traces are useful, they can impact
>> readability of GDB output.  Python developers can always turn it back on
>> again.
>
> Hello,
>
> I have a question regarding this patch, which was committed at the end of June,
> wouldn't GDB let the user know, one way or an other, that there was
> something wrong happened ?
>
> when you run a command, and see to "warning/error" message, you expect
> that everything went right, but that's not true when this flag is set,
> is it?
>
> I don't know if Python allows to do it, but i think it would be nice
> to see something like:
>
>> NameError: global name 'comp' is not define
>
> which is the last line of a python stacktrace

It was set that way so that "shipped" versions of GDB would not have
endless amounts of exceptions when pretty-printing.  However I too have
begun to have second thoughts.  I do multiple builds, and, inevitably, I
forget to turn error-printing back on, so that really throws me
sometimes.  Fortunately, the should-print/should-not-print and the
resultant exception printing code is located centrally.  I think your
idea is a good one, though I am trying to decide how useful that
information would be, and if something additional brief data should be
printed.

Cheers,

Phil

Re: [patch] PR python/12438

[Tom Tromey]

Kevin> I have a question regarding this patch, which was committed at
Kevin> the end of June, wouldn't GDB let the user know, one way or an
Kevin> other, that there was something wrong happened ?

I think gdbpy_print_stack is mostly called in "internal" situations,
where printing something will mess up the output.

Ordinary commands and such that fail should convert the Python exception
to a gdb exception, leading to what you'd expect.

I might be misremembering.  Concrete examples would help.

Kevin> I don't know if Python allows to do it, but i think it would be nice
Kevin> to see something like:

>> NameError: global name 'comp' is not define

Kevin> which is the last line of a python stacktrace

I think it could be done.
We can always add more values for "maint set python print-stack".

Tom

Re: [patch] PR python/12438

[Phil Muldoon]

Tom Tromey <tromey@redhat.com> writes:


>>> NameError: global name 'comp' is not define
>
> Kevin> which is the last line of a python stacktrace
>
> I think it could be done.
> We can always add more values for "maint set python print-stack".

We deprecated that command when we flipped the stack-trace switch.  The
new command is "set python print-stack"

Cheers,

Phil

Re: [patch] PR python/12438

[Kevin Pouget]

On Wed, Oct 19, 2011 at 10:56 PM, Tom Tromey <tromey@redhat.com> wrote:
> Kevin> I have a question regarding this patch, which was committed at
> Kevin> the end of June, wouldn't GDB let the user know, one way or an
> Kevin> other, that there was something wrong happened ?
>
> I think gdbpy_print_stack is mostly called in "internal" situations,
> where printing something will mess up the output.
>
> Ordinary commands and such that fail should convert the Python exception
> to a gdb exception, leading to what you'd expect.
>
> I might be misremembering.  Concrete examples would help.

I don't think so about gdbpy_print_stack, I think it's used every time
an exception is caught, either originating from [user-] python code,
or from GDB's internals

the first example I can thing about it:

class BP(gdb.Breakpoint):	
	def stop(self):
		raise ValueError

which will silently fail if "set python print-stack" is off

> Kevin> I don't know if Python allows to do it, but i think it would be nice
> Kevin> to see something like:
>
>>> NameError: global name 'comp' is not define
>
> Kevin> which is the last line of a python stacktrace
>
> I think it could be done.
> We can always add more values for "maint set python print-stack".

I posted a bug report for that;
http://sourceware.org/bugzilla/show_bug.cgi?id=13329

I'll give in a try in the next weeks or so, if still available



Thanks,

Kevin