Simplify interp::exec / interp_exec - let exceptions propagate
Commit Message
This patch implements a simplication that I suggested here:
https://sourceware.org/pipermail/gdb-patches/2022-March/186320.html
Currently, the interp::exec virtual method interface is such that
subclass implementations must catch exceptions and then return them
via normal function return.
However, higher up the in chain, for the CLI we get to
interpreter_exec_cmd, which does:
for (i = 1; i < nrules; i++)
{
struct gdb_exception e = interp_exec (interp_to_use, prules[i]);
if (e.reason < 0)
{
interp_set (old_interp, 0);
error (_("error in command: \"%s\"."), prules[i]);
}
}
and for MI we get to mi_cmd_interpreter_exec, which has:
void
mi_cmd_interpreter_exec (const char *command, char **argv, int argc)
{
...
for (i = 1; i < argc; i++)
{
struct gdb_exception e = interp_exec (interp_to_use, argv[i]);
if (e.reason < 0)
error ("%s", e.what ());
}
}
Note that if those errors are reached, we lose the original
exception's error code. I can't see why we'd want that.
And, I can't see why we need to have interp_exec catch the exception
and return it via the normal return path. That's normally needed when
we need to handle propagating exceptions across C code, like across
readline or ncurses, but that's not the case here.
It seems to me that we can simplify things by removing some
try/catch-ing and just letting exceptions propagate normally.
Note, the "error in command" error shown above, which only exists in
the CLI interpreter-exec command, is only ever printed AFAICS if you
run "interpreter-exec console" when the top level interpreter is
already the console/tui. Like:
(gdb) interpreter-exec console "foobar"
Undefined command: "foobar". Try "help".
error in command: "foobar".
You won't see it with MI's "-interpreter-exec console" from a top
level MI interpreter:
(gdb)
-interpreter-exec console "foobar"
&"Undefined command: \"foobar\". Try \"help\".\n"
^error,msg="Undefined command: \"foobar\". Try \"help\"."
(gdb)
nor with MI's "-interpreter-exec mi" from a top level MI interpreter:
(gdb)
-interpreter-exec mi "-foobar"
^error,msg="Undefined MI command: foobar",code="undefined-command"
^done
(gdb)
in both these cases because MI's -interpreter-exec just does:
error ("%s", e.what ());
You won't see it either when running an MI command with the CLI's
"interpreter-exec mi":
(gdb) interpreter-exec mi "-foobar"
^error,msg="Undefined MI command: foobar",code="undefined-command"
(gdb)
This last case is because MI's interp::exec implementation never
returns an error:
gdb_exception
mi_interp::exec (const char *command)
{
mi_execute_command_wrapper (command);
return gdb_exception ();
}
Thus I think that "error in command" error is pretty pointless, and
since it simplifies things to not have it, the patch just removes it.
The patch also ends up addressing an old FIXME.
Change-Id: I5a6432a80496934ac7127594c53bf5221622e393
---
gdb/cli/cli-interp.c | 49 +++++++++++++-------------------------------
gdb/interps.c | 21 ++++++++-----------
gdb/interps.h | 5 ++---
gdb/mi/mi-interp.c | 10 ++-------
gdb/mi/mi-interp.h | 2 +-
gdb/python/py-dap.c | 3 +--
gdb/tui/tui-interp.c | 4 ++--
7 files changed, 30 insertions(+), 64 deletions(-)
base-commit: 35e1763185224b2bf634e19cdfd06dab0ba914e3
Comments
>>>>> "Pedro" == Pedro Alves <pedro@palves.net> writes:
Pedro> This patch implements a simplication that I suggested here:
Pedro> https://sourceware.org/pipermail/gdb-patches/2022-March/186320.html
Pedro> Currently, the interp::exec virtual method interface is such that
Pedro> subclass implementations must catch exceptions and then return them
Pedro> via normal function return.
...
Pedro> And, I can't see why we need to have interp_exec catch the exception
Pedro> and return it via the normal return path. That's normally needed when
Pedro> we need to handle propagating exceptions across C code, like across
Pedro> readline or ncurses, but that's not the case here.
I think this is probably a leftover from the old libgdb idea. I don't
really know much about it other than what I saw in comments, but I
gather the idea was to present a more ordinary C API, so return values
rather than longjmp exceptions. Of course we gave up on that ages ago.
Pedro> It seems to me that we can simplify things by removing some
Pedro> try/catch-ing and just letting exceptions propagate normally.
Agreed.
Approved-By: Tom Tromey <tromey@adacore.com>
Tom
On Fri, 27 Jan 2023 23:05:45 +0000
Pedro Alves <pedro@palves.net> wrote:
> This patch implements a simplication that I suggested here:
>
> https://sourceware.org/pipermail/gdb-patches/2022-March/186320.html
>
> Currently, the interp::exec virtual method interface is such that
> subclass implementations must catch exceptions and then return them
> via normal function return.
>
> However, higher up the in chain, for the CLI we get to
> interpreter_exec_cmd, which does:
>
> for (i = 1; i < nrules; i++)
> {
> struct gdb_exception e = interp_exec (interp_to_use, prules[i]);
>
> if (e.reason < 0)
> {
> interp_set (old_interp, 0);
> error (_("error in command: \"%s\"."), prules[i]);
> }
> }
>
> and for MI we get to mi_cmd_interpreter_exec, which has:
>
> void
> mi_cmd_interpreter_exec (const char *command, char **argv, int argc)
> {
> ...
> for (i = 1; i < argc; i++)
> {
> struct gdb_exception e = interp_exec (interp_to_use, argv[i]);
>
> if (e.reason < 0)
> error ("%s", e.what ());
> }
> }
>
> Note that if those errors are reached, we lose the original
> exception's error code. I can't see why we'd want that.
>
> And, I can't see why we need to have interp_exec catch the exception
> and return it via the normal return path. That's normally needed when
> we need to handle propagating exceptions across C code, like across
> readline or ncurses, but that's not the case here.
>
> It seems to me that we can simplify things by removing some
> try/catch-ing and just letting exceptions propagate normally.
>
> Note, the "error in command" error shown above, which only exists in
> the CLI interpreter-exec command, is only ever printed AFAICS if you
> run "interpreter-exec console" when the top level interpreter is
> already the console/tui. Like:
>
> (gdb) interpreter-exec console "foobar"
> Undefined command: "foobar". Try "help".
> error in command: "foobar".
>
> You won't see it with MI's "-interpreter-exec console" from a top
> level MI interpreter:
>
> (gdb)
> -interpreter-exec console "foobar"
> &"Undefined command: \"foobar\". Try \"help\".\n"
> ^error,msg="Undefined command: \"foobar\". Try \"help\"."
> (gdb)
>
> nor with MI's "-interpreter-exec mi" from a top level MI interpreter:
>
> (gdb)
> -interpreter-exec mi "-foobar"
> ^error,msg="Undefined MI command: foobar",code="undefined-command"
> ^done
> (gdb)
>
> in both these cases because MI's -interpreter-exec just does:
>
> error ("%s", e.what ());
>
> You won't see it either when running an MI command with the CLI's
> "interpreter-exec mi":
>
> (gdb) interpreter-exec mi "-foobar"
> ^error,msg="Undefined MI command: foobar",code="undefined-command"
> (gdb)
>
> This last case is because MI's interp::exec implementation never
> returns an error:
>
> gdb_exception
> mi_interp::exec (const char *command)
> {
> mi_execute_command_wrapper (command);
> return gdb_exception ();
> }
>
> Thus I think that "error in command" error is pretty pointless, and
> since it simplifies things to not have it, the patch just removes it.
>
> The patch also ends up addressing an old FIXME.
Thanks for doing this as it'll simplify part of my "Fix
gdb.base/gdb-sigterm.exp failure/error" series.
LGTM.
Kevin
@@ -48,7 +48,7 @@ class cli_interp final : public cli_interp_base
void init (bool top_level) override;
void resume () override;
void suspend () override;
- gdb_exception exec (const char *command_str) override;
+ void exec (const char *command_str) override;
ui_out *interp_ui_out () override;
private:
@@ -75,11 +75,6 @@ as_cli_interp_base (interp *interp)
return dynamic_cast<cli_interp_base *> (interp);
}
-/* Longjmp-safe wrapper for "execute_command". */
-static struct gdb_exception safe_execute_command (struct ui_out *uiout,
- const char *command,
- int from_tty);
-
/* See cli-interp.h.
Breakpoint hits should always be mirrored to a console. Deciding
@@ -314,12 +309,9 @@ cli_interp::suspend ()
gdb_disable_readline ();
}
-gdb_exception
+void
cli_interp::exec (const char *command_str)
{
- struct ui_file *old_stream;
- struct gdb_exception result;
-
/* gdb_stdout could change between the time m_cli_uiout was
initialized and now. Since we're probably using a different
interpreter which has a new ui_file for gdb_stdout, use that one
@@ -327,41 +319,28 @@ cli_interp::exec (const char *command_str)
It is important that it gets reset everytime, since the user
could set gdb to use a different interpreter. */
- old_stream = m_cli_uiout->set_stream (gdb_stdout);
- result = safe_execute_command (m_cli_uiout.get (), command_str, 1);
- m_cli_uiout->set_stream (old_stream);
- return result;
-}
-
-bool
-cli_interp_base::supports_command_editing ()
-{
- return true;
-}
-
-static struct gdb_exception
-safe_execute_command (struct ui_out *command_uiout, const char *command,
- int from_tty)
-{
- struct gdb_exception e;
+ ui_file *old_stream = m_cli_uiout->set_stream (gdb_stdout);
+ SCOPE_EXIT { m_cli_uiout->set_stream (old_stream); };
/* Save and override the global ``struct ui_out'' builder. */
scoped_restore saved_uiout = make_scoped_restore (¤t_uiout,
- command_uiout);
+ m_cli_uiout.get ());
try
{
- execute_command (command, from_tty);
+ execute_command (command_str, 1);
}
- catch (gdb_exception &exception)
+ catch (const gdb_exception_error &ex)
{
- e = std::move (exception);
+ exception_print (gdb_stderr, ex);
+ throw;
}
+}
- /* FIXME: cagney/2005-01-13: This shouldn't be needed. Instead the
- caller should print the exception. */
- exception_print (gdb_stderr, e);
- return e;
+bool
+cli_interp_base::supports_command_editing ()
+{
+ return true;
}
ui_out *
@@ -39,6 +39,7 @@
#include "top.h" /* For command_loop. */
#include "main.h"
#include "gdbsupport/buildargv.h"
+#include "gdbsupport/scope-exit.h"
/* Each UI has its own independent set of interpreters. */
@@ -332,7 +333,7 @@ interp_supports_command_editing (struct interp *interp)
/* interp_exec - This executes COMMAND_STR in the current
interpreter. */
-struct gdb_exception
+void
interp_exec (struct interp *interp, const char *command_str)
{
struct ui_interp_info *ui_interp = get_current_interp_info ();
@@ -341,7 +342,7 @@ interp_exec (struct interp *interp, const char *command_str)
scoped_restore save_command_interp
= make_scoped_restore (&ui_interp->command_interpreter, interp);
- return interp->exec (command_str);
+ interp->exec (command_str);
}
/* A convenience routine that nulls out all the common command hooks.
@@ -393,19 +394,13 @@ interpreter_exec_cmd (const char *args, int from_tty)
error (_("Could not find interpreter \"%s\"."), prules[0]);
interp_set (interp_to_use, false);
-
- for (i = 1; i < nrules; i++)
+ SCOPE_EXIT
{
- struct gdb_exception e = interp_exec (interp_to_use, prules[i]);
-
- if (e.reason < 0)
- {
- interp_set (old_interp, 0);
- error (_("error in command: \"%s\"."), prules[i]);
- }
- }
+ interp_set (old_interp, false);
+ };
- interp_set (old_interp, 0);
+ for (i = 1; i < nrules; i++)
+ interp_exec (interp_to_use, prules[i]);
}
/* See interps.h. */
@@ -36,8 +36,7 @@ typedef struct interp *(*interp_factory_func) (const char *name);
extern void interp_factory_register (const char *name,
interp_factory_func func);
-extern struct gdb_exception interp_exec (struct interp *interp,
- const char *command);
+extern void interp_exec (struct interp *interp, const char *command);
class interp
{
@@ -51,7 +50,7 @@ class interp
virtual void resume () = 0;
virtual void suspend () = 0;
- virtual gdb_exception exec (const char *command) = 0;
+ virtual void exec (const char *command) = 0;
/* Returns the ui_out currently used to collect results for this
interpreter. It can be a formatter for stdout, as is the case
@@ -199,11 +199,10 @@ mi_interp::suspend ()
gdb_disable_readline ();
}
-gdb_exception
+void
mi_interp::exec (const char *command)
{
mi_execute_command_wrapper (command);
- return gdb_exception ();
}
void
@@ -240,12 +239,7 @@ mi_cmd_interpreter_exec (const char *command, char **argv, int argc)
};
for (i = 1; i < argc; i++)
- {
- struct gdb_exception e = interp_exec (interp_to_use, argv[i]);
-
- if (e.reason < 0)
- error ("%s", e.what ());
- }
+ interp_exec (interp_to_use, argv[i]);
}
/* This inserts a number of hooks that are meant to produce
@@ -36,7 +36,7 @@ class mi_interp final : public interp
void init (bool top_level) override;
void resume () override;
void suspend () override;
- gdb_exception exec (const char *command_str) override;
+ void exec (const char *command_str) override;
ui_out *interp_ui_out () override;
void set_logging (ui_file_up logfile, bool logging_redirect,
bool debug_redirect) override;
@@ -45,10 +45,9 @@ class dap_interp final : public interp
{
}
- gdb_exception exec (const char *command) override
+ void exec (const char *command) override
{
/* Just ignore it. */
- return {};
}
void set_logging (ui_file_up logfile, bool logging_redirect,
@@ -49,7 +49,7 @@ class tui_interp final : public cli_interp_base
void init (bool top_level) override;
void resume () override;
void suspend () override;
- gdb_exception exec (const char *command_str) override;
+ void exec (const char *command_str) override;
ui_out *interp_ui_out () override;
};
@@ -149,7 +149,7 @@ tui_interp::interp_ui_out ()
return tui_old_uiout;
}
-gdb_exception
+void
tui_interp::exec (const char *command_str)
{
internal_error (_("tui_exec called"));