From: Mark Kettenis <mark.kettenis@xs4all.nl>
To: jan.kratochvil@redhat.com
Cc: tromey@redhat.com, drow@false.org, gdb-patches@sourceware.org
Subject: Re: [patch] Fix `return' of long/long-long results with no debuginfo
Date: Mon, 09 Mar 2009 22:38:00 -0000 [thread overview]
Message-ID: <200903092233.n29MXEfG019900@brahms.sibelius.xs4all.nl> (raw)
In-Reply-To: <20090309015510.GA28986@host0.dyn.jankratochvil.net> (message from Jan Kratochvil on Mon, 9 Mar 2009 02:55:10 +0100)
> Date: Mon, 9 Mar 2009 02:55:10 +0100
> From: Jan Kratochvil <jan.kratochvil@redhat.com>
>
> On Thu, 05 Mar 2009 01:15:08 +0100, Tom Tromey wrote:
> > >>>>> "Mark" == Mark Kettenis <mark.kettenis@xs4all.nl> writes:
> >
> > Mark> Do we have a way to figure out if such a cast is present in the
> > Mark> expression passed as an argument to the return command?
> >
> > Yeah, it is not too hard. It just means having return_command call
> > parse_expression directly, then look at the resulting expression.
>
> Attached, check-in approval requested.
I think this is a reasonable compromise, and I think the
implementation is ok. But I'd like the opinion of one of the other
global maintainers (Daniel?, Joel?) before going this way.
> gdb/
> 2009-03-09 Jan Kratochvil <jan.kratochvil@redhat.com>
>
> * stack.c (return_command <retval_exp>): New variables retval_expr
> and old_chain. Inline parse_and_eval to initialize retval_expr. Check
> RETVAL_EXPR for UNOP_CAST and set RETURN_TYPE to the RETURN_VALUE type
> if RETURN_TYPE is NULL.
>
> gdb/testsuite/
> 2009-03-09 Jan Kratochvil <jan.kratochvil@redhat.com>
>
> * gdb.base/return-nodebug.exp, gdb.base/return-nodebug.S: New.
>
> --- gdb/stack.c 6 Mar 2009 18:51:05 -0000 1.186
> +++ gdb/stack.c 9 Mar 2009 01:52:55 -0000
> @@ -1796,18 +1796,26 @@ return_command (char *retval_exp, int fr
> message. */
> if (retval_exp)
> {
> + struct expression *retval_expr = parse_expression (retval_exp);
> + struct cleanup *old_chain = make_cleanup (xfree, retval_expr);
> struct type *return_type = NULL;
>
> /* Compute the return value. Should the computation fail, this
> call throws an error. */
> - return_value = parse_and_eval (retval_exp);
> + return_value = evaluate_expression (retval_expr);
>
> /* Cast return value to the return type of the function. Should
> the cast fail, this call throws an error. */
> if (thisfun != NULL)
> return_type = TYPE_TARGET_TYPE (SYMBOL_TYPE (thisfun));
> if (return_type == NULL)
> - return_type = builtin_type (get_frame_arch (thisframe))->builtin_int;
> + {
> + if (retval_expr->elts[0].opcode != UNOP_CAST)
> + error (_("Selected stack frame (with no associated function) "
> + "requires an explicit cast of the value to return"));
> + return_type = value_type (return_value);
> + }
> + do_cleanups (old_chain);
> CHECK_TYPEDEF (return_type);
> return_value = value_cast (return_type, return_value);
>
> --- /dev/null 1 Jan 1970 00:00:00 -0000
> +++ gdb/testsuite/gdb.base/return-nodebug.c 9 Mar 2009 01:52:55 -0000
> @@ -0,0 +1,49 @@
> +/* This testcase is part of GDB, the GNU debugger.
> +
> + Copyright 2009 Free Software Foundation, Inc.
> +
> + This program is free software; you can redistribute it and/or modify
> + it under the terms of the GNU General Public License as published by
> + the Free Software Foundation; either version 3 of the License, or
> + (at your option) any later version.
> +
> + This program is distributed in the hope that it will be useful,
> + but WITHOUT ANY WARRANTY; without even the implied warranty of
> + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> + GNU General Public License for more details.
> +
> + You should have received a copy of the GNU General Public License
> + along with this program. If not, see <http://www.gnu.org/licenses/>. */
> +
> +#include <stdio.h>
> +
> +static TYPE
> +init (void)
> +{
> + return 0;
> +}
> +
> +static TYPE
> +func (void)
> +{
> + return 31;
> +}
> +
> +static void
> +marker (void)
> +{
> +}
> +
> +int
> +main (void)
> +{
> + /* Preinitialize registers to 0 to avoid false PASS by leftover garbage. */
> + init ();
> +
> + printf ("result=" FORMAT "\n", CAST func ());
> +
> + /* Cannot `next' with no debug info. */
> + marker ();
> +
> + return 0;
> +}
> --- /dev/null 1 Jan 1970 00:00:00 -0000
> +++ gdb/testsuite/gdb.base/return-nodebug.exp 9 Mar 2009 01:52:55 -0000
> @@ -0,0 +1,60 @@
> +# Copyright (C) 2009 Free Software Foundation, Inc.
> +
> +# This program is free software; you can redistribute it and/or modify
> +# it under the terms of the GNU General Public License as published by
> +# the Free Software Foundation; either version 3 of the License, or
> +# (at your option) any later version.
> +#
> +# This program is distributed in the hope that it will be useful,
> +# but WITHOUT ANY WARRANTY; without even the implied warranty of
> +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> +# GNU General Public License for more details.
> +#
> +# You should have received a copy of the GNU General Public License
> +# along with this program. If not, see <http://www.gnu.org/licenses/>.
> +
> +proc do_test {type} {
> + set typenospace [string map {{ } -} $type]
> +
> + global pf_prefix
> + set old_prefix $pf_prefix
> + lappend pf_prefix "$typenospace:"
> +
> + if {[runto "func"]} {
> + # Test return from a function with no debug info (symbol; still it may
> + # have a minimal-symbol) if it does not crash.
> +
> + gdb_test "return -1" \
> + "Selected stack frame \\(with no associated function\\) requires an explicit cast of the value to return" \
> + "return from function with no debug info without a cast"
> +
> + # Cast of the result to the proper width must be done explicitely.
> + gdb_test "return ($type) -1" "#0 .* main \\(.*" \
> + "return from function with no debug info with a cast" \
> + "Make selected stack frame return now\\? \\(y or n\\) " "y"
> +
> + # And if it returned the full width of the result.
> + gdb_test "adv marker" "\r\nresult=-1\r\n.* in marker \\(.*" \
> + "full width of the returned result"
> + }
> +
> + set pf_prefix $old_prefix
> +}
> +
> +foreach case {{{signed char} %d (int)} \
> + {{short} %d (int)} \
> + {{int} %d} \
> + {{long} %ld} \
> + {{long long} %lld}} {
> + set type [lindex $case 0]
> + set format [lindex $case 1]
> + set cast [lindex $case 2]
> +
> + set typeesc [string map {{ } {\ }} $type]
> + set typenospace [string map {{ } -} $type]
> +
> + if {[prepare_for_testing return-nodebug.exp "return-nodebug-$typenospace" "return-nodebug.c" \
> + [list "additional_flags=-DFORMAT=\"$format\" -DTYPE=$typeesc -DCAST=$cast"]] == 0} {
> + do_test $type
> + }
> +}
>
next prev parent reply other threads:[~2009-03-09 22:33 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-02-11 19:45 Jan Kratochvil
2009-02-11 20:40 ` Mark Kettenis
2009-02-11 20:50 ` Daniel Jacobowitz
2009-02-11 21:23 ` Mark Kettenis
2009-02-11 21:47 ` Jan Kratochvil
2009-02-11 21:58 ` Mark Kettenis
2009-02-11 22:08 ` Jan Kratochvil
2009-02-11 22:38 ` Mark Kettenis
2009-02-11 22:50 ` Jan Kratochvil
2009-03-03 18:10 ` Tom Tromey
2009-03-04 21:29 ` Mark Kettenis
2009-03-05 0:15 ` Tom Tromey
2009-03-09 1:55 ` Jan Kratochvil
2009-03-09 22:38 ` Mark Kettenis [this message]
2009-03-11 16:49 ` Joel Brobecker
2009-03-11 20:23 ` Tom Tromey
2009-03-11 21:48 ` Joel Brobecker
2009-03-13 19:50 ` Jan Kratochvil
2009-03-13 23:00 ` Joel Brobecker
2009-03-14 10:45 ` Eli Zaretskii
2009-03-14 22:09 ` Jan Kratochvil
2009-03-14 22:18 ` Eli Zaretskii
2009-03-15 9:22 ` Joel Brobecker
2009-03-15 18:15 ` Jan Kratochvil
2009-03-18 4:36 ` Pedro Alves
2009-03-18 14:44 ` Joel Brobecker
2009-03-18 15:39 ` Jan Kratochvil
2009-03-18 15:46 ` Pedro Alves
2009-02-11 22:44 ` Mark Kettenis
2009-02-12 9:41 ` Jan Kratochvil
2009-02-12 14:36 ` Pierre Muller
2009-02-12 14:44 ` Jan Kratochvil
2009-02-14 21:59 ` Mark Kettenis
2009-02-21 12:47 ` Jan Kratochvil
2009-02-21 13:44 ` Mark Kettenis
2009-02-21 15:58 ` Jan Kratochvil
2009-02-21 16:21 ` Mark Kettenis
2009-02-21 16:32 ` Jan Kratochvil
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=200903092233.n29MXEfG019900@brahms.sibelius.xs4all.nl \
--to=mark.kettenis@xs4all.nl \
--cc=drow@false.org \
--cc=gdb-patches@sourceware.org \
--cc=jan.kratochvil@redhat.com \
--cc=tromey@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox