From: Doug Evans <dje@google.com>
To: Tom Tromey <tromey@redhat.com>
Cc: gdb-patches@sourceware.org
Subject: Re: [patch] fix exp/12117
Date: Wed, 13 Oct 2010 20:55:00 -0000 [thread overview]
Message-ID: <AANLkTiktjwOzwReU+8U4R0BmDfAWQoBvrod03Hy4i4T_@mail.gmail.com> (raw)
In-Reply-To: <m3r5ftap9l.fsf@fleche.redhat.com>
[-- Attachment #1: Type: text/plain, Size: 807 bytes --]
On Wed, Oct 13, 2010 at 12:03 PM, Tom Tromey <tromey@redhat.com> wrote:
> The qualified variants are all linked
> together and allocated alongside the main_type, so we don't have to
> worry about memory leaks or excess type reinstantiation.
I'm aware of that.
How about this?
[testsuite results not in yet]
2010-10-13 Doug Evans <dje@google.com>
PR exp/12117
* c-typeprint.c (c_type_print_base, case TYPE_CODE_TYPEDEF):
Verify
assumptions of when this case happens. Print "<unnamed
typedef>".
* gdbtypes.c (check_typedef): Clean up function comment.
Keep track of instance flags as we strip typedefs and create a
new
type to preserve them if necessary.
testsuite/
* gdb.cp/pr12117.cc: New file.
* gdb.cp/pr12117.exp: New file.
[-- Attachment #2: gdb-101013-12117-2.patch.txt --]
[-- Type: text/plain, Size: 7062 bytes --]
2010-10-13 Doug Evans <dje@google.com>
PR exp/12117
* c-typeprint.c (c_type_print_base, case TYPE_CODE_TYPEDEF): Verify
assumptions of when this case happens. Print "<unnamed typedef>".
* gdbtypes.c (check_typedef): Clean up function comment.
Keep track of instance flags as we strip typedefs and create a new
type to preserve them if necessary.
testsuite/
* gdb.cp/pr12117.cc: New file.
* gdb.cp/pr12117.exp: New file.
Index: c-typeprint.c
===================================================================
RCS file: /cvs/src/src/gdb/c-typeprint.c,v
retrieving revision 1.62
diff -u -p -r1.62 c-typeprint.c
--- c-typeprint.c 13 Oct 2010 15:10:10 -0000 1.62
+++ c-typeprint.c 13 Oct 2010 20:47:13 -0000
@@ -707,6 +707,13 @@ c_type_print_base (struct type *type, st
switch (TYPE_CODE (type))
{
case TYPE_CODE_TYPEDEF:
+ /* If we get here, the typedef doesn't have a name, and we couldn't
+ resolve TYPE_TARGET_TYPE. Not much we can do. */
+ gdb_assert (TYPE_NAME (type) == NULL);
+ gdb_assert (TYPE_TARGET_TYPE (type) == NULL);
+ fprintf_filtered (stream, _("<unnamed typedef>"));
+ break;
+
case TYPE_CODE_ARRAY:
case TYPE_CODE_PTR:
case TYPE_CODE_MEMBERPTR:
Index: gdbtypes.c
===================================================================
RCS file: /cvs/src/src/gdb/gdbtypes.c,v
retrieving revision 1.201
diff -u -p -r1.201 gdbtypes.c
--- gdbtypes.c 12 Oct 2010 20:58:17 -0000 1.201
+++ gdbtypes.c 13 Oct 2010 20:47:13 -0000
@@ -1347,7 +1347,17 @@ stub_noname_complaint (void)
complaint (&symfile_complaints, _("stub type has NULL name"));
}
-/* Added by Bryan Boreham, Kewill, Sun Sep 17 18:07:17 1989.
+/* Find the real type of TYPE. This function returns the real type,
+ after removing all layers of typedefs and completing opaque or stub
+ types. Completion changes the TYPE argument, but stripping of
+ typedefs does not.
+
+ NOTE: This will return a typedef if TYPE_TARGET_TYPE for the typedef has
+ not been computed and we're either in the middle of reading symbols, or
+ there was no name for the typedef in the debug info.
+
+ If TYPE is a TYPE_CODE_TYPEDEF, its length is updated to the length of
+ the target type.
If this is a stubbed struct (i.e. declared as struct foo *), see if
we can find a full definition in some other file. If so, copy this
@@ -1355,26 +1365,16 @@ stub_noname_complaint (void)
(but not any code) that if we don't find a full definition, we'd
set a flag so we don't spend time in the future checking the same
type. That would be a mistake, though--we might load in more
- symbols which contain a full definition for the type.
-
- This used to be coded as a macro, but I don't think it is called
- often enough to merit such treatment.
-
- Find the real type of TYPE. This function returns the real type,
- after removing all layers of typedefs and completing opaque or stub
- types. Completion changes the TYPE argument, but stripping of
- typedefs does not.
-
- If TYPE is a TYPE_CODE_TYPEDEF, its length is (also) set to the length of
- the target type instead of zero. However, in the case of TYPE_CODE_TYPEDEF
- check_typedef can still return different type than the original TYPE
- pointer. */
+ symbols which contain a full definition for the type. */
struct type *
check_typedef (struct type *type)
{
struct type *orig_type = type;
int is_const, is_volatile;
+ /* While we're removing typedefs, we don't want to lose qualifiers.
+ E.g., const/volatile. */
+ int instance_flags = TYPE_INSTANCE_FLAGS (type);
gdb_assert (type);
@@ -1407,8 +1407,16 @@ check_typedef (struct type *type)
TYPE_TARGET_TYPE (type) = alloc_type_arch (get_type_arch (type));
}
type = TYPE_TARGET_TYPE (type);
+ instance_flags |= TYPE_INSTANCE_FLAGS (type);
}
+ /* If necessary, create a new type to preserve any instance flags we lost
+ while stripping typedefs. */
+ if (instance_flags != TYPE_INSTANCE_FLAGS (type))
+ type = make_qualified_type (type,
+ instance_flags | TYPE_INSTANCE_FLAGS (type),
+ NULL);
+
is_const = TYPE_CONST (type);
is_volatile = TYPE_VOLATILE (type);
Index: testsuite/gdb.cp/pr12117.cc
===================================================================
RCS file: testsuite/gdb.cp/pr12117.cc
diff -N testsuite/gdb.cp/pr12117.cc
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ testsuite/gdb.cp/pr12117.cc 13 Oct 2010 20:47:13 -0000
@@ -0,0 +1,17 @@
+typedef int my_int;
+typedef const my_int const_my_int;
+typedef volatile my_int volatile_my_int;
+typedef volatile const_my_int volatile_const_my_int;
+typedef const volatile_my_int const_volatile_my_int;
+
+my_int v_my_int (0);
+const_my_int v_const_my_int (1);
+volatile_my_int v_volatile_my_int (2);
+const_volatile_my_int v_const_volatile_my_int (3);
+volatile_const_my_int v_volatile_const_my_int (4);
+
+int
+main ()
+{
+ return 0;
+}
Index: testsuite/gdb.cp/pr12117.exp
===================================================================
RCS file: testsuite/gdb.cp/pr12117.exp
diff -N testsuite/gdb.cp/pr12117.exp
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ testsuite/gdb.cp/pr12117.exp 13 Oct 2010 20:47:13 -0000
@@ -0,0 +1,55 @@
+# Copyright 2010 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/>.
+
+# This file is part of the gdb testsuite.
+
+if { [skip_cplus_tests] } { continue }
+
+load_lib "cp-support.exp"
+
+set testfile "pr12117"
+set srcfile ${testfile}.cc
+set binfile ${objdir}/${subdir}/${testfile}
+
+if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug c++}] != "" } {
+ untested ${testfile}.exp
+ return -1
+}
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+if ![runto_main] then {
+ perror "couldn't run to breakpoint"
+ continue
+}
+
+gdb_test "whatis v_my_int" "my_int"
+gdb_test "ptype v_my_int" "int"
+
+gdb_test "whatis v_const_my_int" "const_my_int"
+gdb_test "ptype v_const_my_int" "const int"
+
+gdb_test "whatis v_volatile_my_int" "volatile_my_int"
+gdb_test "ptype v_volatile_my_int" "volatile int"
+
+gdb_test "whatis v_const_volatile_my_int" "const_volatile_my_int"
+gdb_test "ptype v_const_volatile_my_int" "const volatile int"
+
+gdb_test "whatis v_volatile_const_my_int" "volatile_const_my_int"
+setup_kfail "gcc/45997" *-*-*
+gdb_test "ptype v_volatile_const_my_int" "const volatile int"
next prev parent reply other threads:[~2010-10-13 20:55 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-10-13 17:53 Doug Evans
2010-10-13 18:03 ` Doug Evans
2010-10-13 18:44 ` Tom Tromey
2010-10-13 18:55 ` Doug Evans
2010-10-13 19:04 ` Tom Tromey
2010-10-13 20:55 ` Doug Evans [this message]
2010-10-13 21:10 ` Jan Kratochvil
2010-10-14 1:55 ` Doug Evans
2010-10-14 9:35 ` Jan Kratochvil
2010-10-15 17:42 ` Tom Tromey
2010-10-15 17:49 ` Jan Kratochvil
2010-10-15 17:41 ` Tom Tromey
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=AANLkTiktjwOzwReU+8U4R0BmDfAWQoBvrod03Hy4i4T_@mail.gmail.com \
--to=dje@google.com \
--cc=gdb-patches@sourceware.org \
--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