Skip to content

Commit

Permalink
Fix PR c++/21323: GDB thinks char16_t and char32_t are signed in C++
Browse files Browse the repository at this point in the history
While the C++ standard says that char16_t and char32_t are unsigned types:

 Types char16_t and char32_t denote distinct types with the same size,
 signedness, and alignment as uint_least16_t and uint_least32_t,
 respectively, in <cstdint>, called the underlying types.

... gdb treats them as signed currently:

 (gdb) p (char16_t)-1
 $1 = -1 u'\xffff'

There are actually two places in gdb that hardcode these types:

- gdbtypes.c:gdbtypes_post_init, when creating the built-in types,
  seemingly used by the "x /s" command (judging from commit 9a22f0d).

- dwarf2read.c, when reading base types with DW_ATE_UTF encoding
  (which is what is used for these types, when compiling for C++11 and
  up).  Despite the comment, the type created does end up used.

Both places need fixing.  But since I couldn't tell why dwarf2read.c
needs to create a new type, I've made it use the per-arch built-in
types instead, so that the types are only created once per arch
instead of once per objfile.  That seems to work fine.

While writting the test, I noticed that the C++ language parser isn't
actually aware of these built-in types, so if you try to use them
without a program that uses them, you get:

 (gdb) set language c++
 (gdb) ptype char16_t
 No symbol table is loaded.  Use the "file" command.
 (gdb) ptype u"hello"
 No type named char16_t.
 (gdb) p u"hello"
 No type named char16_t.

That's fixed by simply adding a couple entries to C++'s built-in types
array in c-lang.c.  With that, we get the expected:

 (gdb) ptype char16_t
 type = char16_t
 (gdb) ptype u"hello"
 type = char16_t [6]
 (gdb) p u"hello"
 $1 = u"hello"

gdb/ChangeLog:
2017-04-12  Pedro Alves  <palves@redhat.com>

	PR c++/21323
	* c-lang.c (cplus_primitive_types) <cplus_primitive_type_char16_t,
	cplus_primitive_type_char32_t>: New enum values.
	(cplus_language_arch_info): Register cplus_primitive_type_char16_t
	and cplus_primitive_type_char32_t.
	* dwarf2read.c (read_base_type) <DW_ATE_UTF>: If bit size is 16 or
	32, use the archtecture's built-in type for char16_t and char32_t,
	respectively.  Otherwise, fallback to init_integer_type as before,
	but make the type unsigned, and issue a complaint.
	* gdbtypes.c (gdbtypes_post_init): Make char16_t and char32_t unsigned.

gdb/testsuite/ChangeLog:
2017-04-12  Pedro Alves  <palves@redhat.com>

	PR c++/21323
	* gdb.cp/wide_char_types.c: New file.
	* gdb.cp/wide_char_types.exp: New file.
  • Loading branch information
palves committed Apr 12, 2017
1 parent 5e0e042 commit 53e710a
Show file tree
Hide file tree
Showing 7 changed files with 214 additions and 6 deletions.
13 changes: 13 additions & 0 deletions gdb/ChangeLog
Original file line number Diff line number Diff line change
@@ -1,3 +1,16 @@
2017-04-12 Pedro Alves <palves@redhat.com>

PR c++/21323
* c-lang.c (cplus_primitive_types) <cplus_primitive_type_char16_t,
cplus_primitive_type_char32_t>: New enum values.
(cplus_language_arch_info): Register cplus_primitive_type_char16_t
and cplus_primitive_type_char32_t.
* dwarf2read.c (read_base_type) <DW_ATE_UTF>: If bit size is 16 or
32, use the archtecture's built-in type for char16_t and char32_t,
respectively. Otherwise, fallback to init_integer_type as before,
but make the type unsigned, and issue a complaint.
* gdbtypes.c (gdbtypes_post_init): Make char16_t and char32_t unsigned.

2017-04-12 Alan Hayward <alan.hayward@arm.com>

* m32r-tdep.c (M32R_ARG_REGISTER_SIZE): Added.
Expand Down
6 changes: 6 additions & 0 deletions gdb/c-lang.c
Original file line number Diff line number Diff line change
Expand Up @@ -895,6 +895,8 @@ enum cplus_primitive_types {
cplus_primitive_type_decfloat,
cplus_primitive_type_decdouble,
cplus_primitive_type_declong,
cplus_primitive_type_char16_t,
cplus_primitive_type_char32_t,
nr_cplus_primitive_types
};

Expand Down Expand Up @@ -950,6 +952,10 @@ cplus_language_arch_info (struct gdbarch *gdbarch,
= builtin->builtin_decdouble;
lai->primitive_type_vector [cplus_primitive_type_declong]
= builtin->builtin_declong;
lai->primitive_type_vector [cplus_primitive_type_char16_t]
= builtin->builtin_char16;
lai->primitive_type_vector [cplus_primitive_type_char32_t]
= builtin->builtin_char32;

lai->bool_type_symbol = "bool";
lai->bool_type_default = builtin->builtin_bool;
Expand Down
19 changes: 16 additions & 3 deletions gdb/dwarf2read.c
Original file line number Diff line number Diff line change
Expand Up @@ -15125,9 +15125,22 @@ read_base_type (struct die_info *die, struct dwarf2_cu *cu)
type = init_integer_type (objfile, bits, 1, name);
break;
case DW_ATE_UTF:
/* We just treat this as an integer and then recognize the
type by name elsewhere. */
type = init_integer_type (objfile, bits, 0, name);
{
gdbarch *arch = get_objfile_arch (objfile);

if (bits == 16)
type = builtin_type (arch)->builtin_char16;
else if (bits == 32)
type = builtin_type (arch)->builtin_char32;
else
{
complaint (&symfile_complaints,
_("unsupported DW_ATE_UTF bit size: '%d'"),
bits);
type = init_integer_type (objfile, bits, 1, name);
}
return set_die_type (die, type, cu);
}
break;

default:
Expand Down
5 changes: 2 additions & 3 deletions gdb/gdbtypes.c
Original file line number Diff line number Diff line change
Expand Up @@ -5204,10 +5204,9 @@ gdbtypes_post_init (struct gdbarch *gdbarch)

/* Wide character types. */
builtin_type->builtin_char16
= arch_integer_type (gdbarch, 16, 0, "char16_t");
= arch_integer_type (gdbarch, 16, 1, "char16_t");
builtin_type->builtin_char32
= arch_integer_type (gdbarch, 32, 0, "char32_t");

= arch_integer_type (gdbarch, 32, 1, "char32_t");

/* Default data/code pointer types. */
builtin_type->builtin_data_ptr
Expand Down
6 changes: 6 additions & 0 deletions gdb/testsuite/ChangeLog
Original file line number Diff line number Diff line change
@@ -1,3 +1,9 @@
2017-04-12 Pedro Alves <palves@redhat.com>

PR c++/21323
* gdb.cp/wide_char_types.c: New file.
* gdb.cp/wide_char_types.exp: New file.

2017-04-05 Sergio Durigan Junior <sergiodj@redhat.com>

PR gdb/21352
Expand Down
28 changes: 28 additions & 0 deletions gdb/testsuite/gdb.cp/wide_char_types.c
Original file line number Diff line number Diff line change
@@ -0,0 +1,28 @@
/* This testcase is part of GDB, the GNU debugger.
Copyright 2017 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 <uchar.h>

char16_t u16 = -1;
char32_t u32 = -1;

int
main ()
{
return 0;
}
143 changes: 143 additions & 0 deletions gdb/testsuite/gdb.cp/wide_char_types.exp
Original file line number Diff line number Diff line change
@@ -0,0 +1,143 @@
# This testcase is part of GDB, the GNU debugger.

# Copyright 2017 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/>.

# Test GDB's awareness of the char16_t, char32_t (C++11+) built-in
# types. We also run most tests here in C mode, and check whether the
# built-ins are disabled (gdb uses the typedefs in the debug info
# instead.)

standard_testfile

# Test char16_t/char32_t in language LANG, against symbols in
# a program. Lang can be "c", "c++03" or "c++11". In C++11,
# char16_t/char32_t are built-in types, and the debug information
# reflects that (see
# http://wiki.dwarfstd.org/index.php?title=C%2B%2B0x:_New_string_literals).

proc wide_char_types_program {lang} {
global srcfile testfile

set options {debug}
if {$lang == "c++03"} {
lappend options c++ additional_flags=-std=c++03
set out $testfile-cxx03
} elseif {$lang == "c++11"} {
lappend options c++ additional_flags=-std=c++11
set out $testfile-cxx11
} else {
set out $testfile-c
}

if { [prepare_for_testing "failed to prepare" \
${out} [list $srcfile] $options] } {
return -1
}

if ![runto_main] then {
fail "can't run to main"
return 0
}
do_test_wide_char $lang "u16" "u32"
}

# Test char16_t/char32_t in language LANG. Use CHAR16_EXP and
# CHAR32_EXP as expression for each of the corresponding types.
# (E.g., CHAR16_EXP will be u16 when testing against the program, and
# "(char16_t)-1" when testing the built-in types without a program
# loaded.)

proc do_test_wide_char {lang char16_exp char32_exp} {
global gdb_prompt

# Check that the fixed-width wide types are distinct built-in
# types in C++11+. In other modes, they're instead typedefs,
# found in the debug info.
if {$lang == "c++11"} {
gdb_test "ptype $char16_exp" "type = char16_t" \
"char16_t is distinct"
gdb_test "ptype $char32_exp" "type = char32_t" \
"char32_t is distinct"
} else {
gdb_test "ptype $char16_exp" "type = unsigned (long|int|short)" \
"char16_t is typedef"
gdb_test "ptype $char32_exp" "type = unsigned (long|int|short)" \
"char32_t is typedef"
}

# Check that the fixed-width wide char types are unsigned.
gdb_test "p $char16_exp" " = 65535 u'\\\\xffff'" \
"char16_t is unsigned"
gdb_test "p $char32_exp" " = 4294967295 U'\\\\xffffffff'" \
"char32_t is unsigned"

# Check sizeof. These are fixed-width.
gdb_test "p sizeof($char16_exp)" "= 2" \
"sizeof($char16_exp) == 2"
gdb_test "p sizeof($char32_exp)" "= 4" \
"sizeof(char16_t) == 4"

# Test printing wide literal strings. Note that when testing with
# no program started, this relies on GDB's awareness of the
# built-in wide char types.
gdb_test {p U"hello"} {= U"hello"}
gdb_test {p u"hello"} {= u"hello"}
}

# Make sure that the char16_t/char32_t types are recognized as
# distinct built-in types in C++ mode, even with no program loaded.
# Check that in C mode, the types are not recognized.

proc wide_char_types_no_program {} {
global srcfile testfile

gdb_exit
gdb_start

# These types are not built-in in C.
with_test_prefix "c" {
gdb_test "set language c"

gdb_test "p (char16_t) -1" "No symbol table is loaded.*" \
"char16_t is not built-in"
gdb_test "p (char32_t) -1" "No symbol table is loaded.*" \
"char32_t is not built-in"

gdb_test {p U"hello"} "No type named char32_t\\\."
gdb_test {p u"hello"} "No type named char16_t\\\."
}

# Note GDB does not distinguish C++ dialects, so the fixed-width
# types are always available in C++ mode, even if they were not
# built-in types before C++11.
with_test_prefix "c++" {
gdb_test "set language c++"

do_test_wide_char "c++11" "(char16_t) -1" "(char32_t) -1"
}
}

# Check wide char types with no program loaded.
with_test_prefix "no program" {
wide_char_types_no_program
}

# Check types when a program is loaded.
with_test_prefix "with program" {
foreach_with_prefix lang {"c" "c++03" "c++11"} {
wide_char_types_program $lang
}
}

0 comments on commit 53e710a

Please sign in to comment.