* [patch] Fix internal error on canonicalization of clang types [Re: why in cplus_demangle_operators operation sizeof is expected to be followed with space]
[not found] ` <CAGj4m+46oBXSYQ-wx9g-P7yevZeMDnNNEjvADE8sRSdoWaRiVA@mail.gmail.com>
@ 2012-09-17 6:27 ` Jan Kratochvil
2012-09-17 8:29 ` Vasyl Vavrychuk
0 siblings, 1 reply; 11+ messages in thread
From: Jan Kratochvil @ 2012-09-17 6:27 UTC (permalink / raw)
To: Vasyl Vavrychuk; +Cc: gdb-patches
[-- Attachment #1: Type: text/plain, Size: 1625 bytes --]
On Mon, 17 Sep 2012 07:49:16 +0200, Vasyl Vavrychuk wrote:
> This looks to be not a symbol but type name.
OK, reproduced [attached], patched DWARF of "int" in C++: void f(int i) {}
> But I have found it directly in the binary. Line containing this text looks like
readelf -wi
<1><59>: Abbrev Number: 4 (DW_TAG_base_type)
<5c> DW_AT_name : exact_unsigned_base_helper<sizeof(unsigned int) * 8>
> Do you think the problem is in clang? Could you point out some
> specification, so I can reference it while contacting clang people.
Itanium C++ ABI does not say anything about spacing:
http://www.swag.uwaterloo.ca/acd/docs/ItaniumC++ABI.htm
(the original URL is 404 now)
I guess so far it is only agreed upon between GCC and GDB, it is true GDB
should accept any valid spacing of the language, this is also what
cp_canonicalize_string is there for.
Your patch would belons to the GCC mailing lists where libiberty/ is
maintained.
But I find more correct patch the one below.
For a testcase could you provide a minimal C++ code to produce such type?
No regressions on {x86_64,x86_64-m32,i686}-fedora18-linux-gnu.
Thanks,
Jan
gdb/
2012-09-17 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix internal error on canonicalization of clang types.
* cp-name-parser.y (exp): Use "sizeof ".
diff --git a/gdb/cp-name-parser.y b/gdb/cp-name-parser.y
index 6767fa5..c57ce64 100644
--- a/gdb/cp-name-parser.y
+++ b/gdb/cp-name-parser.y
@@ -1183,7 +1183,7 @@ exp : FLOAT
;
exp : SIZEOF '(' type ')' %prec UNARY
- { $$ = d_unary ("sizeof", $3); }
+ { $$ = d_unary ("sizeof ", $3); }
;
/* C++. */
[-- Attachment #2: type.S --]
[-- Type: text/plain, Size: 4890 bytes --]
.file "type.C"
.text
.Ltext0:
.globl _Z1fi
.type _Z1fi, @function
_Z1fi:
.LFB0:
.file 1 "type.C"
# type.C:1
.loc 1 1 0
.cfi_startproc
# BLOCK 2 seq:0
# PRED: ENTRY (fallthru)
pushq %rbp
.cfi_def_cfa_offset 16
.cfi_offset 6, -16
movq %rsp, %rbp
.cfi_def_cfa_register 6
movl %edi, -4(%rbp)
# type.C:1
.loc 1 1 0
popq %rbp
.cfi_def_cfa 7, 8
# SUCC: EXIT [100.0%]
ret
.cfi_endproc
.LFE0:
.size _Z1fi, .-_Z1fi
.Letext0:
.section .debug_info,"",@progbits
.Ldebug_info0:
.long 1f-2f # Length of Compilation Unit Info
2:
.value 0x4 # DWARF version number
.long .Ldebug_abbrev0 # Offset Into Abbrev. Section
.byte 0x8 # Pointer Size (in bytes)
.uleb128 0x1 # (DIE (0xb) DW_TAG_compile_unit)
.long .LASF0 # DW_AT_producer: "GNU C++ 4.7.1 20120720 (Red Hat 4.7.1-5) -mtune=generic -march=x86-64 -g"
.byte 0x4 # DW_AT_language
.long .LASF1 # DW_AT_name: "type.C"
.long .LASF2 # DW_AT_comp_dir: "/home/jkratoch/t"
.quad .Ltext0 # DW_AT_low_pc
.quad .Letext0 # DW_AT_high_pc
.long .Ldebug_line0 # DW_AT_stmt_list
.uleb128 0x2 # (DIE (0x2d) DW_TAG_subprogram)
# DW_AT_external
.ascii "f\0" # DW_AT_name
.byte 0x1 # DW_AT_decl_file (type.C)
.byte 0x1 # DW_AT_decl_line
.long .LASF3 # DW_AT_linkage_name: "_Z1fi"
.quad .LFB0 # DW_AT_low_pc
.quad .LFE0 # DW_AT_high_pc
.uleb128 0x1 # DW_AT_frame_base
.byte 0x9c # DW_OP_call_frame_cfa
# DW_AT_GNU_all_call_sites
.long 0x59 # DW_AT_sibling
.uleb128 0x3 # (DIE (0x4c) DW_TAG_formal_parameter)
.ascii "i\0" # DW_AT_name
.byte 0x1 # DW_AT_decl_file (type.C)
.byte 0x1 # DW_AT_decl_line
.long 0x59 # DW_AT_type
.uleb128 0x2 # DW_AT_location
.byte 0x91 # DW_OP_fbreg
.sleb128 -20
.byte 0 # end of children of DIE 0x2d
.uleb128 0x4 # (DIE (0x59) DW_TAG_base_type)
.byte 0x4 # DW_AT_byte_size
.byte 0x5 # DW_AT_encoding
#if 0
.ascii "int\0" # DW_AT_name
#else
.ascii "exact_unsigned_base_helper<sizeof(unsigned int) * 8>\0" # DW_AT_name
#endif
.byte 0 # end of children of DIE 0xb
1:
.section .debug_abbrev,"",@progbits
.Ldebug_abbrev0:
.uleb128 0x1 # (abbrev code)
.uleb128 0x11 # (TAG: DW_TAG_compile_unit)
.byte 0x1 # DW_children_yes
.uleb128 0x25 # (DW_AT_producer)
.uleb128 0xe # (DW_FORM_strp)
.uleb128 0x13 # (DW_AT_language)
.uleb128 0xb # (DW_FORM_data1)
.uleb128 0x3 # (DW_AT_name)
.uleb128 0xe # (DW_FORM_strp)
.uleb128 0x1b # (DW_AT_comp_dir)
.uleb128 0xe # (DW_FORM_strp)
.uleb128 0x11 # (DW_AT_low_pc)
.uleb128 0x1 # (DW_FORM_addr)
.uleb128 0x12 # (DW_AT_high_pc)
.uleb128 0x1 # (DW_FORM_addr)
.uleb128 0x10 # (DW_AT_stmt_list)
.uleb128 0x17 # (DW_FORM_sec_offset)
.byte 0
.byte 0
.uleb128 0x2 # (abbrev code)
.uleb128 0x2e # (TAG: DW_TAG_subprogram)
.byte 0x1 # DW_children_yes
.uleb128 0x3f # (DW_AT_external)
.uleb128 0x19 # (DW_FORM_flag_present)
.uleb128 0x3 # (DW_AT_name)
.uleb128 0x8 # (DW_FORM_string)
.uleb128 0x3a # (DW_AT_decl_file)
.uleb128 0xb # (DW_FORM_data1)
.uleb128 0x3b # (DW_AT_decl_line)
.uleb128 0xb # (DW_FORM_data1)
.uleb128 0x6e # (DW_AT_linkage_name)
.uleb128 0xe # (DW_FORM_strp)
.uleb128 0x11 # (DW_AT_low_pc)
.uleb128 0x1 # (DW_FORM_addr)
.uleb128 0x12 # (DW_AT_high_pc)
.uleb128 0x1 # (DW_FORM_addr)
.uleb128 0x40 # (DW_AT_frame_base)
.uleb128 0x18 # (DW_FORM_exprloc)
.uleb128 0x2117 # (DW_AT_GNU_all_call_sites)
.uleb128 0x19 # (DW_FORM_flag_present)
.uleb128 0x1 # (DW_AT_sibling)
.uleb128 0x13 # (DW_FORM_ref4)
.byte 0
.byte 0
.uleb128 0x3 # (abbrev code)
.uleb128 0x5 # (TAG: DW_TAG_formal_parameter)
.byte 0 # DW_children_no
.uleb128 0x3 # (DW_AT_name)
.uleb128 0x8 # (DW_FORM_string)
.uleb128 0x3a # (DW_AT_decl_file)
.uleb128 0xb # (DW_FORM_data1)
.uleb128 0x3b # (DW_AT_decl_line)
.uleb128 0xb # (DW_FORM_data1)
.uleb128 0x49 # (DW_AT_type)
.uleb128 0x13 # (DW_FORM_ref4)
.uleb128 0x2 # (DW_AT_location)
.uleb128 0x18 # (DW_FORM_exprloc)
.byte 0
.byte 0
.uleb128 0x4 # (abbrev code)
.uleb128 0x24 # (TAG: DW_TAG_base_type)
.byte 0 # DW_children_no
.uleb128 0xb # (DW_AT_byte_size)
.uleb128 0xb # (DW_FORM_data1)
.uleb128 0x3e # (DW_AT_encoding)
.uleb128 0xb # (DW_FORM_data1)
.uleb128 0x3 # (DW_AT_name)
.uleb128 0x8 # (DW_FORM_string)
.byte 0
.byte 0
.byte 0
.section .debug_aranges,"",@progbits
.long 0x2c # Length of Address Ranges Info
.value 0x2 # DWARF Version
.long .Ldebug_info0 # Offset of Compilation Unit Info
.byte 0x8 # Size of Address
.byte 0 # Size of Segment Descriptor
.value 0 # Pad to 16 byte boundary
.value 0
.quad .Ltext0 # Address
.quad .Letext0-.Ltext0 # Length
.quad 0
.quad 0
.section .debug_line,"",@progbits
.Ldebug_line0:
.section .debug_str,"MS",@progbits,1
.LASF1:
.string "type.C"
.LASF2:
.string "/home/jkratoch/t"
.LASF0:
.string "GNU C++ 4.7.1 20120720 (Red Hat 4.7.1-5) -mtune=generic -march=x86-64 -g"
.LASF3:
.string "_Z1fi"
.ident "GCC: (GNU) 4.7.1 20120720 (Red Hat 4.7.1-5)"
.section .note.GNU-stack,"",@progbits
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [patch] Fix internal error on canonicalization of clang types [Re: why in cplus_demangle_operators operation sizeof is expected to be followed with space]
2012-09-17 6:27 ` [patch] Fix internal error on canonicalization of clang types [Re: why in cplus_demangle_operators operation sizeof is expected to be followed with space] Jan Kratochvil
@ 2012-09-17 8:29 ` Vasyl Vavrychuk
2012-09-17 15:30 ` [patch] Fix internal error on canonicalization of clang types (+lib/gdb.exp internal error catch) Jan Kratochvil
0 siblings, 1 reply; 11+ messages in thread
From: Vasyl Vavrychuk @ 2012-09-17 8:29 UTC (permalink / raw)
To: Jan Kratochvil; +Cc: gdb-patches
Problem happening for the file
#include <cstdlib>
#include <boost/uuid/uuid_generators.hpp>
int main(int argc, char* argv[]) {
boost::uuids::random_generator gen;
return EXIT_SUCCESS;
}
clang++ -g -O0 sizeof-space-test.cpp
readelf -wi a.out|grep exact_unsigned_base_helper
<4072> DW_AT_name : (indirect string, offset: 0x3f3e):
exact_unsigned_base_helper<sizeof(unsigned int) * 8>
but if compiled with g++ there is no such typename in dwarf
g++ -g -O0 sizeof-space-test.cpp
readelf -wi a.out|grep exact_unsigned_base_helper
Sorry, but I didn't succeed in further reducing mentioned test file.
On Mon, Sep 17, 2012 at 9:26 AM, Jan Kratochvil
<jan.kratochvil@redhat.com> wrote:
> On Mon, 17 Sep 2012 07:49:16 +0200, Vasyl Vavrychuk wrote:
>> This looks to be not a symbol but type name.
>
> OK, reproduced [attached], patched DWARF of "int" in C++: void f(int i) {}
>
>
>> But I have found it directly in the binary. Line containing this text looks like
>
> readelf -wi
> <1><59>: Abbrev Number: 4 (DW_TAG_base_type)
> <5c> DW_AT_name : exact_unsigned_base_helper<sizeof(unsigned int) * 8>
>
>
>> Do you think the problem is in clang? Could you point out some
>> specification, so I can reference it while contacting clang people.
>
> Itanium C++ ABI does not say anything about spacing:
> http://www.swag.uwaterloo.ca/acd/docs/ItaniumC++ABI.htm
> (the original URL is 404 now)
>
> I guess so far it is only agreed upon between GCC and GDB, it is true GDB
> should accept any valid spacing of the language, this is also what
> cp_canonicalize_string is there for.
>
> Your patch would belons to the GCC mailing lists where libiberty/ is
> maintained.
>
> But I find more correct patch the one below.
>
> For a testcase could you provide a minimal C++ code to produce such type?
>
> No regressions on {x86_64,x86_64-m32,i686}-fedora18-linux-gnu.
>
>
> Thanks,
> Jan
>
>
> gdb/
> 2012-09-17 Jan Kratochvil <jan.kratochvil@redhat.com>
>
> Fix internal error on canonicalization of clang types.
> * cp-name-parser.y (exp): Use "sizeof ".
>
> diff --git a/gdb/cp-name-parser.y b/gdb/cp-name-parser.y
> index 6767fa5..c57ce64 100644
> --- a/gdb/cp-name-parser.y
> +++ b/gdb/cp-name-parser.y
> @@ -1183,7 +1183,7 @@ exp : FLOAT
> ;
>
> exp : SIZEOF '(' type ')' %prec UNARY
> - { $$ = d_unary ("sizeof", $3); }
> + { $$ = d_unary ("sizeof ", $3); }
> ;
>
> /* C++. */
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [patch] Fix internal error on canonicalization of clang types (+lib/gdb.exp internal error catch)
2012-09-17 8:29 ` Vasyl Vavrychuk
@ 2012-09-17 15:30 ` Jan Kratochvil
2012-09-17 15:40 ` Doug Evans
0 siblings, 1 reply; 11+ messages in thread
From: Jan Kratochvil @ 2012-09-17 15:30 UTC (permalink / raw)
To: Vasyl Vavrychuk; +Cc: gdb-patches
On Mon, 17 Sep 2012 10:28:52 +0200, Vasyl Vavrychuk wrote:
> Problem happening for the file
>
> #include <cstdlib>
> #include <boost/uuid/uuid_generators.hpp>
> int main(int argc, char* argv[]) {
> boost::uuids::random_generator gen;
> return EXIT_SUCCESS;
> }
>
> clang++ -g -O0 sizeof-space-test.cpp
> readelf -wi a.out|grep exact_unsigned_base_helper
> <4072> DW_AT_name : (indirect string, offset: 0x3f3e):
> exact_unsigned_base_helper<sizeof(unsigned int) * 8>
Not reproducible for me with clang-3.1-9.fc18.x86_64 and
boost-1.50.0-2.fc18.x86_64.
> but if compiled with g++ there is no such typename in dwarf
I believe it belongs into GCC PR
Bug 49130 - discrepancies between DW_AT_name and demangler
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49130
as even GCC puts 'sizeof' into the (mangled) ELF function symbol but it does
not put it into the (demangled) DWARF type name.
So I have at least created a reproducer by hand.
No regressions on {x86_64,x86_64-m32,i686}-fedora18-linux-gnu.
I have also checked the whitespacing of the few other symbols from
libiberty/cp-demangle.c in cp-name-parser.y matches.
Thanks,
Jan
gdb/
2012-09-17 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix internal error on canonicalization of clang types.
* cp-name-parser.y (exp): Use "sizeof ".
gdb/testsuite/
2012-09-17 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix internal error on canonicalization of clang types.
* gdb.dwarf2/dw2-canonicalize-type.S: New file.
* gdb.dwarf2/dw2-canonicalize-type.exp: New file.
* lib/gdb.exp (gdb_file_cmd): Catch also GDB internal error.
diff --git a/gdb/cp-name-parser.y b/gdb/cp-name-parser.y
index 6767fa5..c57ce64 100644
--- a/gdb/cp-name-parser.y
+++ b/gdb/cp-name-parser.y
@@ -1183,7 +1183,7 @@ exp : FLOAT
;
exp : SIZEOF '(' type ')' %prec UNARY
- { $$ = d_unary ("sizeof", $3); }
+ { $$ = d_unary ("sizeof ", $3); }
;
/* C++. */
diff --git a/gdb/testsuite/gdb.dwarf2/dw2-canonicalize-type.S b/gdb/testsuite/gdb.dwarf2/dw2-canonicalize-type.S
new file mode 100644
index 0000000..dbf8f0a
--- /dev/null
+++ b/gdb/testsuite/gdb.dwarf2/dw2-canonicalize-type.S
@@ -0,0 +1,102 @@
+/* Copyright 2012 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/>. */
+
+ .text
+ .globl main
+main: .byte 0
+.Lmain_end:
+ .section .debug_info
+debug_start:
+ .4byte debug_end - 1f /* Length of Compilation Unit Info */
+1:
+ .2byte 0x3 /* DWARF version number */
+ .4byte .Ldebug_abbrev0 /* Offset Into Abbrev. Section */
+ .byte 0x4 /* Pointer Size (in bytes) */
+ .uleb128 0x1 /* (DIE (0xb) DW_TAG_compile_unit) */
+ .ascii "GNU C 4.4.3\0" /* DW_AT_producer */
+ .byte 0x4 /* DW_AT_language = DW_LANG_C_plus_plus */
+ .ascii "1.c\0" /* DW_AT_name */
+ .4byte main /* DW_AT_low_pc */
+ .4byte .Lmain_end /* DW_AT_high_pc */
+ .uleb128 0x4 /* (DIE (0x3c) DW_TAG_subprogram) */
+ .ascii "f\0" /* DW_AT_name */
+/* Value 0 would require has_section_at_zero != 0 (which is true, though). */
+ .4byte main /* DW_AT_low_pc */
+ .4byte .Lmain_end /* DW_AT_high_pc */
+ .byte 0x1 /* DW_AT_prototyped */
+ .uleb128 0x5 /* (DIE (0x42) DW_TAG_formal_parameter) */
+/* Invalid value. */
+ .4byte .Ltype_int - debug_start /* DW_AT_type */
+ .ascii "param\0" /* DW_AT_name */
+ .byte 0x0 /* end of children of DIE 0x3c */
+.Ltype_int:
+ .uleb128 3 /* Abbrev: DW_TAG_base_type */
+ .ascii "typename<sizeof(int)>\0" /* DW_AT_name */
+ .byte 4 /* DW_AT_byte_size */
+ .byte 5 /* DW_AT_encoding */
+ .byte 0x0 /* end of children of DIE 0xb */
+debug_end:
+
+ .section .debug_abbrev
+.Ldebug_abbrev0:
+ .uleb128 0x1 /* (abbrev code) */
+ .uleb128 0x11 /* (TAG: DW_TAG_compile_unit) */
+ .byte 0x1 /* DW_children_yes */
+ .uleb128 0x25 /* (DW_AT_producer) */
+ .uleb128 0x8 /* (DW_FORM_string) */
+ .uleb128 0x13 /* (DW_AT_language) */
+ .uleb128 0xb /* (DW_FORM_data1) */
+ .uleb128 0x3 /* (DW_AT_name) */
+ .uleb128 0x8 /* (DW_FORM_string) */
+ .uleb128 0x11 /* (DW_AT_low_pc) */
+ .uleb128 0x1 /* (DW_FORM_addr) */
+ .uleb128 0x12 /* (DW_AT_high_pc) */
+ .uleb128 0x1 /* (DW_FORM_addr) */
+ .byte 0x0
+ .byte 0x0
+ .uleb128 3 /* Abbrev code */
+ .uleb128 0x24 /* DW_TAG_base_type */
+ .byte 0 /* has_children */
+ .uleb128 0x3 /* DW_AT_name */
+ .uleb128 0x8 /* DW_FORM_string */
+ .uleb128 0xb /* DW_AT_byte_size */
+ .uleb128 0xb /* DW_FORM_data1 */
+ .uleb128 0x3e /* DW_AT_encoding */
+ .uleb128 0xb /* DW_FORM_data1 */
+ .byte 0x0 /* Terminator */
+ .byte 0x0 /* Terminator */
+ .uleb128 0x4 /* (abbrev code) */
+ .uleb128 0x2e /* (TAG: DW_TAG_subprogram) */
+ .byte 0x1 /* DW_children_yes */
+ .uleb128 0x3 /* (DW_AT_name) */
+ .uleb128 0x8 /* (DW_FORM_string) */
+ .uleb128 0x11 /* (DW_AT_low_pc) */
+ .uleb128 0x1 /* (DW_FORM_addr) */
+ .uleb128 0x12 /* (DW_AT_high_pc) */
+ .uleb128 0x1 /* (DW_FORM_addr) */
+ .uleb128 0x27 /* (DW_AT_prototyped) */
+ .uleb128 0xc /* (DW_FORM_flag) */
+ .byte 0x0
+ .byte 0x0
+ .uleb128 0x5 /* (abbrev code) */
+ .uleb128 0x5 /* (TAG: DW_TAG_formal_parameter) */
+ .byte 0x0 /* DW_children_no */
+ .uleb128 0x49 /* (DW_AT_type) */
+ .uleb128 0x13 /* (DW_FORM_ref4) */
+ .uleb128 0x3 /* DW_AT_name */
+ .uleb128 0x8 /* DW_FORM_string */
+ .byte 0x0
+ .byte 0x0
+ .byte 0x0
diff --git a/gdb/testsuite/gdb.dwarf2/dw2-canonicalize-type.exp b/gdb/testsuite/gdb.dwarf2/dw2-canonicalize-type.exp
new file mode 100644
index 0000000..2591bc7
--- /dev/null
+++ b/gdb/testsuite/gdb.dwarf2/dw2-canonicalize-type.exp
@@ -0,0 +1,29 @@
+# Copyright 2012 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/>.
+load_lib dwarf.exp
+
+# This test can only be run on targets which support DWARF-2 and use gas.
+if {![dwarf2_support]} {
+ return 0
+}
+
+standard_testfile .S
+set executable ${testfile}
+
+if [prepare_for_testing $testfile.exp $testfile $srcfile {nodebug}] {
+ return -1
+}
+
+gdb_test "ptype f" {type = void \(typename<sizeof \(int\)>\)}
diff --git a/gdb/testsuite/lib/gdb.exp b/gdb/testsuite/lib/gdb.exp
index 0b4c679..8715c5e 100644
--- a/gdb/testsuite/lib/gdb.exp
+++ b/gdb/testsuite/lib/gdb.exp
@@ -1277,6 +1277,11 @@ proc gdb_file_cmd { arg } {
perror "($arg) No such file or directory"
return -1
}
+ -re "A problem internal to GDB has been detected" {
+ fail "($arg) GDB internal error"
+ gdb_internal_error_resync
+ return -1
+ }
-re "$gdb_prompt $" {
perror "couldn't load $arg into $GDB."
return -1
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [patch] Fix internal error on canonicalization of clang types (+lib/gdb.exp internal error catch)
2012-09-17 15:30 ` [patch] Fix internal error on canonicalization of clang types (+lib/gdb.exp internal error catch) Jan Kratochvil
@ 2012-09-17 15:40 ` Doug Evans
2012-09-17 16:13 ` Jan Kratochvil
0 siblings, 1 reply; 11+ messages in thread
From: Doug Evans @ 2012-09-17 15:40 UTC (permalink / raw)
To: Jan Kratochvil; +Cc: Vasyl Vavrychuk, gdb-patches
On Mon, Sep 17, 2012 at 8:29 AM, Jan Kratochvil
<jan.kratochvil@redhat.com> wrote:
> gdb/
> 2012-09-17 Jan Kratochvil <jan.kratochvil@redhat.com>
>
> Fix internal error on canonicalization of clang types.
> * cp-name-parser.y (exp): Use "sizeof ".
>
>
> diff --git a/gdb/cp-name-parser.y b/gdb/cp-name-parser.y
> index 6767fa5..c57ce64 100644
> --- a/gdb/cp-name-parser.y
> +++ b/gdb/cp-name-parser.y
> @@ -1183,7 +1183,7 @@ exp : FLOAT
> ;
>
> exp : SIZEOF '(' type ')' %prec UNARY
> - { $$ = d_unary ("sizeof", $3); }
> + { $$ = d_unary ("sizeof ", $3); }
> ;
>
> /* C++. */
I had to look hard to see the change here.
Comment explaining "why things are they way they are" is required IMO.
[I'm assuming the above change is in fact TRTTD.]
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [patch] Fix internal error on canonicalization of clang types (+lib/gdb.exp internal error catch)
2012-09-17 15:40 ` Doug Evans
@ 2012-09-17 16:13 ` Jan Kratochvil
2012-09-17 19:52 ` Tom Tromey
0 siblings, 1 reply; 11+ messages in thread
From: Jan Kratochvil @ 2012-09-17 16:13 UTC (permalink / raw)
To: Doug Evans; +Cc: Vasyl Vavrychuk, gdb-patches
On Mon, 17 Sep 2012 17:40:00 +0200, Doug Evans wrote:
> I had to look hard to see the change here.
> Comment explaining "why things are they way they are" is required IMO.
> [I'm assuming the above change is in fact TRTTD.]
That makes sense. Also extended the testcase for other make_operator calls as
removing and adding these spaces is a common patch topic here.
I did not find a regression, I could not get working GDB from 2005 but the bug
seems to be there since that time.
commit c20637db7b29f4886af5f689e00fedfa1dc4ecd8
Author: Daniel Jacobowitz <dan@debian.org>
Date: Fri Mar 11 02:24:23 2005 +0000
Thanks,
Jan
gdb/
2012-09-17 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix internal error on canonicalization of clang types.
* cp-name-parser.y (operator): New comment at make_operator call for
new, delete, new[] and delete[].
* cp-name-parser.y (exp): Use "sizeof ". Add new comment at
make_operator call.
gdb/testsuite/
2012-09-17 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix internal error on canonicalization of clang types.
* gdb.dwarf2/dw2-canonicalize-type.S: New file.
* gdb.dwarf2/dw2-canonicalize-type.exp: New file.
* lib/gdb.exp (gdb_file_cmd): Catch also GDB internal error.
diff --git a/gdb/cp-name-parser.y b/gdb/cp-name-parser.y
index 6767fa5..c218244 100644
--- a/gdb/cp-name-parser.y
+++ b/gdb/cp-name-parser.y
@@ -446,13 +446,29 @@ demangler_special
;
operator : OPERATOR NEW
- { $$ = make_operator ("new", 3); }
+ {
+ /* Match the whitespacing of cplus_demangle_operators.
+ It would abort on unrecognized string otherwise. */
+ $$ = make_operator ("new", 3);
+ }
| OPERATOR DELETE
- { $$ = make_operator ("delete ", 1); }
+ {
+ /* Match the whitespacing of cplus_demangle_operators.
+ It would abort on unrecognized string otherwise. */
+ $$ = make_operator ("delete ", 1);
+ }
| OPERATOR NEW '[' ']'
- { $$ = make_operator ("new[]", 3); }
+ {
+ /* Match the whitespacing of cplus_demangle_operators.
+ It would abort on unrecognized string otherwise. */
+ $$ = make_operator ("new[]", 3);
+ }
| OPERATOR DELETE '[' ']'
- { $$ = make_operator ("delete[] ", 1); }
+ {
+ /* Match the whitespacing of cplus_demangle_operators.
+ It would abort on unrecognized string otherwise. */
+ $$ = make_operator ("delete[] ", 1);
+ }
| OPERATOR '+'
{ $$ = make_operator ("+", 2); }
| OPERATOR '-'
@@ -1183,7 +1199,11 @@ exp : FLOAT
;
exp : SIZEOF '(' type ')' %prec UNARY
- { $$ = d_unary ("sizeof", $3); }
+ {
+ /* Match the whitespacing of cplus_demangle_operators.
+ It would abort on unrecognized string otherwise. */
+ $$ = d_unary ("sizeof ", $3);
+ }
;
/* C++. */
diff --git a/gdb/testsuite/gdb.dwarf2/dw2-canonicalize-type.S b/gdb/testsuite/gdb.dwarf2/dw2-canonicalize-type.S
new file mode 100644
index 0000000..2d1c7c3
--- /dev/null
+++ b/gdb/testsuite/gdb.dwarf2/dw2-canonicalize-type.S
@@ -0,0 +1,146 @@
+/* Copyright 2012 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/>. */
+
+ .text
+ .globl main
+main: .byte 0
+.Lmain_end:
+ .section .debug_info
+debug_start:
+ .4byte debug_end - 1f /* Length of Compilation Unit Info */
+1:
+ .2byte 0x3 /* DWARF version number */
+ .4byte .Ldebug_abbrev0 /* Offset Into Abbrev. Section */
+ .byte 0x4 /* Pointer Size (in bytes) */
+ .uleb128 0x1 /* (DIE (0xb) DW_TAG_compile_unit) */
+ .ascii "GNU C 4.4.3\0" /* DW_AT_producer */
+ .byte 0x4 /* DW_AT_language = DW_LANG_C_plus_plus */
+ .ascii "1.c\0" /* DW_AT_name */
+ .4byte main /* DW_AT_low_pc */
+ .4byte .Lmain_end /* DW_AT_high_pc */
+ .uleb128 0x4 /* (DIE (0x3c) DW_TAG_subprogram) */
+ .ascii "f\0" /* DW_AT_name */
+/* Value 0 would require has_section_at_zero != 0 (which is true, though). */
+ .4byte main /* DW_AT_low_pc */
+ .4byte .Lmain_end /* DW_AT_high_pc */
+ .byte 0x1 /* DW_AT_prototyped */
+
+ .uleb128 0x5 /* (DIE (0x42) DW_TAG_formal_parameter) */
+ .4byte .Ltype_sizeof - debug_start /* DW_AT_type */
+ .ascii "param_sizeof\0" /* DW_AT_name */
+
+ .uleb128 0x5 /* (DIE (0x42) DW_TAG_formal_parameter) */
+ .4byte .Ltype_opnew - debug_start /* DW_AT_type */
+ .ascii "param_opnew\0" /* DW_AT_name */
+
+ .uleb128 0x5 /* (DIE (0x42) DW_TAG_formal_parameter) */
+ .4byte .Ltype_opdelete - debug_start /* DW_AT_type */
+ .ascii "param_opdelete\0" /* DW_AT_name */
+
+ .uleb128 0x5 /* (DIE (0x42) DW_TAG_formal_parameter) */
+ .4byte .Ltype_opnewarray - debug_start /* DW_AT_type */
+ .ascii "param_opnewarray\0" /* DW_AT_name */
+
+ .uleb128 0x5 /* (DIE (0x42) DW_TAG_formal_parameter) */
+ .4byte .Ltype_opdeletearray - debug_start /* DW_AT_type */
+ .ascii "param_opdeletearray\0" /* DW_AT_name */
+ .byte 0x0 /* end of children of DIE 0x3c */
+
+.Ltype_sizeof:
+ .uleb128 3 /* Abbrev: DW_TAG_base_type */
+ .ascii "typename<sizeof(int)>\0" /* DW_AT_name */
+ .byte 4 /* DW_AT_byte_size */
+ .byte 5 /* DW_AT_encoding */
+
+.Ltype_opnew:
+ .uleb128 3 /* Abbrev: DW_TAG_base_type */
+ .ascii "C::operator new(int)\0" /* DW_AT_name */
+ .byte 4 /* DW_AT_byte_size */
+ .byte 5 /* DW_AT_encoding */
+
+.Ltype_opdelete:
+ .uleb128 3 /* Abbrev: DW_TAG_base_type */
+ .ascii "C::operator delete(void *)\0" /* DW_AT_name */
+ .byte 4 /* DW_AT_byte_size */
+ .byte 5 /* DW_AT_encoding */
+
+.Ltype_opnewarray:
+ .uleb128 3 /* Abbrev: DW_TAG_base_type */
+ /* The correct new[] syntax is probably unsupported by GDB. */
+ .ascii "C::operator new[](int)\0" /* DW_AT_name */
+ .byte 4 /* DW_AT_byte_size */
+ .byte 5 /* DW_AT_encoding */
+
+.Ltype_opdeletearray:
+ .uleb128 3 /* Abbrev: DW_TAG_base_type */
+ /* The correct delete[] syntax is probably unsupported by GDB. */
+ .ascii "C::operator delete[](void *)\0" /* DW_AT_name */
+ .byte 4 /* DW_AT_byte_size */
+ .byte 5 /* DW_AT_encoding */
+
+ .byte 0x0 /* end of children of DIE 0xb */
+debug_end:
+
+ .section .debug_abbrev
+.Ldebug_abbrev0:
+ .uleb128 0x1 /* (abbrev code) */
+ .uleb128 0x11 /* (TAG: DW_TAG_compile_unit) */
+ .byte 0x1 /* DW_children_yes */
+ .uleb128 0x25 /* (DW_AT_producer) */
+ .uleb128 0x8 /* (DW_FORM_string) */
+ .uleb128 0x13 /* (DW_AT_language) */
+ .uleb128 0xb /* (DW_FORM_data1) */
+ .uleb128 0x3 /* (DW_AT_name) */
+ .uleb128 0x8 /* (DW_FORM_string) */
+ .uleb128 0x11 /* (DW_AT_low_pc) */
+ .uleb128 0x1 /* (DW_FORM_addr) */
+ .uleb128 0x12 /* (DW_AT_high_pc) */
+ .uleb128 0x1 /* (DW_FORM_addr) */
+ .byte 0x0
+ .byte 0x0
+ .uleb128 3 /* Abbrev code */
+ .uleb128 0x24 /* DW_TAG_base_type */
+ .byte 0 /* has_children */
+ .uleb128 0x3 /* DW_AT_name */
+ .uleb128 0x8 /* DW_FORM_string */
+ .uleb128 0xb /* DW_AT_byte_size */
+ .uleb128 0xb /* DW_FORM_data1 */
+ .uleb128 0x3e /* DW_AT_encoding */
+ .uleb128 0xb /* DW_FORM_data1 */
+ .byte 0x0 /* Terminator */
+ .byte 0x0 /* Terminator */
+ .uleb128 0x4 /* (abbrev code) */
+ .uleb128 0x2e /* (TAG: DW_TAG_subprogram) */
+ .byte 0x1 /* DW_children_yes */
+ .uleb128 0x3 /* (DW_AT_name) */
+ .uleb128 0x8 /* (DW_FORM_string) */
+ .uleb128 0x11 /* (DW_AT_low_pc) */
+ .uleb128 0x1 /* (DW_FORM_addr) */
+ .uleb128 0x12 /* (DW_AT_high_pc) */
+ .uleb128 0x1 /* (DW_FORM_addr) */
+ .uleb128 0x27 /* (DW_AT_prototyped) */
+ .uleb128 0xc /* (DW_FORM_flag) */
+ .byte 0x0
+ .byte 0x0
+ .uleb128 0x5 /* (abbrev code) */
+ .uleb128 0x5 /* (TAG: DW_TAG_formal_parameter) */
+ .byte 0x0 /* DW_children_no */
+ .uleb128 0x49 /* (DW_AT_type) */
+ .uleb128 0x13 /* (DW_FORM_ref4) */
+ .uleb128 0x3 /* DW_AT_name */
+ .uleb128 0x8 /* DW_FORM_string */
+ .byte 0x0
+ .byte 0x0
+ .byte 0x0
diff --git a/gdb/testsuite/gdb.dwarf2/dw2-canonicalize-type.exp b/gdb/testsuite/gdb.dwarf2/dw2-canonicalize-type.exp
new file mode 100644
index 0000000..91d718e
--- /dev/null
+++ b/gdb/testsuite/gdb.dwarf2/dw2-canonicalize-type.exp
@@ -0,0 +1,30 @@
+# Copyright 2012 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/>.
+load_lib dwarf.exp
+
+# This test can only be run on targets which support DWARF-2 and use gas.
+if {![dwarf2_support]} {
+ return 0
+}
+
+standard_testfile .S
+set executable ${testfile}
+
+if [prepare_for_testing $testfile.exp $testfile $srcfile {nodebug}] {
+ return -1
+}
+
+# The correct new[] and delete[] syntax is probably unsupported by GDB.
+gdb_test "ptype f" {type = void \(typename<sizeof \(int\)>, C::operator new\(int\), C::operator delete\(void\*\), C::operator new\[\]\(int\), C::operator delete\[\]\(void\*\)\)}
diff --git a/gdb/testsuite/lib/gdb.exp b/gdb/testsuite/lib/gdb.exp
index 0b4c679..8715c5e 100644
--- a/gdb/testsuite/lib/gdb.exp
+++ b/gdb/testsuite/lib/gdb.exp
@@ -1277,6 +1277,11 @@ proc gdb_file_cmd { arg } {
perror "($arg) No such file or directory"
return -1
}
+ -re "A problem internal to GDB has been detected" {
+ fail "($arg) GDB internal error"
+ gdb_internal_error_resync
+ return -1
+ }
-re "$gdb_prompt $" {
perror "couldn't load $arg into $GDB."
return -1
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [patch] Fix internal error on canonicalization of clang types (+lib/gdb.exp internal error catch)
2012-09-17 16:13 ` Jan Kratochvil
@ 2012-09-17 19:52 ` Tom Tromey
2012-09-21 19:36 ` Jan Kratochvil
0 siblings, 1 reply; 11+ messages in thread
From: Tom Tromey @ 2012-09-17 19:52 UTC (permalink / raw)
To: Jan Kratochvil; +Cc: Doug Evans, Vasyl Vavrychuk, gdb-patches
>>>>> "Jan" == Jan Kratochvil <jan.kratochvil@redhat.com> writes:
Jan> | OPERATOR DELETE
Jan> - { $$ = make_operator ("delete ", 1); }
Jan> + {
Jan> + /* Match the whitespacing of cplus_demangle_operators.
Jan> + It would abort on unrecognized string otherwise. */
The indentation here (and elsewhere) seems funny.
Jan> +set executable ${testfile}
This line shouldn't be needed.
I think the rest is fine though.
Tom
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [patch] Fix internal error on canonicalization of clang types (+lib/gdb.exp internal error catch)
2012-09-17 19:52 ` Tom Tromey
@ 2012-09-21 19:36 ` Jan Kratochvil
2012-09-21 19:40 ` Tom Tromey
0 siblings, 1 reply; 11+ messages in thread
From: Jan Kratochvil @ 2012-09-21 19:36 UTC (permalink / raw)
To: Tom Tromey; +Cc: Doug Evans, Vasyl Vavrychuk, gdb-patches
On Mon, 17 Sep 2012 21:52:08 +0200, Tom Tromey wrote:
> >>>>> "Jan" == Jan Kratochvil <jan.kratochvil@redhat.com> writes:
>
> Jan> | OPERATOR DELETE
> Jan> - { $$ = make_operator ("delete ", 1); }
> Jan> + {
> Jan> + /* Match the whitespacing of cplus_demangle_operators.
> Jan> + It would abort on unrecognized string otherwise. */
>
> The indentation here (and elsewhere) seems funny.
Sorry but I do not understand if it was a request for a change (and which
change) or not.
Thanks,
Jan
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [patch] Fix internal error on canonicalization of clang types (+lib/gdb.exp internal error catch)
2012-09-21 19:36 ` Jan Kratochvil
@ 2012-09-21 19:40 ` Tom Tromey
2012-09-21 19:52 ` Jan Kratochvil
0 siblings, 1 reply; 11+ messages in thread
From: Tom Tromey @ 2012-09-21 19:40 UTC (permalink / raw)
To: Jan Kratochvil; +Cc: Doug Evans, Vasyl Vavrychuk, gdb-patches
>>>>> "Jan" == Jan Kratochvil <jan.kratochvil@redhat.com> writes:
Jan> - { $$ = make_operator ("delete ", 1); }
Jan> + {
Jan> + /* Match the whitespacing of cplus_demangle_operators.
Jan> + It would abort on unrecognized string otherwise. */
>>
>> The indentation here (and elsewhere) seems funny.
Jan> Sorry but I do not understand if it was a request for a change (and which
Jan> change) or not.
At least from the patch it appears that the comment is over by the
leftmost column, but I think the comment should be indented like the
code is.
Tom
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [patch] Fix internal error on canonicalization of clang types (+lib/gdb.exp internal error catch)
2012-09-21 19:40 ` Tom Tromey
@ 2012-09-21 19:52 ` Jan Kratochvil
2012-09-21 19:57 ` Tom Tromey
0 siblings, 1 reply; 11+ messages in thread
From: Jan Kratochvil @ 2012-09-21 19:52 UTC (permalink / raw)
To: Tom Tromey; +Cc: Doug Evans, Vasyl Vavrychuk, gdb-patches
On Fri, 21 Sep 2012 21:39:54 +0200, Tom Tromey wrote:
> At least from the patch it appears that the comment is over by the
> leftmost column, but I think the comment should be indented like the
> code is.
This is after patch from the mail, grep -3 xxx | expand, inserted into this
mail.
I see it has 81 columns but with the large tabs there this file has very
common violation of > 80 columns; I can fix this patch, though.
Thanks,
Jan
------------------------------------------------------------------------------
operator : OPERATOR NEW
{
/* Match the whitespacing of cplus_demangle_operators.
It would abort on unrecognized string otherwise. */
$$ = make_operator ("new", 3);
}
| OPERATOR DELETE
{
/* Match the whitespacing of cplus_demangle_operators.
It would abort on unrecognized string otherwise. */
$$ = make_operator ("delete ", 1);
}
| OPERATOR NEW '[' ']'
{
/* Match the whitespacing of cplus_demangle_operators.
It would abort on unrecognized string otherwise. */
$$ = make_operator ("new[]", 3);
}
| OPERATOR DELETE '[' ']'
{
/* Match the whitespacing of cplus_demangle_operators.
It would abort on unrecognized string otherwise. */
$$ = make_operator ("delete[] ", 1);
}
--
exp : SIZEOF '(' type ')' %prec UNARY
{
/* Match the whitespacing of cplus_demangle_operators.
It would abort on unrecognized string otherwise. */
$$ = d_unary ("sizeof ", $3);
}
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [patch] Fix internal error on canonicalization of clang types (+lib/gdb.exp internal error catch)
2012-09-21 19:52 ` Jan Kratochvil
@ 2012-09-21 19:57 ` Tom Tromey
2012-09-21 20:01 ` [commit] " Jan Kratochvil
0 siblings, 1 reply; 11+ messages in thread
From: Tom Tromey @ 2012-09-21 19:57 UTC (permalink / raw)
To: Jan Kratochvil; +Cc: Doug Evans, Vasyl Vavrychuk, gdb-patches
Jan> This is after patch from the mail, grep -3 xxx | expand, inserted into this
Jan> mail.
Thanks, it looks ok now.
Tom
^ permalink raw reply [flat|nested] 11+ messages in thread
* [commit] [patch] Fix internal error on canonicalization of clang types (+lib/gdb.exp internal error catch)
2012-09-21 19:57 ` Tom Tromey
@ 2012-09-21 20:01 ` Jan Kratochvil
0 siblings, 0 replies; 11+ messages in thread
From: Jan Kratochvil @ 2012-09-21 20:01 UTC (permalink / raw)
To: Tom Tromey; +Cc: Doug Evans, Vasyl Vavrychuk, gdb-patches
On Fri, 21 Sep 2012 21:56:56 +0200, Tom Tromey wrote:
> Jan> This is after patch from the mail, grep -3 xxx | expand, inserted into this
> Jan> mail.
>
> Thanks, it looks ok now.
Checked in:
http://sourceware.org/ml/gdb-cvs/2012-09/msg00124.html
Thanks,
Jan
^ permalink raw reply [flat|nested] 11+ messages in thread
end of thread, other threads:[~2012-09-21 20:01 UTC | newest]
Thread overview: 11+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
[not found] <CAGj4m+4au-jiC2ueSyC+wEY76YRQ5zNuxfY+-mfZ39N4fnpN_g@mail.gmail.com>
[not found] ` <20120916080035.GA28890@host2.jankratochvil.net>
[not found] ` <CAGj4m+46oBXSYQ-wx9g-P7yevZeMDnNNEjvADE8sRSdoWaRiVA@mail.gmail.com>
2012-09-17 6:27 ` [patch] Fix internal error on canonicalization of clang types [Re: why in cplus_demangle_operators operation sizeof is expected to be followed with space] Jan Kratochvil
2012-09-17 8:29 ` Vasyl Vavrychuk
2012-09-17 15:30 ` [patch] Fix internal error on canonicalization of clang types (+lib/gdb.exp internal error catch) Jan Kratochvil
2012-09-17 15:40 ` Doug Evans
2012-09-17 16:13 ` Jan Kratochvil
2012-09-17 19:52 ` Tom Tromey
2012-09-21 19:36 ` Jan Kratochvil
2012-09-21 19:40 ` Tom Tromey
2012-09-21 19:52 ` Jan Kratochvil
2012-09-21 19:57 ` Tom Tromey
2012-09-21 20:01 ` [commit] " Jan Kratochvil
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox