binutils-gdb/gdbsupport/packed.h

169 lines
5.2 KiB
C
Raw Normal View History

Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
/* Copyright (C) 2022 Free Software Foundation, Inc.
This file is part of GDB.
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/>. */
#ifndef PACKED_H
#define PACKED_H
#include "traits.h"
#include <atomic>
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
/* Each instantiation and full specialization of the packed template
defines a type that behaves like a given scalar type, but that has
byte alignment, and, may optionally have a smaller size than the
given scalar type. This is typically used as alternative to
bit-fields (and ENUM_BITFIELD), when the fields must have separate
memory locations to avoid data races. */
/* There are two implementations here -- one standard compliant, using
a byte array for internal representation, and another that relies
on bitfields and attribute packed (and attribute gcc_struct on
Windows). The latter is preferable, as it is more convenient when
debugging GDB -- printing a struct packed variable prints its field
using its natural type, which is particularly useful if the type is
an enum -- but may not work on all compilers. */
/* Clang targeting Windows does not support attribute gcc_struct, so
we use the alternative byte array implemention there. */
#if defined _WIN32 && defined __clang__
# define PACKED_USE_ARRAY 1
#else
# define PACKED_USE_ARRAY 0
#endif
/* For the preferred implementation, we need gcc_struct on Windows, as
otherwise the size of e.g., "packed<int, 1>" will be larger than
what we want. Clang targeting Windows does not support attribute
gcc_struct. */
#if !PACKED_USE_ARRAY && defined _WIN32 && !defined __clang__
struct packed: Use gcc_struct on Windows Building GDB on mingw/gcc hosts is currently broken, due to a static assertion failure in gdbsupport/packed.h: In file included from ../../../../../binutils-gdb/gdb/../gdbsupport/common-defs.h:201, from ../../../../../binutils-gdb/gdb/defs.h:28, from ../../../../../binutils-gdb/gdb/dwarf2/read.c:31: ../../../../../binutils-gdb/gdb/../gdbsupport/packed.h: In instantiation of 'packed<T, Bytes>::packed(T) [with T = dwarf_unit_type; long long unsigned int Bytes = 1]': ../../../../../binutils-gdb/gdb/dwarf2/read.h:181:74: required from here ../../../../../binutils-gdb/gdb/../gdbsupport/packed.h:41:40: error: static assertion failed 41 | gdb_static_assert (sizeof (packed) == Bytes); | ~~~~~~~~~~~~~~~~^~~~~~~~ ../../../../../binutils-gdb/gdb/../gdbsupport/gdb_assert.h:27:48: note: in definition of macro 'gdb_static_assert' 27 | #define gdb_static_assert(expr) static_assert (expr, "") | ^~~~ ../../../../../binutils-gdb/gdb/../gdbsupport/packed.h:41:40: note: the comparison reduces to '(4 == 1)' 41 | gdb_static_assert (sizeof (packed) == Bytes); | ~~~~~~~~~~~~~~~~^~~~~~~~ The issue is that mingw gcc defaults to "-mms-bitfields", which affects how bitfields are laid out. We can however tell GCC that we want the regular GCC layout instead using attribute gcc_struct. Attribute gcc_struct is not implemented in "clang -target x86_64-pc-windows-gnu", so that will need a different fix. Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=29373 Change-Id: I023315ee03622c59c397bf4affc0b68179c32374
2022-07-19 07:26:33 +08:00
# define ATTRIBUTE_GCC_STRUCT __attribute__((__gcc_struct__))
#else
# define ATTRIBUTE_GCC_STRUCT
#endif
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
template<typename T, size_t Bytes = sizeof (T)>
struct packed: Use gcc_struct on Windows Building GDB on mingw/gcc hosts is currently broken, due to a static assertion failure in gdbsupport/packed.h: In file included from ../../../../../binutils-gdb/gdb/../gdbsupport/common-defs.h:201, from ../../../../../binutils-gdb/gdb/defs.h:28, from ../../../../../binutils-gdb/gdb/dwarf2/read.c:31: ../../../../../binutils-gdb/gdb/../gdbsupport/packed.h: In instantiation of 'packed<T, Bytes>::packed(T) [with T = dwarf_unit_type; long long unsigned int Bytes = 1]': ../../../../../binutils-gdb/gdb/dwarf2/read.h:181:74: required from here ../../../../../binutils-gdb/gdb/../gdbsupport/packed.h:41:40: error: static assertion failed 41 | gdb_static_assert (sizeof (packed) == Bytes); | ~~~~~~~~~~~~~~~~^~~~~~~~ ../../../../../binutils-gdb/gdb/../gdbsupport/gdb_assert.h:27:48: note: in definition of macro 'gdb_static_assert' 27 | #define gdb_static_assert(expr) static_assert (expr, "") | ^~~~ ../../../../../binutils-gdb/gdb/../gdbsupport/packed.h:41:40: note: the comparison reduces to '(4 == 1)' 41 | gdb_static_assert (sizeof (packed) == Bytes); | ~~~~~~~~~~~~~~~~^~~~~~~~ The issue is that mingw gcc defaults to "-mms-bitfields", which affects how bitfields are laid out. We can however tell GCC that we want the regular GCC layout instead using attribute gcc_struct. Attribute gcc_struct is not implemented in "clang -target x86_64-pc-windows-gnu", so that will need a different fix. Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=29373 Change-Id: I023315ee03622c59c397bf4affc0b68179c32374
2022-07-19 07:26:33 +08:00
struct ATTRIBUTE_GCC_STRUCT packed
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
{
public:
packed () noexcept = default;
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
packed (T val)
{
gdb_static_assert (sizeof (ULONGEST) >= sizeof (T));
#if PACKED_USE_ARRAY
ULONGEST tmp = val;
for (int i = (Bytes - 1); i >= 0; --i)
{
m_bytes[i] = (gdb_byte) tmp;
tmp >>= HOST_CHAR_BIT;
}
#else
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
m_val = val;
#endif
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
/* Ensure size and aligment are what we expect. */
gdb_static_assert (sizeof (packed) == Bytes);
gdb_static_assert (alignof (packed) == 1);
/* Make sure packed can be wrapped with std::atomic. */
#if HAVE_IS_TRIVIALLY_COPYABLE
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
gdb_static_assert (std::is_trivially_copyable<packed>::value);
#endif
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
gdb_static_assert (std::is_copy_constructible<packed>::value);
gdb_static_assert (std::is_move_constructible<packed>::value);
gdb_static_assert (std::is_copy_assignable<packed>::value);
gdb_static_assert (std::is_move_assignable<packed>::value);
}
operator T () const noexcept
{
#if PACKED_USE_ARRAY
ULONGEST tmp = 0;
for (int i = 0;;)
{
tmp |= m_bytes[i];
if (++i == Bytes)
break;
tmp <<= HOST_CHAR_BIT;
}
return (T) tmp;
#else
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
return m_val;
#endif
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
}
private:
#if PACKED_USE_ARRAY
gdb_byte m_bytes[Bytes];
#else
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
T m_val : (Bytes * HOST_CHAR_BIT) ATTRIBUTE_PACKED;
#endif
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
};
/* Add some comparisons between std::atomic<packed<T>> and packed<T>
and T. We need this because even though std::atomic<T> doesn't
define these operators, the relational expressions still work via
implicit conversions. Those wouldn't work when wrapped in packed
without these operators, because they'd require two implicit
conversions to go from T to packed<T> to std::atomic<packed<T>>
(and back), and C++ only does one. */
#define PACKED_ATOMIC_OP(OP) \
template<typename T, size_t Bytes> \
bool operator OP (const std::atomic<packed<T, Bytes>> &lhs, \
const std::atomic<packed<T, Bytes>> &rhs) \
{ \
return lhs.load () OP rhs.load (); \
} \
\
template<typename T, size_t Bytes> \
bool operator OP (T lhs, const std::atomic<packed<T, Bytes>> &rhs) \
{ \
return lhs OP rhs.load (); \
} \
\
template<typename T, size_t Bytes> \
bool operator OP (const std::atomic<packed<T, Bytes>> &lhs, T rhs) \
{ \
return lhs.load () OP rhs; \
} \
\
template<typename T, size_t Bytes> \
bool operator OP (const std::atomic<packed<T, Bytes>> &lhs, \
packed<T, Bytes> rhs) \
{ \
return lhs.load () OP rhs; \
} \
\
template<typename T, size_t Bytes> \
bool operator OP (packed<T, Bytes> lhs, \
const std::atomic<packed<T, Bytes>> &rhs) \
{ \
return lhs OP rhs.load (); \
}
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
PACKED_ATOMIC_OP (==)
PACKED_ATOMIC_OP (!=)
PACKED_ATOMIC_OP (>)
PACKED_ATOMIC_OP (<)
PACKED_ATOMIC_OP (>=)
PACKED_ATOMIC_OP (<=)
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
#undef PACKED_ATOMIC_OP
Introduce struct packed template When building gdb with -fsanitize=thread and gcc 12, and running test-case gdb.dwarf2/dwz.exp, we run into a few data races. For example, between: ... Write of size 1 at 0x7b200000300e by thread T4: #0 process_psymtab_comp_unit gdb/dwarf2/read.c:6789 (gdb+0x830720) ... and: ... Previous read of size 1 at 0x7b200000300e by main thread: #0 cutu_reader::cutu_reader(dwarf2_per_cu_data*, dwarf2_per_objfile*, \ abbrev_table*, dwarf2_cu*, bool, abbrev_cache*) gdb/dwarf2/read.c:6164 \ (gdb+0x82edab) ... In other words, between: ... this_cu->unit_type = DW_UT_partial; ... and: ... if (this_cu->reading_dwo_directly) ... The problem is that the written fields are part of the same memory location as the read fields, so executing a read and write in different threads is undefined behavour. Making the written fields separate memory locations, like this: ... struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8; }; ... fixes it, however that also increases the size of struct dwarf2_per_cu_data, because it introduces padding due to alignment of these new structs, which align on the natural alignment of the specified type of their fields. We can fix that with __attribute__((packed)), like so: struct { ENUM_BITFIELD (dwarf_unit_type) unit_type : 8 __attribute__((packed)); }; but to avoid having to write that in several places and add suitable comments explaining how that concoction works, introduce a new struct packed template that wraps/hides this. Instead of the above, we'll be able to write: packed<dwarf_unit_type, 1> unit_type; Note that we can't change the type of dwarf_unit_type, as that is defined in include/, and shared with other projects, some of those written in C. This patch just adds the struct packed type. Following patches will make use of it. One of those patches will want to wrap a struct packed in an std::atomic, like: std::atomic<std::packed<language, 1>> m_lang; so the new gdbsupport/packed.h header adds some operators to make comparisions between that std::atomic and the type that the wrapped struct packed wraps work, like in: if (m_lang == language_c) It would be possible to implement struct packed without using __attribute__((packed)), by having it store an array of bytes of the appropriate size instead, however that would make it less convenient to debug GDB. The way it's implemented, printing a struct packed variable just prints its field using its natural type, which is particularly useful if the type is an enum. I believe that __attribute__((packed)) is supported by all compilers that are able to build GDB. Even a few BFD headers use on ATTRIBUTE_PACKED on external types: include/coff/external.h: } ATTRIBUTE_PACKED include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/external.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED ; include/coff/pe.h:} ATTRIBUTE_PACKED; include/elf/external.h:} ATTRIBUTE_PACKED Elf_External_Versym; It is not possible to build GDB with MSVC today, but if it could, that would be one compiler that doesn't support this attribute. However, it supports packing via pragmas, so there's a way to cross that bridge if we ever get to it. I believe any compiler worth its salt supports some way of packing. In any case, the worse that happens without the attribute is that some types become larger than ideal. Regardless, I've added a couple static assertions to catch such compilers in action: /* Ensure size and aligment are what we expect. */ gdb_static_assert (sizeof (packed) == Bytes); gdb_static_assert (alignof (packed) == 1); Change-Id: Ifa94f0a2cebfae5e8f6ddc73265f05e7fd9e1532
2022-07-12 18:20:13 +08:00
#endif