Pedro Alves [Wed, 9 Aug 2017 14:04:32 +0000 (15:04 +0100)]
Make cp_comp_to_string return a gdb::unique_xmalloc_ptr<char>
To help avoid issues like the one fixed by
e88e8651cf34 ("Fix memory
leak in cp-support.c").
gdb/ChangeLog:
2017-08-09 Pedro Alves <palves@redhat.com>
* cp-name-parser.y (cp_comp_to_string): Return a
gdb::unique_xmalloc_ptr<char>.
* cp-support.c (replace_typedefs_qualified_name)
(replace_typedefs): Adjust to use gdb::unique_xmalloc_ptr<char>.
(cp_canonicalize_string_full): Use op= instead of explicit
convertion.
(cp_class_name_from_physname, method_name_from_physname)
(cp_func_name, cp_remove_params): Adjust to use
gdb::unique_xmalloc_ptr<char>.
* cp-support.h (cp_comp_to_string): Return a
gdb::unique_xmalloc_ptr<char>.
* python/py-type.c (typy_lookup_type): Adjust to use
gdb::unique_xmalloc_ptr<char>.
H.J. Lu [Wed, 9 Aug 2017 12:01:55 +0000 (05:01 -0700)]
gdb: Fix build failure with GCC 7
Fix:
/export/gnu/import/git/sources/binutils-gdb/gdb/dwarf2read.c: In function ‘const char* dwarf2_string_attr(die_info*, unsigned int, dwarf2_cu*)’:
/export/gnu/import/git/sources/binutils-gdb/gdb/dwarf2read.c:17626:39: error: enum constant in boolean context [-Werror=int-in-bool-context]
|| attr->form == DW_FORM_string || DW_FORM_GNU_str_index
* dwarf2read.c (dwarf2_string_attr): Fix a typo.
Yao Qi [Wed, 9 Aug 2017 11:39:16 +0000 (12:39 +0100)]
Fix memory leak in cp-support.c
The return value of cp_comp_to_string was never freed, creating a
sizable memory leak detectable with valgrind.
==21225== 8 bytes in 1 blocks are definitely lost in loss record 4,599 of 10,949^M
==21225== at 0x4C2DB8F: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)^M
==21225== by 0x4C2FDEF: realloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)^M
==21225== by 0x76CB31: d_growable_string_resize (cp-demangle.c:3963)^M
==21225== by 0x76CB31: d_growable_string_init (cp-demangle.c:3942)^M
==21225== by 0x76CB31: cplus_demangle_print (cp-demangle.c:4308)^M
==21225== by 0x4C9535: cp_comp_to_string(demangle_component*, int) (cp-name-parser.y:1972)^M
==21225== by 0x53E1D4: cp_canonicalize_string_full[abi:cxx11](char const*, char const* (*)(type*, void*), void*) (cp-support.c:530)^M
==21225== by 0x53E360: cp_canonicalize_string_no_typedefs[abi:cxx11](char const*) (cp-support.c:548)^M
==21225== by 0x5D51D2: find_linespec_symbols(linespec_state*, VEC_symtab_ptr*, char const*, VEC_symbolp**, VEC_bound_minimal_symbol_d**) (linespec.c:4030)^M
==21225== by 0x5D6CF6: linespec_parse_basic (linespec.c:1907)
==21279== 32 bytes in 1 blocks are definitely lost in loss record 6,066 of 10,947^M
==21279== at 0x4C2DB8F: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)^M
==21279== by 0x4C2FDEF: realloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)^M
==21279== by 0x76CB31: d_growable_string_resize (cp-demangle.c:3963)^M
==21279== by 0x76CB31: d_growable_string_init (cp-demangle.c:3942)^M
==21279== by 0x76CB31: cplus_demangle_print (cp-demangle.c:4308)^M
==21279== by 0x4C9535: cp_comp_to_string(demangle_component*, int) (cp-name-parser.y:1972)^M
==21279== by 0x53EF14: cp_canonicalize_string[abi:cxx11](char const*) (cp-support.c:569)^M
==21279== by 0x561B75: dwarf2_canonicalize_name(char const*, dwarf2_cu*, obstack*) [clone .isra.210] (dwarf2read.c:20159)
This patch fixes the leak. It is a regression by
2f408ecb.
gdb:
2017-08-09 Alex Lindsay <alexlindsay239@gmail.com>
Yao Qi <yao.qi@linaro.org>
* cp-support.c (cp_canonicalize_string_full): Use
gdb::unique_xmalloc_ptr<char>.
(cp_canonicalize_string): Likewise.
Yao Qi [Wed, 9 Aug 2017 11:27:24 +0000 (12:27 +0100)]
Clean up x86 non-linux GDBserver target descriptions
In GDBserver, only tdesc_i386 and tdesc_amd64 are used. There is no point
of generating these *.dat files (which are used to generate *.c files during
GDBserver build.).
gdb:
2017-08-09 Yao Qi <yao.qi@linaro.org>
* features/Makefile (WHICH): Remove i386/ non-linux stuff.
* regformats/i386/amd64-avx-avx512.dat: Remove.
* regformats/i386/amd64-avx-mpx-avx512-pku.dat: Remove.
* regformats/i386/amd64-avx-mpx.dat:Remove.
* regformats/i386/amd64-avx.dat: Remove.
* regformats/i386/amd64-mpx.dat: Remove.
* regformats/i386/i386-avx-avx512.dat: Remove.
* regformats/i386/i386-avx-mpx-avx512-pku.dat: Remove.
* regformats/i386/i386-avx-mpx.dat: Remove.
* regformats/i386/i386-mmx.dat: Remove.
* regformats/i386/i386-mpx.dat: Remove.
gdb/gdbserver:
2017-08-09 Yao Qi <yao.qi@linaro.org>
* configure.srv (srv_i386_regobj): Remove i386-avx.o,
i386-avx-avx512.o, i386-avx-mpx-avx512-pku.o, i386-mpx.o,
i386-avx-mpx.o and i386-mmx.o.
(srv_amd64_regobj): Remove amd64-avx.o, amd64-avx-avx512.o,
amd64-avx-mpx-avx512-pku.o, amd64-mpx.o and amd64-avx-mpx.o.
(srv_i386_xmlfiles): Remove i386/i386-avx.xml,
i386/i386-avx-avx512.xml, i386/i386-avx-mpx-avx512-pku.xml,
i386/i386-mpx.xml, i386/i386-avx-mpx.xml and i386/i386-mmx.xml.
(srv_amd64_xmlfile):i386/amd64-avx.xml, i386/amd64-avx-avx512.xml,
i386/amd64-avx-mpx-avx512-pku.xml, i386/amd64-mpx.xml,
i386/amd64-avx-mpx.xml.
Yao Qi [Wed, 9 Aug 2017 11:27:23 +0000 (12:27 +0100)]
Remove x32 non-linux target descriptions
x32 non-linux target descriptions are not used in GDB or GDBserver. This
patch removes them.
gdb:
2017-08-09 Yao Qi <yao.qi@linaro.org>
* amd64-tdep.h (tdesc_x32): Remove the declaration.
* amd64-tdep.c: Don't include features/i386/x32*.c.
(_initialize_amd64_tdep): Don't call initialize_tdesc_x32*
functions.
* features/Makefile (WHICH): Remove i386/x32, i386/x32-avx,
and i386/x32-avx-avx512.
(XMLTOC): Remove i386/x32-avx.xml, i386/x32-avx-avx512.xml,
and i386/x32.xml.
* features/i386/x32-avx-avx512.c: Removed.
* features/i386/x32-avx-avx512.xml: Removed.
* features/i386/x32-avx.c: Removed.
* features/i386/x32-avx.xml: Removed.
* features/i386/x32.c: Removed.
* features/i386/x32.xml: Removed.
* regformats/i386/x32-avx-avx512.dat: Removed.
* regformats/i386/x32-avx.dat: Removed.
* regformats/i386/x32.dat: Removed.
gdb/gdbserver:
2017-08-09 Yao Qi <yao.qi@linaro.org>
* configure.srv (srv_amd64_regobj): Remove x32.o, x32-avx.o
and x32-avx-avx512.o.
(srv_amd64_xmlfiles): Remove i386/x32.xml, i386/x32-avx.xml
i386/x32-avx-avx512.xml.
H.J. Lu [Wed, 9 Aug 2017 03:20:43 +0000 (20:20 -0700)]
ld: Require GCC 5 or above for 3 x86 tests
Require GCC 5 or above for 3 x86 tests which fail with GCC 4.9.3.
PR ld/21924
* testsuite/ld-i386/i386.exp: Require GCC 5 or above for
"weakundef1 with PIE" test.
* testsuite/ld-x86-64/tls.exp: Require GCC 5 or above for
"tlsdesc1" and "tlsdesc1 with PIE" tests.
H.J. Lu [Wed, 9 Aug 2017 03:01:28 +0000 (20:01 -0700)]
ld: Add a ',' after -Wl,--rpath
* configure.ac (TESTBFDLIB): Add a ',' after -Wl,--rpath.
* configure: Regenerated.
GDB Administrator [Wed, 9 Aug 2017 00:00:58 +0000 (00:00 +0000)]
Automatic date update in version.in
H.J. Lu [Tue, 8 Aug 2017 21:13:46 +0000 (14:13 -0700)]
ld: Replace --rpath with -Wl,--rpath
"--rpath ../bfd/.libs ../bfd/.libs/libbfd.so" isn't a valid compiler
command-line option. We should use -Wl,--rpath instead of --rpath.
PR ld/21923
* configure.ac (TESTBFDLIB): Replace --rpath with -Wl,--rpath
for --disable-static.
* configure: Regenerated.
Nick Clifton [Tue, 8 Aug 2017 12:20:02 +0000 (13:20 +0100)]
Fix address violation problems when parsing corrupt ELF binaries.
PR 21916
* elf-attrs.c (_bfd_elf_parse_attributes): Complain about very
small section lengths.
* elf.c (_bfd_elf_setup_sections): Skip empty entries in the group
table.
(elfcore_grok_freebsd_prstatus): Add checks to make sure that
there is enough data present in the note.
Nick Clifton [Tue, 8 Aug 2017 10:59:32 +0000 (11:59 +0100)]
Updated Serbian translation for gprof
Nick Clifton [Tue, 8 Aug 2017 10:57:22 +0000 (11:57 +0100)]
Fix address violation bugs when writing beyond the end of a local string buffer.
PR 21909
* prdbg.c (pr_int_type): Increase size of local string buffer.
(pr_float_type): Likewise.
(pr_bool_type): Likewise.
Alan Modra [Tue, 8 Aug 2017 04:35:33 +0000 (14:05 +0930)]
PR21017, microblaze missing _GLOBAL_OFFSET_TABLE_ symbol
PR 21017
* elf32-microblaze.c (microblaze_elf_check_relocs): Don't bump
got.refcount for GOTOFF relocs, just create .got section.
GDB Administrator [Tue, 8 Aug 2017 00:00:54 +0000 (00:00 +0000)]
Automatic date update in version.in
Romain Geissler [Mon, 7 Aug 2017 21:55:39 +0000 (07:25 +0930)]
Add configure flag to enable gnu hash style by default.
ld/
* configure.ac: Add --enable-default-hash-style option.
* ldmain.c (main): Set link_info.emit_hash to DEFAULT_EMIT_SYSV_HASH.
Set link_info.emit_gnu_hash to DEFAULT_EMIT_GNU_HASH.
* configure: Regenerate.
* config.in: Regenerate.
gold/
* configure.ac: Add --enable-default-hash-style option.
* options.h (hash_style): Use DEFAULT_HASH_STYLE as default value.
* configure: Regenerate.
* config.in: Regenerate.
Alan Modra [Mon, 7 Aug 2017 21:45:29 +0000 (07:15 +0930)]
More support for ld --hash-style in the ld testsuite
These were discovered when running --hash-style=gnu as default, the
previous batch being --hash-style=both.
* testsuite/ld-aarch64/ifunc-1-local.d: Run ld with --hash-style=sysv.
* testsuite/ld-aarch64/ifunc-2-local.d: Likewise.
* testsuite/ld-aarch64/ifunc-3a.d: Likewise.
* testsuite/ld-frv/fdpic-pie-1.d: Likewise.
* testsuite/ld-frv/fdpic-pie-2.d: Likewise.
* testsuite/ld-frv/fdpic-pie-7.d: Likewise.
* testsuite/ld-frv/fdpic-pie-8.d: Likewise.
* testsuite/ld-arm/arm-elf.exp: Add --hash-style=sysv to "Using
Thumb lib by another lib" test's ld options.
* testsuite/ld-elf/note-3.l: Match .gnu.hash.
* testsuite/ld-elf/note-3.t: Add .gnu.hash output section.
Simon Marchi [Mon, 7 Aug 2017 16:25:59 +0000 (18:25 +0200)]
Add missing PR mention in ChangeLog
I noticed that the patch pushed previously had an open bug about it, so
add a reference to it.
Maciej W. Rozycki [Mon, 7 Aug 2017 16:02:04 +0000 (17:02 +0100)]
PR breakpoints/21886: mem-break: Fix breakpoint insertion location
Fix a commit
cd6c3b4ffc4e ("New gdbarch methods breakpoint_kind_from_pc
and sw_breakpoint_from_kind") regression and restore the use of
`->placed_address' rather than `->reqstd_address' as the location for a
memory breakpoint to be inserted at. Previously
`gdbarch_breakpoint_from_pc' was used that made that adjustment in
`default_memory_insert_breakpoint' from the preinitialized value,
however with the said commit that call is gone, so the passed
`->placed_address' has to be used for the initialization.
The regression manifests itself as the inability to debug any MIPS/Linux
compressed ISA dynamic executable as GDB corrupts the dynamic loader
with one of its implicit breakpoints, causing the program to crash, as
seen for example with the `mips-linux-gnu' target, o32 ABI, MIPS16 code,
and the gdb.base/advance.exp test case:
(gdb) continue
Continuing.
Program received signal SIGBUS, Bus error.
_dl_debug_initialize (ldbase=0, ns=0) at dl-debug.c:51
51 r = &_r_debug;
(gdb) FAIL: gdb.base/advance.exp: Can't run to main
gdb/
PR breakpoints/21886
* mem-break.c (default_memory_insert_breakpoint): Use
`->placed_address' rather than `->reqstd_address' for the
breakpoint location.
H.J. Lu [Mon, 7 Aug 2017 15:07:26 +0000 (08:07 -0700)]
Mark big and mach with ATTRIBUTE_UNUSED
Fix build on x86:
opcodes/disassemble.c: In function ‘disassembler’:
opcodes/disassemble.c:113:52: error: unused parameter ‘big’ [-Werror=unused-parameter]
disassembler (enum bfd_architecture a, bfd_boolean big, unsigned long mach,
^~~
opcodes/disassemble.c:113:71: error: unused parameter ‘mach’ [-Werror=unused-parameter]
disassembler (enum bfd_architecture a, bfd_boolean big, unsigned long mach,
^~~~
cc1: all warnings being treated as errors
* disassemble.c (disassembler): Mark big and mach with
ATTRIBUTE_UNUSED.
H.J. Lu [Mon, 7 Aug 2017 15:02:12 +0000 (08:02 -0700)]
Move common symbol check after bed->common_definition
bfd/
* elflink.c (elf_link_add_object_symbols): Move common symbol
check after bed->common_definition.
ld/
* testsuite/ld-elf/pr21903.s (start): Removed.
(_start): Likewise.
(__start): Likewise.
(main): Likewise.
(bar): New.
Maciej W. Rozycki [Mon, 7 Aug 2017 14:53:54 +0000 (15:53 +0100)]
GDB/opcodes: Remove arch/mach/endian disassembler assertions
Fix `set architecture' and `set endian' command disassembly regressions
from commit
39503f82427e ("Delegate opcodes to select disassembler in
GDB"), and commit
003ca0fd2286 ("Refactor disassembler selection"), as
well as a MIPS compressed ISA disassembly target regression from commit
6394c606997f ("Don't use print_insn_XXX in GDB"), which caused assertion
failures to trigger.
For example with the `mips-linux-gnu' target we get:
$ cat main.c
int
main (void)
{
return 0;
}
$ gcc -mips32r2 -O2 main.c -o main
$ gcc -mips16 -mips32r2 -O2 main.c -o main16
$ gdb
GNU gdb (GDB) 8.0.50.
20170731-git
[...]
(gdb) file main
Reading symbols from main...done.
(gdb) show architecture
The target architecture is set automatically (currently mips:isa32r2)
(gdb) show endian
The target endianness is set automatically (currently big endian)
(gdb) disassemble main
Dump of assembler code for function main:
0x00400500 <+0>: jr ra
0x00400504 <+4>: move v0,zero
End of assembler dump.
(gdb) set architecture mips:isa64r2
The target architecture is assumed to be mips:isa64r2
(gdb) disassemble main
Dump of assembler code for function main:
0x00400500 <+0>:
.../gdb/arch-utils.c:979: internal-error: int default_print_insn(bfd_vma, disassemble_info*): Assertion `info->mach == bfd_get_mach (exec_bfd)' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n) n
[...]
Command aborted.
(gdb) set architecture auto
The target architecture is set automatically (currently mips:isa32r2)
(gdb) set endian little
The target is assumed to be little endian
(gdb) disassemble main
Dump of assembler code for function main:
0x00400500 <+0>:
.../gdb/arch-utils.c:978: internal-error: int default_print_insn(bfd_vma, disassemble_info*): Assertion `info->endian == (bfd_big_endian (exec_bfd) ? BFD_ENDIAN_BIG : BFD_ENDIAN_LITTLE)' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n) n
[...]
Command aborted.
(gdb) set endian auto
The target endianness is set automatically (currently big endian)
(gdb) set architecture i386
The target architecture is assumed to be i386
(gdb) disassemble main
Dump of assembler code for function main:
0x00400500 <+0>:
.../gdb/arch-utils.c:976: internal-error: int default_print_insn(bfd_vma, disassemble_info*): Assertion `info->arch == bfd_get_arch (exec_bfd)' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n) n
[...]
Command aborted.
(gdb) set architecture auto
The target architecture is set automatically (currently mips:isa32r2)
(gdb) file main16
Load new symbol table from "main16"? (y or n) y
Reading symbols from main16...done.
(gdb) disassemble main
Dump of assembler code for function main:
0x00400501 <+0>:
.../gdb/arch-utils.c:979: internal-error: int default_print_insn(bfd_vma, disassemble_info*): Assertion `info->mach == bfd_get_mach (exec_bfd)' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n) n
Command aborted.
(gdb)
Remove the assertions then, restoring previous semantics:
(gdb) file main
Reading symbols from main...done.
(gdb) set architecture mips:isa64r2
The target architecture is assumed to be mips:isa64r2
(gdb) disassemble main
Dump of assembler code for function main:
0x00400500 <+0>: jr ra
0x00400504 <+4>: move v0,zero
End of assembler dump.
(gdb) set endian little
The target is assumed to be little endian
(gdb) disassemble main
Dump of assembler code for function main:
0x00400500 <+0>: j 0x3800c
0x00400504 <+4>: addiu s0,t0,0
End of assembler dump.
(gdb) set architecture i386
The target architecture is assumed to be i386
(gdb) disassemble main
Dump of assembler code for function main:
0x00400500 <+0>: add %eax,%esp
0x00400502 <+2>: add %cl,(%eax)
0x00400504 <+4>: add %al,(%eax)
0x00400506 <+6>: adc %ah,0x0
End of assembler dump.
(gdb) set architecture auto
The target architecture is set automatically (currently mips:isa32r2)
(gdb) set endian auto
The target endianness is set automatically (currently big endian)
(gdb) file main16
Load new symbol table from "main16"? (y or n) y
Reading symbols from main16...done.
(gdb) disassemble main
Dump of assembler code for function main:
0x00400501 <+0>: jr ra
0x00400503 <+2>: li v0,0
End of assembler dump.
(gdb)
gdb/
* arch-utils.c (default_print_insn): Remove arch/mach/endian
assertions.
opcodes/
* disassemble.c (disassembler): Remove arch/mach/endian
assertions.
Maciej W. Rozycki [Mon, 7 Aug 2017 14:53:54 +0000 (15:53 +0100)]
gdbarch: Use an anonymous union for target data in `gdbarch_info'
As an update to commit
ede5f15146ae ("gdbarch.h: Change
gdbarch_info::tdep_info's type to void *") replace the definition of the
`tdep_info' member in `struct gdbarch_info' with an anonymous union,
comprising the original member, with its type reverted to `struct
gdbarch_tdep_info *', a `tdesc_data' member of a `struct tdesc_arch_data
*' type and an `id' member of an `int *' type. Remove now unnecessary
casts throughout use places then, making code easier to read an less
prone to errors, which may happen with casting.
gdb/
* gdbarch.sh (gdbarch_info): Replace the `tdep_info' member with
a union of `tdep_info', `tdesc_data' and `id'.
* aarch64-tdep.c (aarch64_gdbarch_init): Use `info.tdesc_data'
rather than `info.tdep_info'.
* amd64-linux-tdep.c (amd64_linux_init_abi): Likewise.
* i386-linux-tdep.c (i386_linux_init_abi): Likewise.
* i386-tdep.c (i386_gdbarch_init): Likewise.
* mips-linux-tdep.c (mips_linux_init_abi): Likewise.
* mips-tdep.c (mips_gdbarch_init): Likewise.
* nds32-tdep.c (nds32_gdbarch_init): Likewise.
* rs6000-tdep.c (rs6000_gdbarch_init): Likewise.
* ppc-linux-tdep.c (ppu2spu_sniffer): Use `info.id' rather than
`info.tdep_info'.
(ppc_linux_init_abi): Use `info.tdesc_data' rather than
`info.tdep_info'.
* sparc-tdep.c (sparc32_gdbarch_init): Likewise.
* spu-multiarch.c (spu_gdbarch): Use `info.id' rather than
`info.tdep_info'.
* spu-tdep.c (spu_gdbarch_init): Likewise.
* gdbarch.h: Regenerate.
Leszek Swirski [Mon, 7 Aug 2017 14:40:38 +0000 (16:40 +0200)]
Fix dwarf2_string_attr for -gsplit-dwarf
The dwarf2_string_attr did not allow DW_FORM_GNU_str_index as a form for
string types. This manifested as null strings in the namespace_name
lookup (replaced with "(anonymous namespace)") when debugging
Fission-compiled code.
gdb/ChangeLog:
* dwarf2read.c (dwarf2_string_attr): Allow DW_FORM_GNU_strp_alt.
Alan Modra [Sun, 6 Aug 2017 12:34:25 +0000 (22:04 +0930)]
Support different ld --hash-style in the ld testsuite
When ld is running with something other than --hash-style=sysv by
default, numerous tests need tweaking. Most of the changes just add
--hash-style=sysv. I didn't want to make --hash-style=sysv global as
that means --hash-style=gnu isn't well tested.
* testsuite/ld-alpha/alpha.exp: Add --hash-style=sysv to various
test's ld options.
* testsuite/ld-arm/arm-elf.exp: Likewise.
* testsuite/ld-elf/elf.exp: Likewise.
* testsuite/ld-elf/readelf.exp: Likewise.
* testsuite/ld-elfvsb/elfvsb.exp: Likewise.
* testsuite/ld-i386/i386.exp: Likewise.
* testsuite/ld-ia64/ia64.exp: Likewise.
* testsuite/ld-m68k/m68k.exp: Likewise.
* testsuite/ld-metag/metag.exp: Likewise.
* testsuite/ld-powerpc/powerpc.exp: Likewise.
* testsuite/ld-s390/s390.exp: Likewise.
* testsuite/ld-sh/sh-vxworks.exp: Likewise.
* testsuite/ld-shared/shared.exp: Likewise.
* testsuite/ld-sparc/sparc.exp: Likewise.
* testsuite/ld-tic6x/tic6x.exp: Likewise.
* testsuite/ld-vax-elf/vax-elf.exp: Likewise.
* testsuite/ld-x86-64/mpx.exp: Likewise.
* testsuite/ld-x86-64/x86-64.exp: Likewise.
* testsuite/ld-xtensa/xtensa.exp: Likewise.
* testsuite/ld-elf/comm-data2.ld: Add .gnu.hash output section.
* testsuite/ld-elf/pr20828-v.ld: Likewise.
* testsuite/ld-elf/pr20828.ld: Likewise.
* testsuite/ld-elf/pr21233.ld: Likewise.
* testsuite/ld-elf/pr21384.ld: Likewise.
* testsuite/ld-elf/provide-hidden-1.ld: Likewise.
* testsuite/ld-elf/provide-hidden-2.ld: Likewise.
* testsuite/ld-elf/provide-hidden-s.ld: Likewise.
* testsuite/ld-scripts/cross1.t: Likewise.
* testsuite/ld-elf/stab.d: Adjust allowed section indices.
* testsuite/ld-i386/pie1.d: Don't match addresses.
* testsuite/ld-i386/plt-pic2.dd: Likewise.
* testsuite/ld-i386/pr19636-1d.d: Likewise.
* testsuite/ld-i386/pr19636-2c.d: Likewise.
* testsuite/ld-powerpc/elfv2so.d: Likewise.
* testsuite/ld-powerpc/tlsopt5.d: Likewise.
* testsuite/ld-powerpc/tlsopt5.wf: Likewise.
* testsuite/ld-powerpc/tlsopt5_32.d: Likewise.
* testsuite/ld-i386/pr19636-2b.d: Don't match _start.
* testsuite/ld-powerpc/ambiguousv1.d: Relax symbol count, index
and address match.
* testsuite/ld-powerpc/ambiguousv1b.d: Likewise.
* testsuite/ld-powerpc/ambiguousv2.d: Likewise.
* testsuite/ld-powerpc/ambiguousv2b.d: Likewise.
* testsuite/ld-aarch64/gc-plt-relocs.d: Run ld with --hash-style=sysv.
* testsuite/ld-aarch64/ifunc-1.d: Likewise.
* testsuite/ld-aarch64/ifunc-2.d: Likewise.
* testsuite/ld-aarch64/ifunc-21.d: Likewise.
* testsuite/ld-aarch64/relasz.d: Likewise.
* testsuite/ld-aarch64/tls-small-ld.d: Likewise.
* testsuite/ld-aarch64/tls-tiny-ld.d: Likewise.
* testsuite/ld-arc/tls_gd-01.d: Likewise.
* testsuite/ld-cris/libdso-10.d: Likewise.
* testsuite/ld-cris/libdso-2.d: Likewise.
* testsuite/ld-cris/pic-gc-72.d: Likewise.
* testsuite/ld-cris/pic-gc-73.d: Likewise.
* testsuite/ld-cris/tls-gd-1.d: Likewise.
* testsuite/ld-cris/tls-gd-1h.d: Likewise.
* testsuite/ld-cris/tls-gd-2.d: Likewise.
* testsuite/ld-cris/tls-gd-2h.d: Likewise.
* testsuite/ld-cris/tls-ie-10.d: Likewise.
* testsuite/ld-cris/tls-ie-11.d: Likewise.
* testsuite/ld-cris/tls-ie-8.d: Likewise.
* testsuite/ld-cris/tls-ie-9.d: Likewise.
* testsuite/ld-cris/tls-ld-4.d: Likewise.
* testsuite/ld-cris/tls-ld-5.d: Likewise.
* testsuite/ld-cris/tls-ld-6.d: Likewise.
* testsuite/ld-cris/tls-ld-7.d: Likewise.
* testsuite/ld-cris/tls-ldgd-14.d: Likewise.
* testsuite/ld-cris/tls-ldgd-15.d: Likewise.
* testsuite/ld-cris/tls-ldgdx-14.d: Likewise.
* testsuite/ld-cris/tls-ldgdx-15.d: Likewise.
* testsuite/ld-cris/tls-local-54.d: Likewise.
* testsuite/ld-cris/tls-local-60.d: Likewise.
* testsuite/ld-cris/tls-local-61.d: Likewise.
* testsuite/ld-cris/weakhiddso.d: Likewise.
* testsuite/ld-elf/linkinfo1a.d: Likewise.
* testsuite/ld-elf/linkinfo1b.d: Likewise.
* testsuite/ld-elf/pr19617a.d: Likewise.
* testsuite/ld-elfvsb/hidden2.d: Likewise.
* testsuite/ld-frv/fdpic-pie-6.d: Likewise.
* testsuite/ld-frv/fdpic-shared-2.d: Likewise.
* testsuite/ld-frv/fdpic-shared-5.d: Likewise.
* testsuite/ld-frv/fdpic-shared-6.d: Likewise.
* testsuite/ld-frv/fdpic-shared-8.d: Likewise.
* testsuite/ld-frv/fdpic-shared-local-2.d: Likewise.
* testsuite/ld-frv/fdpic-shared-local-8.d: Likewise.
* testsuite/ld-frv/tls-dynamic-2.d: Likewise.
* testsuite/ld-i386/ibt-plt-1.d: Likewise.
* testsuite/ld-i386/ibt-plt-2a.d: Likewise.
* testsuite/ld-i386/ibt-plt-2c.d: Likewise.
* testsuite/ld-i386/ibt-plt-3a.d: Likewise.
* testsuite/ld-i386/ibt-plt-3c.d: Likewise.
* testsuite/ld-i386/pr20830.d: Likewise.
* testsuite/ld-ia64/merge1.d: Likewise.
* testsuite/ld-ia64/merge2.d: Likewise.
* testsuite/ld-ia64/merge3.d: Likewise.
* testsuite/ld-ia64/merge4.d: Likewise.
* testsuite/ld-ia64/merge5.d: Likewise.
* testsuite/ld-ifunc/ifunc-1-local-x86.d: Likewise.
* testsuite/ld-ifunc/ifunc-1-x86.d: Likewise.
* testsuite/ld-ifunc/ifunc-2-i386-now.d: Likewise.
* testsuite/ld-ifunc/ifunc-2-local-i386-now.d: Likewise.
* testsuite/ld-ifunc/ifunc-2-local-x86-64-now.d: Likewise.
* testsuite/ld-ifunc/ifunc-2-local-x86-64.d: Likewise.
* testsuite/ld-ifunc/ifunc-2-x86-64-now.d: Likewise.
* testsuite/ld-ifunc/ifunc-2-x86-64.d: Likewise.
* testsuite/ld-ifunc/ifunc-3a-x86.d: Likewise.
* testsuite/ld-ifunc/pr17154-i386-now.d: Likewise.
* testsuite/ld-ifunc/pr17154-i386.d: Likewise.
* testsuite/ld-ifunc/pr17154-x86-64-now.d: Likewise.
* testsuite/ld-ifunc/pr17154-x86-64.d: Likewise.
* testsuite/ld-m68k/got-1.d: Likewise.
* testsuite/ld-m68k/got-multigot-12-13-14-34-35-ok.d: Likewise.
* testsuite/ld-m68k/got-multigot-14-ok.d: Likewise.
* testsuite/ld-m68k/got-multigot-15-er.d: Likewise.
* testsuite/ld-m68k/got-negative-12-13-14-34-ok.d: Likewise.
* testsuite/ld-m68k/got-negative-12-13-14-35-er.d: Likewise.
* testsuite/ld-m68k/got-negative-14-ok.d: Likewise.
* testsuite/ld-m68k/got-negative-15-er.d: Likewise.
* testsuite/ld-m68k/got-single-12-ok.d: Likewise.
* testsuite/ld-m68k/got-single-13-er.d: Likewise.
* testsuite/ld-m68k/got-xgot-12-13-14-15-34-35-ok.d: Likewise.
* testsuite/ld-m68k/got-xgot-15-ok.d: Likewise.
* testsuite/ld-m68k/tls-gd-1.d: Likewise.
* testsuite/ld-m68k/tls-gd-2.d: Likewise.
* testsuite/ld-m68k/tls-gd-ie-1.d: Likewise.
* testsuite/ld-m68k/tls-ie-1.d: Likewise.
* testsuite/ld-m68k/tls-ld-1.d: Likewise.
* testsuite/ld-m68k/tls-ld-2.d: Likewise.
* testsuite/ld-sh/shared-2.d: Likewise.
* testsuite/ld-sh/tlsbin-2.d: Likewise.
* testsuite/ld-sh/tlspic-2.d: Likewise.
* testsuite/ld-x86-64/bnd-branch-1-now.d: Likewise.
* testsuite/ld-x86-64/bnd-ifunc-1-now.d: Likewise.
* testsuite/ld-x86-64/bnd-ifunc-1.d: Likewise.
* testsuite/ld-x86-64/bnd-ifunc-2-now.d: Likewise.
* testsuite/ld-x86-64/bnd-ifunc-2.d: Likewise.
* testsuite/ld-x86-64/bnd-plt-1-now.d: Likewise.
* testsuite/ld-x86-64/bnd-plt-1.d: Likewise.
* testsuite/ld-x86-64/ibt-plt-1-x32.d: Likewise.
* testsuite/ld-x86-64/ibt-plt-1.d: Likewise.
* testsuite/ld-x86-64/ibt-plt-2a-x32.d: Likewise.
* testsuite/ld-x86-64/ibt-plt-2a.d: Likewise.
* testsuite/ld-x86-64/ibt-plt-2c-x32.d: Likewise.
* testsuite/ld-x86-64/ibt-plt-2c.d: Likewise.
* testsuite/ld-x86-64/ibt-plt-3a-x32.d: Likewise.
* testsuite/ld-x86-64/ibt-plt-3a.d: Likewise.
* testsuite/ld-x86-64/ibt-plt-3c-x32.d: Likewise.
* testsuite/ld-x86-64/ibt-plt-3c.d: Likewise.
* testsuite/ld-x86-64/ilp32-4-nacl.d: Likewise.
* testsuite/ld-x86-64/ilp32-4.d: Likewise.
* testsuite/ld-x86-64/load1c-nacl.d: Likewise.
* testsuite/ld-x86-64/load1c.d: Likewise.
* testsuite/ld-x86-64/load1d-nacl.d: Likewise.
* testsuite/ld-x86-64/load1d.d: Likewise.
* testsuite/ld-x86-64/pie3-nacl.d: Likewise.
* testsuite/ld-x86-64/pie3.d: Likewise.
* testsuite/ld-x86-64/pr14207.d: Likewise.
* testsuite/ld-x86-64/pr19162.d: Likewise.
* testsuite/ld-x86-64/pr19636-2d-nacl.d: Likewise.
* testsuite/ld-x86-64/pr19636-2d.d: Likewise.
* testsuite/ld-x86-64/pr20253-1d.d: Likewise.
* testsuite/ld-x86-64/pr20253-1f.d: Likewise.
* testsuite/ld-x86-64/pr20253-1j.d: Likewise.
* testsuite/ld-x86-64/pr20253-1l.d: Likewise.
* testsuite/ld-x86-64/pr20830a-now.d: Likewise.
* testsuite/ld-x86-64/pr20830a.d: Likewise.
* testsuite/ld-x86-64/pr20830b-now.d: Likewise.
* testsuite/ld-x86-64/pr20830b.d: Likewise.
* testsuite/ld-x86-64/pr21038a-now.d: Likewise.
* testsuite/ld-x86-64/pr21038a.d: Likewise.
* testsuite/ld-x86-64/pr21038b-now.d: Likewise.
* testsuite/ld-x86-64/pr21038b.d: Likewise.
* testsuite/ld-x86-64/pr21038c-now.d: Likewise.
* testsuite/ld-x86-64/pr21038c.d: Likewise.
Alan Modra [Mon, 7 Aug 2017 08:32:49 +0000 (18:02 +0930)]
PR21910, segfault in common symbol override test when hash-style=gnu
PR 21910
* elflink.c (bfd_elf_final_link): Don't segfault when sections
needed to define various dynamic tags have been discarded.
Simon Marchi [Mon, 7 Aug 2017 10:09:37 +0000 (12:09 +0200)]
remote-sim.c: Fix arg variables conflicts
The recent change introducing gdb_argv introduced some build failures in
remote-sim.c.
/home/emaisin/src/binutils-gdb/gdb/remote-sim.c: In function 'void gdbsim_load(target_ops*, const char*, int)':
/home/emaisin/src/binutils-gdb/gdb/remote-sim.c:573:22: error: conflicting declaration 'gdb_argv argv'
gdb_argv argv (args);
^
/home/emaisin/src/binutils-gdb/gdb/remote-sim.c:565:10: note: previous declaration as 'char** argv'
char **argv;
^~~~
/home/emaisin/src/binutils-gdb/gdb/remote-sim.c: In function 'void gdbsim_open(const char*, int)':
/home/emaisin/src/binutils-gdb/gdb/remote-sim.c:730:25: error: declaration of 'gdb_argv args' shadows a parameter
gdb_argv args (arg_buf);
In gdbsim_load, the new gdb_argv object conflicts with old char **argv
variable. I think the old variable should be removed.
In gdbsim_open, the new gdb_argv object conflicts with the args
parameter. This patch renames it to argv.
Built-tested for a mips host.
gdb/ChangeLog:
* remote-sim.c (gdbsim_load): Remove char **argv local variable.
(gdbsim_open): Rename gdb_argv args object to argv.
Nick Clifton [Mon, 7 Aug 2017 09:09:51 +0000 (10:09 +0100)]
Do not choose a non-ELF format input file to hold the linker created GOT sections.
PR 21884
* elf32-i386.c (elf_i386_link_setup_gnu_properties): If the dynobj
has not been set then use the bfd returned by
_bfd_elf_link_setup_gnu_properties. If that is null then search
through all the input bfds selecting the first normal, ELF format
one.
* elf64-x86-64.c (elf_x86_64_link_setup_gnu_properties): Likewise.
GDB Administrator [Mon, 7 Aug 2017 00:01:03 +0000 (00:01 +0000)]
Automatic date update in version.in
H.J. Lu [Sun, 6 Aug 2017 15:40:56 +0000 (08:40 -0700)]
x86: Lookup __tls_get_addr or ___tls_get_addr once
Instead of checking if a symbol is __tls_get_addr or ___tls_get_addr,
we check if there is a reference to __tls_get_addr or ___tls_get_addr
before starting relocation check.
* elf32-i386.c (elf_i386_link_hash_entry): Change tls_get_addr
to 1 bit.
(elf_i386_link_hash_newfunc): Initialize tls_get_addr to 0.
(elf_i386_check_tls_transition): Check tls_get_addr directly.
(elf_i386_convert_load_reloc): Update tls_get_addr check.
(elf_i386_link_check_relocs): New function.
(bfd_elf32_bfd_link_check_relocs): New.
* elf64-x86-64.c (elf_x86_64_link_hash_entry): Change tls_get_addr
to 1 bit.
(elf_x86_64_link_hash_newfunc): Initialize tls_get_addr to 0.
(elf_x86_64_check_tls_transition): Check tls_get_addr directly.
(elf_x86_64_convert_load_reloc): Update tls_get_addr check.
(elf_x86_64_link_check_relocs): New function.
(bfd_elf64_bfd_link_check_relocs): New.
(bfd_elf32_bfd_link_check_relocs): Likewise.
H.J. Lu [Sun, 6 Aug 2017 15:18:53 +0000 (08:18 -0700)]
Treat common symbol as undefined for --no-define-common
When --no-define-common is used to build shared library, treat common
symbol as undefined so that common symbols that are referenced from a
shared library to be assigned addresses only in the main program. This
eliminates the unused duplicate space in the shared library, and also
prevents any possible confusion over resolving to the wrong duplicate
when there are many dynamic modules with specialized search paths for
runtime symbol resolution.
--no-define-common is only allowed when building a shared library.
bfd/
PR ld/21903:
* elflink.c (elf_link_add_object_symbols): Treat common symbol
as undefined for --no-define-common.
include/
PR ld/21903:
* bfdlink.h (bfd_link_info): Add inhibit_common_definition.
ld/
PR ld/21903:
* ld.h (command_line): Remove inhibit_common_definition.
* ldgram.y: Replace command_line.inhibit_common_definition with
link_info.inhibit_common_definition.
* ldlang.c (lang_common): Likewise.
* lexsup.c (parse_args): Likewise.
* ldmain.c (main): Only allow --no-define-common with -shared.
* testsuite/ld-elf/pr21903.s: New file.
* testsuite/ld-elf/pr21903a.d: Likewise.
* testsuite/ld-elf/pr21903b.d: Likewise.
* testsuite/ld-elf/pr21903c.d: Likewise.
* testsuite/ld-elf/pr21903d.d: Likewise.
* testsuite/ld-elf/pr21903e.d: Likewise.
GDB Administrator [Sun, 6 Aug 2017 00:00:44 +0000 (00:00 +0000)]
Automatic date update in version.in
Tom Tromey [Mon, 31 Jul 2017 21:49:21 +0000 (15:49 -0600)]
Use gdb::unique_xmalloc_ptr when calling tilde_expand
This patch changes most sites calling tilde_expand to use
gdb::unique_xmalloc_ptr, rather than a cleanup. It also changes
scan_expression_with_cleanup to return a unique pointer, because the
patch was already touching code in that area.
Regression tested on the buildbot.
ChangeLog
2017-08-05 Tom Tromey <tom@tromey.com>
* compile/compile-object-load.c (compile_object_load): Use
gdb::unique_xmalloc_ptr.
* cli/cli-dump.c (scan_filename): Rename from
scan_filename_with_cleanup. Change return type.
(scan_expression): Rename from scan_expression_with_cleanup.
Change return type.
(dump_memory_to_file, dump_value_to_file, restore_command):
Use gdb::unique_xmalloc_ptr. Update.
* cli/cli-cmds.c (find_and_open_script): Use
gdb::unique_xmalloc_ptr.
* tracefile-tfile.c (tfile_open): Use gdb::unique_xmalloc_ptr.
* symmisc.c (maintenance_print_symbols)
(maintenance_print_msymbols): Use gdb::unique_xmalloc_ptr.
* symfile.c (symfile_bfd_open, generic_load)
(add_symbol_file_command, remove_symbol_file_command): Use
gdb::unique_xmalloc_ptr.
* source.c (openp): Use gdb::unique_xmalloc_ptr.
* psymtab.c (maintenance_print_psymbols): Use
gdb::unique_xmalloc_ptr.
* corelow.c (core_open): Use gdb::unique_xmalloc_ptr.
* breakpoint.c (save_breakpoints): Use gdb::unique_xmalloc_ptr.
* solib.c (solib_map_sections): Use gdb::unique_xmalloc_ptr.
(reload_shared_libraries_1): Likewise.
Tom Tromey [Sat, 5 Aug 2017 21:38:32 +0000 (15:38 -0600)]
Fix Rust test suite for 1.20 beta
I ran the gdb.rust tests against Rust 1.20 (beta) and saw a few
failures. The failures all came because a particular item moved to a
different module. Since the particular choice of module name isn't
important here, I simply widened the allowable results.
Tested locally against rustc 1.19, 1.20, and 1.21.
testsuite/ChangeLog
2017-08-05 Tom Tromey <tom@tromey.com>
* gdb.rust/simple.exp: Allow String to appear in a different
namespace.
Tom Tromey [Fri, 4 Aug 2017 13:30:26 +0000 (07:30 -0600)]
Remove cleanups from Rust parser
This removes the few remaining cleanups in the Rust language code.
The main difficulty here was that the earlier code allocated VEC heads
on an obstack. The new code instead introduces an object that
allocates and maintains the storage for whatever vectors are needed
during the parse.
Regression tested on the buildbot.
ChangeLog
2017-08-05 Tom Tromey <tom@tromey.com>
* rust-exp.y (rust_op_ptr, set_field): Remove typedefs.
(rust_op_vector, rust_set_vector): New typedefs.
(current_parser): New global.
(work_obstack): Change to pointer type. Update all users.
(rust_ast, pstate): Remove globals.
(struct rust_parser): New.
(%union) <params, field_inits>: Change type.
(start, tuple_expr, unit_expr, struct_expr_list, literal)
(field_expr, expr_list, maybe_expr_list, type_list): Update.
(ast_call_ish, ast_path, ast_function_type, ast_tuple_type)
(convert_params_to_types, convert_params_to_expression): Change
type of "params".
(ast_string): Change type of "fields".
(rust_parse): Make a rust_parser. Remove cleanups.
(rust_lex_tests): Make and install an auto_obstack.
Alan Modra [Sat, 5 Aug 2017 13:58:21 +0000 (23:28 +0930)]
Don't fail in elf32_hppa_set_gp
The base bfd_link_hash_table works fine here, the only thing to watch
out for is to only set elf_gp if the output is ELF.
bfd/
* elf32-hppa.c (elf32_hppa_set_gp): Don't require an
hppa_link_hash_table.
ld/
* testsuite/ld-unique/pr21529.d: Don't xfail hppa.
GDB Administrator [Sat, 5 Aug 2017 00:00:19 +0000 (00:00 +0000)]
Automatic date update in version.in
Yao Qi [Fri, 4 Aug 2017 15:02:51 +0000 (16:02 +0100)]
Unbreak GDBserver build for x32
When I verify my target description changes, I build GDB and GDBserver for
x32, but it failed.
/../../binutils-gdb/gdb/gdbserver/linux-amd64-ipa.c
../../../binutils-gdb/gdb/gdbserver/linux-amd64-ipa.c: In function ‘const target_desc* get_ipa_tdesc(int)’:
../../../binutils-gdb/gdb/gdbserver/linux-amd64-ipa.c:184:10: error: ‘X86_TDESC_AVX512’ was not declared in this scope
case X86_TDESC_AVX512:
^
../../../binutils-gdb/gdb/gdbserver/linux-amd64-ipa.c:185:14: error: ‘tdesc_x32_avx512_linux’ was not declared in this scope
return tdesc_x32_avx512_linux;
^
../../../binutils-gdb/gdb/gdbserver/linux-amd64-ipa.c: In function ‘void initialize_low_tracepoint()’:
../../../binutils-gdb/gdb/gdbserver/linux-amd64-ipa.c:282:36: error: ‘init_registers_x32_avx512_linux’ was not declared in this scope
init_registers_x32_avx512_linux ();
^
ipa_x32_linux_regobj use to be there, but removed by
22049425ce40324139be82d9a6ec518c46b65815 by mistake.
gdb/gdbserver:
2017-08-04 Yao Qi <yao.qi@linaro.org>
* configure.srv (ipa_x32_linux_regobj): New.
* linux-amd64-ipa.c (get_ipa_tdesc): Use X86_TDESC_AVX_AVX512
instead of X86_TDESC_AVX512.
(initialize_low_tracepoint): Call
init_registers_x32_avx_avx512_linux.
Yao Qi [Fri, 4 Aug 2017 13:27:58 +0000 (14:27 +0100)]
Add namespace std to nullptr_t
This patch fixes the build failure for target i686-w64-mingw32,
In file included from ../../binutils-gdb/gdb/defs.h:786:0,
from ../../binutils-gdb/gdb/gdb.c:19:
../../binutils-gdb/gdb/utils.h:188:20: error: ‘nullptr_t’ has not been declared
bool operator!= (nullptr_t)
^
../../binutils-gdb/gdb/utils.h:193:20: error: ‘nullptr_t’ has not been declared
bool operator== (nullptr_t)
^
gdb:
2017-08-04 Yao Qi <yao.qi@linaro.org>
* utils.h (gdb_argv): Add namespace std for nullptr_t.
H.J. Lu [Fri, 4 Aug 2017 12:39:56 +0000 (05:39 -0700)]
Add a testcase for "ld -d"
"ld -d" assigns space to common symbols even if a relocatable output
file is specified (with '-r').
PR ld/21904
* testsuite/ld-elf/pr21904.d: New file.
* testsuite/ld-elf/pr21904.s: Likewise.
James Clarke [Fri, 4 Aug 2017 05:44:21 +0000 (22:44 -0700)]
Add --no-relax option.
gold/
* options.h (General_options): Set a non-NULL second help string
argument for relax to allow --no-relax.
GDB Administrator [Fri, 4 Aug 2017 00:00:52 +0000 (00:00 +0000)]
Automatic date update in version.in
Ruslan Kabatsayev [Thu, 3 Aug 2017 06:11:45 +0000 (09:11 +0300)]
Add myself as a write-after-approval GDB maintainer.
Tom Tromey [Sun, 23 Jul 2017 21:52:59 +0000 (15:52 -0600)]
Remove make_cleanup_freeargv and gdb_buildargv
After the previous patches in this series, make_cleanup_freeargv and
gdb_buildargv are now unused and can be removed.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* utils.c (make_cleanup_freeargv, do_freeargv, gdb_buildargv):
Remove.
* utils.h (make_cleanup_freeargv, gdb_buildargv): Remove.
Tom Tromey [Mon, 1 May 2017 05:03:58 +0000 (23:03 -0600)]
Use gdb_argv in Python
This changes one spot in the Python code to use gdb_argv. This
removes the last cleanup from the Python layer.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* python/py-param.c (compute_enum_values): Use gdb_argv.
Tom Tromey [Mon, 1 May 2017 05:02:30 +0000 (23:02 -0600)]
Introduce gdb_argv, a class wrapper for buildargv
This introduces gdb_argv, a class wrapping an "argv" pointer; that is,
a pointer to a NULL-terminated array of char*, where both the array
and each non-NULL element in the array are xmalloc'd.
This patch then changes most users of gdb_buildargv to use gdb_argv
instead.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* utils.h (struct gdb_argv_deleter): New.
(gdb_argv): New class.
* utils.c (gdb_argv::reset): New method.
* tracepoint.c (delete_trace_variable_command): Use gdb_argv.
* tracefile.c (tsave_command): Use gdb_argv.
* top.c (new_ui_command): Use gdb_argv.
* symmisc.c (maintenance_print_symbols)
(maintenance_print_msymbols, maintenance_expand_symtabs): Use gdb_argv.
* symfile.c (symbol_file_command, generic_load)
(remove_symbol_file_command): Use gdb_argv.
* stack.c (backtrace_command): Use gdb_argv.
* source.c (add_path, show_substitute_path_command)
(unset_substitute_path_command, set_substitute_path_command):
Use gdb_argv.
* skip.c (skip_command): Use gdb_argv. Use gdb_buildargv.
* ser-mingw.c (pipe_windows_open): Use gdb_argv.
* remote.c (extended_remote_run, remote_put_command)
(remote_get_command, remote_delete_command): Use gdb_argv.
* remote-sim.c (gdbsim_load, gdbsim_create_inferior)
(gdbsim_open): Use gdb_argv.
* python/py-cmd.c (gdbpy_string_to_argv): Use gdb_argv.
* psymtab.c (maintenance_print_psymbols): Use gdb_argv.
* procfs.c (procfs_info_proc): Use gdb_argv.
* interps.c (interpreter_exec_cmd): Use gdb_argv.
* infrun.c (handle_command): Use gdb_argv.
* inferior.c (add_inferior_command, clone_inferior_command):
Use gdb_argv.
* guile/scm-string.c (gdbscm_string_to_argv): Use gdb_argv.
* exec.c (exec_file_command): Use gdb_argv.
* cli/cli-cmds.c (alias_command): Use gdb_argv.
* compile/compile.c (build_argc_argv): Use gdb_argv.
Tom Tromey [Mon, 1 May 2017 04:12:13 +0000 (22:12 -0600)]
Remove a cleanup in Python
This removes cleanups from gdbpy_decode_line, in favor of a use of
unique_xmalloc_ptr.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* python/python.c (gdbpy_decode_line): Use unique_xmalloc_ptr.
Tom Tromey [Mon, 1 May 2017 04:10:41 +0000 (22:10 -0600)]
Avoid some manual memory management in Python
This changes a few places in the Python code to avoid manual memory
management, in favor of letting std::string do the work.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* python/python.c (compute_python_string): Return std::string.
(gdbpy_eval_from_control_command): Update.
(do_start_initialization): Use std::string.
* python/py-varobj.c (py_varobj_iter_next): Use string_printf, not
xstrprintf.
* python/py-breakpoint.c (local_setattro): Use string_printf, not
xstrprintf.
Tom Tromey [Mon, 1 May 2017 03:17:58 +0000 (21:17 -0600)]
Replace do_restore_instream_cleanup with scoped_restore
This changes the users of do_restore_instream_cleanup to use a
scoped_restore instead. This patch is broken out because it warrants
some additional attention: in particular it's unclear to me whether
current_ui can change in the body of these functions -- but if it can,
then the cleanup would have modified a different UI's instream member.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* top.h (do_restore_instream_cleanup): Remove.
* top.c (do_restore_instream_cleanup): Remove.
(read_command_file): Use scoped_restore.
* cli/cli-script.c (execute_user_command): Use scoped_restore.
Tom Tromey [Mon, 1 May 2017 03:15:58 +0000 (21:15 -0600)]
Use a scoped_restore for command_nest_depth
This changes a couple of places to use a scoped_restore when
manipulating command_nest_depth.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* cli/cli-script.c (execute_user_command)
(execute_control_command): Use scoped_restore.
Tom Tromey [Mon, 1 May 2017 03:08:31 +0000 (21:08 -0600)]
Remove user_call_depth
This changes execute_user_command to remove user_call_depth, using the
size of user_args_stack instead. This avoids a cleanup.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* cli/cli-script.c (do_restore_user_call_depth): Remove.
(execute_user_command): Remove user_call_depth; use
user_args_stack's size instead.
Tom Tromey [Mon, 1 May 2017 03:04:51 +0000 (21:04 -0600)]
Remove in_user_command
While working on the next patch in this series, I found that the
global in_user_command is not used. This patch removes it. (I didn't
think to check Insight until submitting this series; and it's not very
convenient to do so, so if someone has it checked out and could look
at it, that would be nice.)
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* top.h (in_user_command): Remove.
* top.c (in_user_command): Remove.
* cli/cli-script.c (do_restore_user_call_depth)
(execute_user_command): Update.
Tom Tromey [Sat, 29 Apr 2017 05:34:32 +0000 (23:34 -0600)]
Use containers to avoid cleanups
This patch introduces the use of various containers -- std::vector,
std::string, or gdb::byte_vector -- in several spots in gdb that were
using xmalloc and a cleanup.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* valops.c (search_struct_method): Use gdb::byte_vector.
* valarith.c (value_concat): Use std::vector.
* target.c (memory_xfer_partial): Use gdb::byte_vector.
(simple_search_memory): Likewise.
* printcmd.c (find_string_backward): Use gdb::byte_vector.
* mi/mi-main.c (mi_cmd_data_write_memory): Use gdb::byte_vector.
* gcore.c (gcore_copy_callback): Use gdb::byte_vector.
* elfread.c (elf_rel_plt_read): Use std::string.
* cp-valprint.c (cp_print_value): Use gdb::byte_vector.
* cli/cli-dump.c (restore_section_callback): Use
gdb::byte_vector.
Tom Tromey [Sat, 29 Apr 2017 05:08:16 +0000 (23:08 -0600)]
Use unique_xmalloc_ptr in jit.c
This removes some cleanups from jit.c by using unique_xmalloc_ptr
instead.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* jit.c (jit_reader_load_command): Use unique_xmalloc_ptr.
Tom Tromey [Sat, 29 Apr 2017 04:59:24 +0000 (22:59 -0600)]
Replace tui_restore_gdbout with scoped_restore
This patch replaces tui_restore_gdbout (a cleaup function) with a use
of scoped_restore. This one is broken out into its own patch because
it might slightly change the behavior of gdb: it saves and restores
pagination_enabled, whereas the tui_restore_gdbout unconditionally set
pagination_enabled to 1; and I think this warrants closer review.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* tui/tui-regs.c (tui_restore_gdbout): Remove.
(tui_register_format): Use scoped_restore.
Tom Tromey [Thu, 27 Apr 2017 04:53:40 +0000 (22:53 -0600)]
More uses of scoped_restore
There were a few more places in gdb that could easily use
scoped_restore, replacing some cleanups.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* reverse.c (exec_direction_default): Remove.
(exec_reverse_once): Use scoped_restore.
* remote.c (restore_remote_timeout): Remove.
(remote_flash_erase, remote_flash_write, remote_flash_done)
(readchar, remote_serial_write): Use scoped_restore.
* cli/cli-script.c (struct source_cleanup_lines_args)
(source_cleanup_lines): Remove.
(script_from_file): Use scoped_restore.
* cli/cli-cmds.c (source_verbose_cleanup): Remove.
(source_command): Use scoped_restore.
Tom Tromey [Sat, 29 Apr 2017 04:43:54 +0000 (22:43 -0600)]
Remove make_cleanup_free_so
make_cleanup_free_so is used in a single spot. This patch introduces
a unique pointer wrapper for struct so_list, and changes this spot to
use it.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* utils.h (make_cleanup_free_so): Remove.
* utils.c (do_free_so, make_cleanup_free_so): Remove.
* solist.h (struct so_deleter): New.
(so_list_up): New typedef.
* solib-svr4.c (svr4_read_so_list): Use so_list_up.
Tom Tromey [Thu, 27 Apr 2017 04:41:30 +0000 (22:41 -0600)]
Remove make_cleanup_restore_current_language
This patch replaces make_cleanup_restore_current_language with an RAII
class that saves the current language, and restores it when the object
is destroyed.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* utils.h (make_cleanup_restore_current_language): Remove.
* utils.c (do_restore_current_language)
(make_cleanup_restore_current_language): Remove.
* parse.c (parse_exp_in_context_1)
(parse_expression_with_language): Use
scoped_restore_current_language.
* mi/mi-main.c (mi_cmd_execute): Use
scoped_restore_current_language.
* language.h (scoped_restore_current_language): New class.
Tom Tromey [Thu, 27 Apr 2017 04:12:50 +0000 (22:12 -0600)]
Remove an unlink cleanup
compile/compile.c had its own cleanup to unlink a file. This patch
replaces this cleanup with gdb::unlinker.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* compile/compile.c (cleanup_unlink_file): Remove.
(compile_to_object): Use gdb::unlinker.
(eval_compile_command): Likewise.
Tom Tromey [Thu, 27 Apr 2017 04:03:42 +0000 (22:03 -0600)]
Remove make_cleanup_fclose
After the preceding patches, make_cleanup_fclose is no longer used, so
remove it.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* utils.h (make_cleanup_fclose): Remove.
* utils.c (do_fclose_cleanup, make_cleanup_fclose): Remove.
Tom Tromey [Thu, 27 Apr 2017 04:03:08 +0000 (22:03 -0600)]
Change open_terminal_stream to return a gdb_file_up
This changes open_terminal_stream to return a gdb_file_up, eliminating
another use of make_cleanup_fclose. Arguably perhaps new_ui should
take ownership of the files using a move, but there is at least one
spot where this isn't appropriate (or at least not currently done), so
I elected to use a more minimal approach.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* top.c (open_terminal_stream): Return gdb_file_up.
(new_ui_command): Update.
Tom Tromey [Thu, 27 Apr 2017 03:56:14 +0000 (21:56 -0600)]
Use gdb_file_up in source.c
This changes some functions in source.c to use gdb_file_up.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* source.c (print_source_lines_base, forward_search_command)
(reverse_search_command): Use gdb_file_up.
Tom Tromey [Thu, 27 Apr 2017 03:52:52 +0000 (21:52 -0600)]
Use gdb_file_up in fbsd-nat.c
This updates fbsd-nat.c to use gdb_file_up. This removes a use of a
cleanup, and helps remove make_cleanup_fclose in a later patch.
I have no way to test this patch.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* fbsd-nat.c (fbsd_find_memory_regions): Update.
Tom Tromey [Thu, 27 Apr 2017 03:49:55 +0000 (21:49 -0600)]
Change return type of find_and_open_script
This changes find_and_open_script to return a
gdb::optional<open_script>, where open_script is a new type
encapsulating the two return values. The new type helps avoid
cleanups in the callers.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* cli/cli-cmds.c (find_and_open_script): Change return type.
Remove "streamp" and "full_path" parameters.
(source_script_with_search): Update.
* auto-load.c (source_script_file): Update.
* cli/cli-cmds.h (find_and_open_script): Change type.
(open_script): New struct.
Tom Tromey [Thu, 27 Apr 2017 03:39:46 +0000 (21:39 -0600)]
Introduce and use gdb_file_up
This introduces gdb_file_up, a unique pointer holding a FILE*, and
then changes some code in gdb to use it. In particular
gdb_fopen_cloexec now returns a gdb_file_up. This allow removing some
cleanups.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* xml-support.c (xml_fetch_content_from_file): Update.
* ui-file.c (stdio_file::open): Update.
* tracefile-tfile.c (tfile_start): Update.
* remote.c (remote_file_put, remote_file_get): Update.
* nat/linux-procfs.c (linux_proc_get_int)
(linux_proc_pid_get_state, linux_proc_tid_get_name): Update.
* nat/linux-osdata.c (linux_common_core_of_thread): Update.
(command_from_pid, commandline_from_pid, linux_xfer_osdata_cpus)
(print_sockets, linux_xfer_osdata_shm, linux_xfer_osdata_sem)
(linux_xfer_osdata_msg, linux_xfer_osdata_modules): Update.
* nat/linux-btrace.c (linux_determine_kernel_start): Update.
* linux-nat.c (linux_proc_pending_signals): Update.
* dwarf2read.c (write_psymtabs_to_index): Use gdb_file_up.
(file_closer): Remove.
* compile/compile.c (compile_to_object): Update.
* common/filestuff.h (struct gdb_file_deleter): New.
(gdb_file_up): New typedef.
(gdb_fopen_cloexec): Change return type.
* common/filestuff.c (gdb_fopen_cloexec): Return gdb_file_up.
* cli/cli-dump.c (fopen_with_cleanup): Remove.
(dump_binary_file, restore_binary_file): Update.
* auto-load.c (auto_load_objfile_script_1): Update.
Tom Tromey [Sun, 23 Apr 2017 16:21:50 +0000 (10:21 -0600)]
Introduce and use ui_out_emit_table
This introduces ui_out_emit_table, similar to the other existing
ui_out RAII classes, and then uses it in a number of places. This
replaces some cleanups.
ChangeLog
2017-08-03 Tom Tromey <tom@tromey.com>
* tracepoint.c (tvariables_info_1): Use ui_out_emit_table.
(info_static_tracepoint_markers_command): Likewise.
* solib.c (info_sharedlibrary_command): Use ui_out_emit_table.
* skip.c (skip_info): Use ui_out_emit_table.
* progspace.c (print_program_space): Use ui_out_emit_table.
* osdata.c (info_osdata): Use ui_out_emit_table.
* mi/mi-cmd-info.c (mi_cmd_info_ada_exceptions): Use
ui_out_emit_table.
* linux-thread-db.c (info_auto_load_libthread_db): Use
ui_out_emit_table.
* inferior.c (print_inferior): Use ui_out_emit_table.
* gdb_bfd.c (maintenance_info_bfds): Use ui_out_emit_table.
* breakpoint.c (breakpoint_1): Use ui_out_emit_table.
* auto-load.c (auto_load_info_scripts): Use ui_out_emit_table.
* ada-tasks.c (print_ada_task_info): Use ui_out_emit_table.
* ui-out.h (class ui_out_emit_table): New.
Nick Clifton [Thu, 3 Aug 2017 10:59:32 +0000 (11:59 +0100)]
Add more targets to the list of architectures not supporting format changing during linking. Fix seg-faults triggered when this is attempted.
PR ld/21884
* testsuite/ld-elf/pr21884.d: Add AVR, HPPA, IA64, M68HC1x and
SCORE to list of targets not supporting file format changes during
linking.
* testsuite/ld-unique/pr21529.d: Likewise.
* emultempl/avrelf.em (_before_allocation): Skip for non-ELF
output formats.
(avr_elf_create_output_section_statements): Fail if the output
format is not ELF.
(avr_finish): Do not access the ELF header in non-ELF format
output bfds.
* emultempl/m68hc1xelf.em (_before_allocation): Skip for non-ELF
output formats.
(m68hc11elf_create_output_section_statements): Fail if the putput
format is not ELF.
(m68hc11elf_after_allocation): Skip for non-ELF output formats.
Alan Modra [Thu, 3 Aug 2017 04:31:34 +0000 (14:01 +0930)]
ELF checks for orphan placement
The loop checking for previous orphan placement should run even when
the output is non-ELF.
PR ld/21884
* emultempl/elf32.em (gld${EMULATION_NAME}_place_orphan): Revert
last change. Rename iself to elfinput. Expand comments. Condition
ELF checks on having both input and output ELF files. Extract..
(elf_orphan_compatible): ..this new function.
GDB Administrator [Thu, 3 Aug 2017 00:01:14 +0000 (00:01 +0000)]
Automatic date update in version.in
Maciej W. Rozycki [Wed, 2 Aug 2017 15:40:16 +0000 (16:40 +0100)]
MIPS: Factor out FPU type naming in `gdbarch' debug
Replace chains of conditional expressions used in target-dependent MIPS
`gdbarch' debug output to get a textual name of the FPU type with calls
to a helper decoder function, improving code readability. No functional
change.
gdb/
* mips-tdep.c (mips_fpu_type_str): New function.
(mips_dump_tdep): Call it.
H.J. Lu [Wed, 2 Aug 2017 12:10:29 +0000 (05:10 -0700)]
Check ELF section header only for ELF output
When placing an orphan input section, check ELF section header only for
ELF output.
PR ld/21884
* emultempl/elf32.em (gld${EMULATION_NAME}_place_orphan): Check
ELF section header only for ELF output.
* testsuite/ld-elf/pr21884.d: New test.
* testsuite/ld-elf/pr21884.t: Likewise.
* testsuite/ld-elf/pr21884a.s: Likewise.
* testsuite/ld-elf/pr21884b.s: Likewise.
H.J. Lu [Wed, 2 Aug 2017 12:08:41 +0000 (05:08 -0700)]
Revert the last change to emultempl/elf32.em
It will be replaced by a different fix.
* emultempl/elf32.em (_place_orphan): Revert the last change.
Nick Clifton [Wed, 2 Aug 2017 11:12:37 +0000 (12:12 +0100)]
Add support for creating archives of slim-LTO modules using MRi script commands.
PR 21702
* arsup.c (ar_addmod): Add plugin support for the MRI ADDMOD
command.
Max Filippov [Wed, 2 Aug 2017 10:48:07 +0000 (11:48 +0100)]
Fix assertion failure for xtensa linker.
* emultempl/xtensaelf.em (xtensa_wild_group_interleave_callback):
Only check for by_name sorting.
Nick Clifton [Wed, 2 Aug 2017 10:45:05 +0000 (11:45 +0100)]
Fix seg-fault when trying to place non-ELF orphan sections.
PR 21884
* emultempl/elf32.em (_place_orphan): Skip non-ELF binaries when
looking for sections to merge.
Max Filippov [Wed, 2 Aug 2017 09:38:09 +0000 (10:38 +0100)]
Fix potential illegal memory access in Xtensa port.
* xtensa-isa.c (xtensa_isa_init): Don't update lookup table
entries for sysregs with negative indices.
Nick Clifton [Wed, 2 Aug 2017 09:19:22 +0000 (10:19 +0100)]
Fix gas and binutils testsuite failures for am33_2.0-linux target.
gas * testsuite/gas/all/gas.exp: Add am33 to the skip lists of tests
passed over by the mn10300 target.
* testsuite/gas/elf/elf.exp: Likewise.
* testsuite/gas/elf/dwarf2-11.d: Correct skip of am33 target.
* testsuite/gas/elf/dwarf2-12.d: Likewise.
* testsuite/gas/elf/dwarf2-13.d: Likewise.
* testsuite/gas/elf/dwarf2-14.d: Likewise.
* testsuite/gas/elf/dwarf2-15.d: Likewise.
* testsuite/gas/elf/dwarf2-16.d: Likewise.
* testsuite/gas/elf/dwarf2-17.d: Likewise.
* testsuite/gas/elf/dwarf2-18.d: Likewise.
* testsuite/gas/elf/dwarf2-5.d: Likewise.
* testsuite/gas/elf/dwarf2-6.d: Likewise.
* testsuite/gas/elf/dwarf2-7.d: Likewise.
binutils * testsuite/binutils-all/objdump.exp (cpus_expected): Add am33-2.
Alan Modra [Wed, 2 Aug 2017 05:48:56 +0000 (15:18 +0930)]
Teach readelf about IA64 MSB relocs.
* readelf.c (is_32bit_abs_reloc): Add R_IA64_SECREL32MSB and
R_IA64_DIR32MSB.
(is_64bit_abs_reloc): Add R_IA64_DIR64MSB.
(is_64bit_pcrel_reloc): Add R_IA64_PCREL64MSB.
GDB Administrator [Wed, 2 Aug 2017 00:00:53 +0000 (00:00 +0000)]
Automatic date update in version.in
Maciej W. Rozycki [Tue, 1 Aug 2017 18:32:25 +0000 (19:32 +0100)]
MIPS: Consistently use MIPS_FPU_TYPE for `gdbarch' member
Complement commit
74ed0bb4148e ("Replace current_gdbarch in *mips*"),
<https://sourceware.org/ml/gdb-patches/2008-06/msg00490.html>, and
consistently use the MIPS_FPU_TYPE macro to access the `->mips_fpu_type'
target-dependent `gdbarch' member. No functional change.
gdb/
* mips-tdep.c (mips_gdbarch_init): Use MIPS_FPU_TYPE to access
`->mips_fpu_type'.
Andreas Krebbel [Tue, 1 Aug 2017 13:13:40 +0000 (15:13 +0200)]
S/390: Fix claimfile failures
This fixes a segfault when trying to access the local_plt field in the
s390 specific elf data althoug the underlaying object is a generic elf
object.
This fixes the following testsuite failures:
< FAIL: plugin claimfile replace symbol with source
< FAIL: plugin claimfile resolve symbol with source
< FAIL: plugin claimfile replace file with source
< FAIL: plugin set symbol visibility with source
< FAIL: plugin ignore lib with source
< FAIL: plugin claimfile replace lib with source
< FAIL: plugin 2 with source lib
< FAIL: load plugin 2 with source
< FAIL: plugin 3 with source lib
< FAIL: load plugin 3 with source
bfd/ChangeLog:
2017-08-01 Andreas Krebbel <krebbel@linux.vnet.ibm.com>
* elf32-s390.c (elf_s390_finish_dynamic_sections): Skip if it
isn't the S/390 specific elf data.
* elf64-s390.c (elf_s390_finish_dynamic_sections): Likewise.
H.J. Lu [Tue, 1 Aug 2017 12:53:27 +0000 (05:53 -0700)]
x86: Update segment register check in Intel syntax
https://sourceware.org/ml/binutils/2009-04/msg00223.html
introduced a new Intel syntax parser which accepts
mov eax, fs:gs:[eax]
It ignores anything between ':'s after fs and treats
mov eax, DWORD PTR fs:foobar:16
mov eax, DWORD PTR fs:foobar:barfoo:16
mov eax, DWORD PTR fs:ds:16
mov eax, DWORD PTR fs:ds:cs:16
as
mov eax, DWORD PTR fs:16
This patch updates segment register check and only allows a single ':'.
PR gas/21874
* config/tc-i386-intel.c (i386_intel_operand): Update segment
register check.
* testsuite/gas/i386/intelok.s: Replace "fs:gs:[eax]" with
"fs:[eax]".
* testsuite/gas/i386/inval-seg.s: Add tests for invalid segment
register.
* testsuite/gas/i386/x86-64-inval-seg.s: Likewise.
* testsuite/gas/i386/inval-seg.l: Updated.
* testsuite/gas/i386/x86-64-inval-seg.l: Likewise.
Nick Clifton [Tue, 1 Aug 2017 07:24:22 +0000 (08:24 +0100)]
Updated Swedish translation for the binutils sub-directory.
* po/sv.po: Updated Swedish translation.
Alan Modra [Tue, 1 Aug 2017 04:38:53 +0000 (14:08 +0930)]
[GOLD] PowerPC recreate eh_frame for stubs on each relax pass
There is a very small but non-zero probability that a stub group
contains stubs on one relax pass, but does not on the next. In that
case we would get an FDE covering a zero length address range.
(Actually, it's even worse. Alignment padding for stubs can mean the
address for the non-existent stubs is past the end of the original
section to which stubs are attached, and due to the way
do_plt_fde_location calculates the length we can get a negative
length.) Fixing this properly requires removing the FDE.
Also, I have been implementing the __tls_get_addr_opt support for
gold, and that stub needs something other than the default FDE. The
necessary FDE will depend on the offset to the __tls_get_addr_opt
stub, which of course can change during relaxation. That means at the
very least, rewriting the FDE on each pass, possibly changing the FDE
size. I think that is better done by completely recreating PLT
eh_frame FDEs.
* ehframe.cc (Fde::operator==): New.
(Cie::remove_fde, Eh_frame::remove_ehframe_for_plt): New.
* ehframe.h (Fde::operator==): Declare.
(Cie::remove_fde, Eh_frame::remove_ehframe_for_plt): Likewise.
* layout.cc (Layout::remove_eh_frame_for_plt): New.
* layout.h (Layout::remove_eh_frame_for_plt): Declare.
* powerpc.cc (Target_powerpc::do_relax): Remove old eh_frame FDEs.
(Stub_table::add_eh_frame): Delete eh_frame_added_ condition.
Don't add eh_frame for empty stub section.
(Stub_table::remove_eh_frame): New.
GDB Administrator [Tue, 1 Aug 2017 00:00:54 +0000 (00:00 +0000)]
Automatic date update in version.in
Richard Henderson [Mon, 31 Jul 2017 18:44:29 +0000 (11:44 -0700)]
Update myself in MAINTAINERS
John David Anglin [Mon, 31 Jul 2017 16:51:25 +0000 (12:51 -0400)]
Fix bb instructions with double-word condition on hppa.
Kuan-Lin Chen [Mon, 31 Jul 2017 14:33:48 +0000 (15:33 +0100)]
Improve handling of ADD and SUB relocations on the RISCV target.
* elfxx-riscv.c (riscv_elf_add_sub_reloc): New function.
* (howto_table) [R_RISCV_ADD8]: Use riscv_elf_add_sub_reloc.
[R_RISCV_ADD16]: Likewise.
[R_RISCV_ADD32]: Likewise.
[R_RISCV_ADD64]: Likewise.
[R_RISCV_SUB6]: Likewise.
[R_RISCV_SUB8]: Likewise.
[R_RISCV_SUB16]: Likewise.
[R_RISCV_SUB32]: Likewise.
[R_RISCV_SUB64]: Likewise.
Nick Clifton [Mon, 31 Jul 2017 14:08:32 +0000 (15:08 +0100)]
Fix compile time error when using ansidecl.h with an old version of GCC.
PR 21850
* ansidecl.h (OVERRIDE): Protect check of __cplusplus value with
#idef __cplusplus.
Xavier Roirand [Mon, 31 Jul 2017 12:36:24 +0000 (14:36 +0200)]
Update ChangeLog (missed that in my previous commit)
gdb/
* ChangeLog: Add previous commit entry
Change-Id: Ic4f07cca66126e73340f032f19a2d5029bf0d40a
Alan Modra [Mon, 31 Jul 2017 10:23:49 +0000 (19:53 +0930)]
[GOLD] PowerPC --no-tls-optimize
This adds a --no-tls-optimize option for people who want to keep
__tls_get_addr calls in an executable rather than optimizing such code
sequences to IE/LE.
Also tidy some formatting errors, rename a variable to better reflect
its use, and tweak two functions that create pairs of GOT entries to
first check whether the GOT entry already exists before potentially
inserting the header via reserve(2). Without the check it is possible
to waste one GOT entry.
* options.h (no_tls_optimize): New powerpc option.
* powerpc.cc (Target_powerpc::abiversion, set_abiversion): Formatting.
(Target_powerpc::stk_toc): Formatting, fix comment.
(Target_powerpc::Track_tls::tls_get_addr_state): Rename from
tls_get_addr.
(Target_powerpc::optimize_tls_gd, optimize_tls_ld, optimize_tls_ie):
Return TLSOPT_NONE when !tls_optimize.
(Target_powerpc::add_global_pair_with_rel): Check
for existing reloc before reserving.
(Target_powerpc::add_local_tls_pair): Likewise.
Marty Plummer [Mon, 31 Jul 2017 09:31:18 +0000 (10:31 +0100)]
Fix compiling zlib using x86_64-musl.
PR 21861
* winduni.c (codepages): Use cp1252 for codepage 0.
Simon Marchi [Mon, 31 Jul 2017 09:22:21 +0000 (11:22 +0200)]
Fix documentation about rot agent expression bytecode
The rot agent expression bytecode rotates the three items on the top of
the stack. It is not clear which way the rotation is. However, the
documentation currently shows this as the effect of the instructions:
a b c => c b a
which doesn't make sense, since the value b doesn't move. The two
valid possibilities I see are
a b c => b c a
a b c => c a b
depending on which way you rotate.
When looking at the gdbserver code, the top of the stack becomes the
third item, and the next-to-top item becomes the top. So the second
form would be the right one, since in this notation the top of the stack
is the rightmost element:
a b c => c a b
I adjusted the symbolic description and added a bit of text to make it
more obvious.
gdb/doc/ChangeLog:
* agentexpr.texi (rot): Fix symbolic description, improve
textual description.
Xavier Roirand [Fri, 9 Jun 2017 09:36:36 +0000 (11:36 +0200)]
darwin: handle recent version of dyld
gdb/
* solib-darwin.c (DYLD_VERSION_MAX): Increase value.
Change-Id: I45da25f46b3e452d44993b122a994c818d00020b
Alan Modra [Mon, 31 Jul 2017 00:48:46 +0000 (10:18 +0930)]
PR 21847, PowerPC64 --plt-localentry again
This makes ld warn about --plt-localentry if a version of glibc
without the necessary ld.so checks is detected, and revises the
documentation.
bfd/
* elf64-ppc.c (ppc64_elf_tls_setup): Warn on --plt-localentry
without ld.so checks.
gold/
* powerpc.cc (Target_powerpc::scan_relocs): Warn on --plt-localentry
without ld.so checks.
ld/
* ld.texinfo (plt-localentry): Revise.
GDB Administrator [Mon, 31 Jul 2017 00:00:25 +0000 (00:00 +0000)]
Automatic date update in version.in
GDB Administrator [Sun, 30 Jul 2017 00:00:45 +0000 (00:00 +0000)]
Automatic date update in version.in
Alan Modra [Sat, 29 Jul 2017 03:03:35 +0000 (12:33 +0930)]
PR 21847, Don't default PowerPC64 to --plt-localentry
The big comment in ppc64_elf_tls_setup says why. I've also added some
code to the bfd linker that catches the -lpthread -lc symbol
differences and disable generation of optimized call stubs even when
--plt-localentry is activated. Gold doesn't yet have that.
PR 21847
bfd/
* elf64-ppc.c (struct ppc_link_hash_entry): Add non_zero_localentry.
(ppc64_elf_merge_symbol): Set non_zero_localentry.
(is_elfv2_localentry0): Test non_zero_localentry.
(ppc64_elf_tls_setup): Default to --no-plt-localentry.
gold/
* powerpc.cc (Target_powerpc::scan_relocs): Default to
--no-plt-localentry.
ld/
* ld.texinfo (plt-localentry): Document.
GDB Administrator [Sat, 29 Jul 2017 00:01:03 +0000 (00:01 +0000)]
Automatic date update in version.in
H.J. Lu [Fri, 28 Jul 2017 20:39:42 +0000 (13:39 -0700)]
gold: Add put_ch_reserved to 64-bit Chdr_write
The 64-bit ELF compression header has a reserved field. It should be
cleared to avoid random bits in it.
elfcpp/
PR gold/21857
* elfcpp.h (Chdr_write): Add put_ch_reserved.
(Chdr_write<64, true>::put_ch_reserved): New.
(Chdr_write<64, false>::put_ch_reserved): Likewise.
gold/
PR gold/21857
* compressed_output.cc (Output_compressed_section::set_final_data_size):
Call put_ch_reserved to clear the reserved field for 64-bit ELF.
Andreas Krebbel [Fri, 28 Jul 2017 16:34:34 +0000 (18:34 +0200)]
Add missing changelog