binutils_os216

Fork

gold: aarch64 split stack support

This patch adds gold support for split-stack function calling non-split
function on aarch64.

gold/ChangeLog:

* gold/aarch64.c (AArch64_insn_utilities::aarch64_movn_decode_imm):
New function.
(AArch64_insn_utilities::aarch64_movk_decode_imm): Likewise.
(AArch64_insn_utilities::clz_hwi): Likewise.
(AArch64_insn_utilities::aarch64_bitmask_imm): Likewise.
(Target_aarch64::do_calls_non_split): Likewise.
* gold/testsuite/Makefile.am [DEFAULT_TARGET_AARCH64] (check_SCRIPTS):
add split_aarch64.sh.
[DEFAULT_TARGET_AARCH64] (check_DATA): Add split_aarch64_{1,2,3,4,r)
tests.
[DEFAULT_TARGET_AARCH64] (SPLIT_DEFSYMS): New rule.
[DEFAULT_TARGET_AARCH64] (split_aarch64_1.o): Likewise.
[DEFAULT_TARGET_AARCH64] (split_aarch64_2.o): Likewise.
[DEFAULT_TARGET_AARCH64] (split_aarch64_3.o): Likewise.
[DEFAULT_TARGET_AARCH64] (split_aarch64_4.o): Likewise.
[DEFAULT_TARGET_AARCH64] (split_aarch64_r.o): Likewise.
[DEFAULT_TARGET_AARCH64] (split_aarch64_n.o): Likewise.
[DEFAULT_TARGET_AARCH64] (split_aarch64_1): Likewise.
[DEFAULT_TARGET_AARCH64] (split_aarch64_2): Likewise.
[DEFAULT_TARGET_AARCH64] (split_aarch64_2.stdout): Likewise.
[DEFAULT_TARGET_AARCH64] (split_aarch64_3.stdout): Likewise.
[DEFAULT_TARGET_AARCH64] (split_aarch64_4): Likewise.
[DEFAULT_TARGET_AARCH64] (split_aarch64_4.stdout): Likewise.
[DEFAULT_TARGET_AARCH64] (split_aarch64_r.stdout): Likewise.
[DEFAULT_TARGET_AARCH64] (MOSTLYCLEANFILES): Likewise.
* gold/testsuite/split_aarch64.sh: New file.
* gold/testsuite/split_aarch64_1.s: Likewise.
* gold/testsuite/split_aarch64_2.s: Likewise.
* gold/testsuite/split_aarch64_3.s: Likewise.
* gold/testsuite/split_aarch64_4.s: Likewise.
* gold/testsuite/split_aarch64_n.s: Likewise.

Revert "PowerPC PLT speculative execution barriers"

This reverts most of commit 1be5d8d3bb.
Left in place are addition of --no-plt-align to some ppc32 ld tests
and the ld.texinfo --no-plt-thread-safe fix.

RISC-V: Eliminate spurious error w/ reloc truncated message

bfd/
* elfnn-riscv.c (riscv_elf_relocate_section): Return TRUE if used
callback to report an error.

Automatic date update in version.in

Remove some $ARCH_read_pc and $ARCH_write_pc

Nowadays, gdbarch_read_pc is called in this way,

if (gdbarch_read_pc_p (gdbarch))
pc_val = gdbarch_read_pc (gdbarch, regcache);
/* Else use per-frame method on get_current_frame. */
else if (gdbarch_pc_regnum (gdbarch) >= 0)
{
ULONGEST raw_val;

if (regcache_cooked_read_unsigned (regcache,
gdbarch_pc_regnum (gdbarch),
&raw_val) == REG_UNAVAILABLE)

some ports don't have to define its own gdbarch read_pc method if the
pc value is simply a unsigned value from "pc" register. The same rule
applies to regcache_write_pc. This patch removes these $ARCH_read_pc
and $ARCH_write_pc functions.

gdb:

2018-02-06 Yao Qi <yao.qi@linaro.org>

* ft32-tdep.c (ft32_read_pc): Remove.
(ft32_write_pc): Remove.
(ft32_gdbarch_init): Update.
* m32r-tdep.c (m32r_read_pc): Remove.
(m32r_gdbarch_init): Update.
* mep-tdep.c (mep_read_pc): Remove.
(mep_gdbarch_init): Update.
* microblaze-tdep.c (microblaze_write_pc): Remove.
(microblaze_gdbarch_init): Update.
* mn10300-tdep.c (mn10300_read_pc): Remove.
(mn10300_write_pc): Remove.
(mn10300_gdbarch_init): Update.
* moxie-tdep.c (moxie_read_pc): Remove.
(moxie_write_pc): Remove.
(moxie_gdbarch_init): Update.

Fix PR ld/22263 on SPARC.

This is -fpie -pie generating dynamic relocations in the text section,
simply because no TLS transitions are applied in PIE mode. The meat
of the patch is to turn calls to bfd_link_pic (info) in TLS-related code
into !bfd_link_executable (info) and there are quite a lot of them...

bfd/
* elfxx-sparc.c (sparc_elf_tls_transition): Turn call to bfd_link_pic
into call to !bfd_link_executable and tidy up.
(_bfd_sparc_elf_check_relocs): Fix formatting and tidy up.
<R_SPARC_TLS_LE_HIX22>: Turn call to bfd_link_pic into call to
!bfd_link_executable.
<R_SPARC_TLS_IE_HI22>: Likewise.
<GOT relocations>: Remove useless code, tidy and merge blocks.
<R_SPARC_TLS_GD_CALL>: Turn call to bfd_link_pic into call to
!bfd_link_executable.
<R_SPARC_WPLT30>: Tidy up.
(_bfd_sparc_elf_gc_mark_hook): Turn call to bfd_link_pic into call
to !bfd_link_executable.
(allocate_dynrelocs): Likewise.
(_bfd_sparc_elf_relocate_section): Fix formatting and tidy up.
<R_SPARC_TLS_GD_HI22>: Merge into...
<R_SPARC_TLS_GD_LO10>: ...this. Adjust 4th argument in call to
sparc_elf_tls_transition and remove redundant code.
<R_SPARC_TLS_LDM_HI22>: Turn call to bfd_link_pic into call to
!bfd_link_executable.
<R_SPARC_TLS_LDO_HIX22>: Likewise.
<R_SPARC_TLS_LE_HIX22>: Likewise. Tidy up.
<R_SPARC_TLS_LDM_CALL>: Likewise.
<R_SPARC_TLS_GD_CALL>: Likewise. Tidy up.
<R_SPARC_TLS_GD_ADD>: Likewise.
<R_SPARC_TLS_LDM_ADD>: Likewise.
<R_SPARC_TLS_LDO_ADD>: Likewise.
<R_SPARC_TLS_IE_LD>: Likewise.
ld/
* testsuite/ld-elf/tls.exp (AFLAGS_PIC): Define on SPARC.
(pr22263-1): Pass AFLAGS_PIC to the assembler.
* testsuite/ld-sparc/tlspie32.s: Add test for other 3 transitions.
* testsuite/ld-sparc/tlspie32.dd: Adjust to above.
* testsuite/ld-sparc/tlspie64.s: Add test for other 3 transitions.
* testsuite/ld-sparc/tlspie64.dd: Adjust to above.

Treat OP_F77_UNDETERMINED_ARGLIST as OP_FUNCALL

When I debug some fortran expression parsing, I got

(gdb) set debug expression 1
(gdb) p intvla(5,5,5)
Dump of expression @ 0x205fa80, before conversion to prefix form:
Language fortran, 19 elements, 16 bytes each.
Index Opcode Hex Value String Value
0 OP_VAR_VALUE 40 (...............
1 <unknown 31863232> 31863232 .1..............
....
14 BINOP_REM 5 ................
15 OP_LONG 38 &...............
16 OP_F77_UNDETERMINED_ARGLIST 48 0...............
17 BINOP_MUL 3 ................
18 OP_F77_UNDETERMINED_ARGLIST 48 0...............
Dump of expression @ 0x205fa80, after conversion to prefix form:
Expression: `Invalid expression
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

This patch fixes this problem by handling OP_F77_UNDETERMINED_ARGLIST in
the same way as handling OP_FUNCALL. With this patch applied, the output
looks better,

(gdb) p intvla (5,5,5)
Dump of expression @ 0x2d75590, before conversion to prefix form:
Language fortran, 19 elements, 16 bytes each.
Index Opcode Hex Value String Value
0 OP_VAR_VALUE 40 (...............
....
16 OP_F77_UNDETERMINED_ARGLIST 48 0...............
17 BINOP_MUL 3 ................
18 OP_F77_UNDETERMINED_ARGLIST 48 0...............
Dump of expression @ 0x2d75590, after conversion to prefix form:
Expression: `vla_primitives::intvla (5, 5, 5)'
Language fortran, 19 elements, 16 bytes each.

0 OP_F77_UNDETERMINED_ARGLIST Number of args: 3
3 OP_VAR_VALUE Block @0x297e1c0, symbol @0x297cd50 (intvla)
7 OP_LONG Type @0x2976900 (int), value 5 (0x5)
11 OP_LONG Type @0x2976900 (int), value 5 (0x5)
15 OP_LONG Type @0x2976900 (int), value 5 (0x5)

gdb:

2018-02-06 Yao Qi <yao.qi@linaro.org>

* expprint.c (print_subexp_standard): Handle
OP_F77_UNDETERMINED_ARGLIST.
(dump_subexp_body_standard): Likewise.

Fix up one ChangeLog entry date

The patch was committed on 5th Feb, and the first line of ChangeLog
entry is too long. The max line length is 74.
https://sourceware.org/gdb/wiki/ContributionChecklist#Properly_Formatted_GNU_ChangeLog

Improve the find_nearest_line function for the MIPS target so that it tries harder to find a function name.

PR 22789
* elfxx-mips.c (_bfd_mips_elf_find_nearest_line): If the dwarf
functions failed to find the function name, try the generic elf
find function instead.

Fix GCC 8's -Wstringop-overflow on bfd/coff-rs6000.c

GCC 8 will bring a new warning option which will detect possible
overflow and truncation on string manipulation functions. For more
details, see:

https://gcc.gnu.org/ml/gcc-patches/2017-08/msg00471.html

While compiling BFD with it, I can see one place on bfd/coff-rs6000.c
where the warning is triggered. This:

(void) strncpy (fhdr.magic, XCOFFARMAG, SXCOFFARMAG);

will not include the trailing NUL on fhdr.magic, but that's fine
because it's a magic number. The fix is trivial: just use memcpy
instead.

OK to push?

2018-02-06 Sergio Durigan Junior <sergiodj@redhat.com>

* coff-rs6000.c (xcoff_write_archive_contents_old): Use
'memcpy' instead of 'strncpy' when writing the magic number.

Replace reachable assertion with a test and return of NULL.

PR 22793
* readelf.c (find_section): Replace assertion with test and return
of NULL.
(find_section_by_address): Add test of section header table
existance.
(find_section_by_type): Likewise.
(find_section_in_set): Likewise.

Prevent attempts to call strncpy with a zero-length field by chacking the size of debuglink sections.

PR 22794
* opncls.c (bfd_get_debug_link_info_1): Check the size of the
section before attempting to read it in.
(bfd_get_alt_debug_link_info): Likewise.

Allow the find_abstract_instance_name() function in the BFD library to also return file and line number information.

The nm utility supports -l for using debug information to obtain file and line information for each symbol, if available.
We have a tool that consumes this information and displays it.
This identified a problem with the 'nm' utility.

When a source is compiled with -O2, functions can be inlined. The compiler also produces an uninlined copy of the function, normally for linking to other object files.
In the case of DWARF2 debug information, the compiler generates debug information to describe a function. If that function is inlined, the compiler then references that debug information from the inlined and uninlined copies of the routine through the use of the DW_AT_abstract_origin reference.
When nm is used on such a file, it is not able to find file and line information because that information is present in the common debug information and not at each actual implementation of the function.
The 'nm' utility only retrieves the name of the function from the abstract origin debug information and no more.

What I am proposing is to modify the find_abstract_instance_name() function (which I renamed to find_abstract_instance() ) to return the name of the function as well as any file and line information. The routine is already parsing all of the debug information in the abstract instance, so it is easy to pick up the file and line information at that time. If, for some reason, the file and line information is not present, the routine behaves as before.

For example, if I have a simple test case:

int foo(int j)
{
if (j < 15)
j += j << 2;
else
j += j << 6;
return j;
}

int main (int argc,char **argv)
{
int i = argc;
i += foo(i);
return i;
}

If that test case is compiled with -O2 and then 'nm -l' reads that executable, it currently produces this symbol output (ignoring a lot of library symbols):

8048400 T foo
080482e0 T main /scratch/pcarroll/its254/test/mytest.c:12

If I modify 'nm' to return file and line information for abstract instances, it produces the following output:

08048400 T foo /scratch/pcarroll/its254/test/mytest.c:1
080482e0 T main /scratch/pcarroll/its254/test/mytest.c:12
--------------------------------------------------------------------------
bfd * bfd/dwarf2.c (find_abstract_name): Modified to return file and
line information in addition to name, if they can be found.

Automatic date update in version.in

ppc64: Fix stwux encoding

With gcc-8.0.1-0.9.fc28.x86_64 I get:

../../gdb/rs6000-tdep.c: In function 'CORE_ADDR skip_prologue(gdbarch*, CORE_ADDR, CORE_ADDR, rs6000_framedata*)':
../../gdb/rs6000-tdep.c:1911:34: error: bitwise comparison always evaluates to false [-Werror=tautological-compare]
else if ((op & 0xfc1f016a) == 0x7c01016e)
~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~

https://www.ibm.com/support/knowledgecenter/en/ssw_aix_72/com.ibm.aix.alangref/idalangref_stwux_stux_instrs.htm
says
bit 21 - 30 = 183
Those are bits 1..10 in normal bit order: 183<<1 = 0x16e

gdb/ChangeLog
2018-02-04 Jan Kratochvil <jan.kratochvil@redhat.com>

* rs6000-tdep.c (skip_prologue): Fix stwux encoding.

[PR22764][LD][AARCH64]Allow R_AARCH64_ABS16 and R_AARCH64_ABS32 against absolution symbol or undefine symbol in shared object.

The assumption that R_AARCH64_ABS16 and R_AARCH64_ABS32 relocation in LP64 abi
will be used to generate an address does not hold for absolute symbol.
In this case, it is a value fixed at static linking time.

The condition to check the relocations is relax to allow absolute symbol and
undefined symbol case.

bfd/

2018-02-05 Renlin Li <renlin.li@arm.com>

PR ld/22764
* elfnn-aarch64.c (elfNN_aarch64_check_relocs): Relax the
R_AARCH64_ABS32 and R_AARCH64_ABS16 for absolute symbol. Apply the
check for writeable section as well.

ld/

2018-02-05 Renlin Li <renlin.li@arm.com>

PR ld/22764
* testsuite/ld-aarch64/emit-relocs-258.s: Define symbol as an address.
* testsuite/ld-aarch64/emit-relocs-259.s: Likewise.
* testsuite/ld-aarch64/aarch64-elf.exp: Run new test.
* testsuite/ld-aarch64/pr22764.s: New.
* testsuite/ld-aarch64/pr22764.d: New.

Remove myself as a write-after-approval GDB maintainer.

Since I'll be leaving Ericsson.

gdb/
* MAINTAINERS (Write After Approval): Remove Antoine Tremblay.

x86: Remove the unused _GLOBAL_OFFSET_TABLE_

Since _GLOBAL_OFFSET_TABLE_ may be referenced implicitly on x86,
checking ref_regular_nonweak leaves the unused _GLOBAL_OFFSET_TABLE_
in output. This patch checks explicit GOT references instead.

ld-i386/discarded1.s and ld-x86-64/discarded1.s are updated to avoid
linker optimization which removes GOT references.

bfd/

PR ld/22782
* elf32-i386.c (elf_i386_check_relocs): Set got_referenced if
_GLOBAL_OFFSET_TABLE_ is referenced or GOT is needed to resolve
undefined weak symbol to 0.
* elf64-x86-64.c (elf_x86_64_check_relocs): Set got_referenced
if _GLOBAL_OFFSET_TABLE_ is referenced.
* elfxx-x86.c (_bfd_x86_elf_size_dynamic_sections): Check
got_referenced instead of ref_regular_nonweak. Remove the
unused _GLOBAL_OFFSET_TABLE_ from symbol table.
* elfxx-x86.h (elf_x86_link_hash_table): Add got_referenced.

ld/

PR ld/22782
* testsuite/ld-i386/discarded1.s: Replace mov with div.
* testsuite/ld-x86-64/discarded1.s: Likewise.
* testsuite/ld-i386/i386.exp: Run pr22782.
* testsuite/ld-i386/load1-nacl.d: Updated for removing
_GLOBAL_OFFSET_TABLE_ from output.
* testsuite/ld-i386/load1.d: Likewise.
* testsuite/ld-x86-64/load1a-nacl.d: Likewise.
* testsuite/ld-x86-64/load1a.d: Likewise.
* testsuite/ld-x86-64/load1b-nacl.d: Likewise.
* testsuite/ld-x86-64/load1b.d: Likewise.
* testsuite/ld-i386/pr22782.d: New file.
* testsuite/ld-i386/pr22782.s: Likewise.
* testsuite/ld-x86-64/pr22782.s: Likewise.
* testsuite/ld-x86-64/pr22782a.d: Likewise.
* testsuite/ld-x86-64/pr22782b.d: Likewise.
* testsuite/ld-x86-64/x86-64.exp: Run pr22782a and pr22782b.

Use visitors for make_gdb_type

Remove the make_gdb_type functions from the tdesc_type_ classes.
Replace with a static make_gdb_type function that uses a element
visitor called gdb_type_creator.

gdb/
* target-descriptions.c (tdesc_element_visitor) Add empty implementations.
(tdesc_type): Move make_gdb_type from here.
(tdesc_type_builtin): Likewise.
(tdesc_type_vector): Likewise.
(tdesc_type_with_fields): Move make_gdb_type_ functions from here.
(make_gdb_type_struct): Move from tdesc_type_with_fields.
(make_gdb_type_union): Likewise.
(make_gdb_type_flags): Likewise.
(make_gdb_type_enum): Likewise.
(make_gdb_type): New function.
(tdesc_register_type): Use static make_gdb_type.

RISC-V/GAS: Correct an `expr' global shadowing error for pre-4.8 GCC

Correct a commit f0531ed6a429 ("Compress loads/stores with implicit 0
offset.") regression and remove a `-Wshadow' compilation error:

cc1: warnings being treated as errors
.../gas/config/tc-riscv.c: In function 'riscv_handle_implicit_zero_offset':
.../gas/config/tc-riscv.c:1194: error: declaration of 'expr' shadows a global declaration
.../gas/expr.h:180: error: shadowed declaration is here
make[4]: *** [tc-riscv.o] Error 1

which for versions of GCC before 4.8 prevents GAS for RISC-V targets
from being built. See also GCC PR c/53066.

gas/
* config/tc-riscv.c (riscv_handle_implicit_zero_offset): Rename
`expr' parameter to `ep'.

RISC-V/BFD: Correct a missing initializer error with pre-4.7 GCC

Correct a commit 9d06997adb88 ("RISC-V: Relax RISCV_PCREL_* to
RISCV_GPREL_*") regression and remove a `-Wmissing-field-initializers'
compilation error:

cc1: warnings being treated as errors
.../bfd/elfnn-riscv.c: In function '_bfd_riscv_relax_pc':
.../bfd/elfnn-riscv.c:3117: error: missing initializer
.../bfd/elfnn-riscv.c:3117: error: (near initialization for 'hi_reloc.hi_addend')
make[4]: *** [elf64-riscv.lo] Error 1

which for versions of GCC before 4.7 prevents support for RISC-V targets
from being built. See also GCC PR c/36750.

bfd/
* elfnn-riscv.c (_bfd_riscv_relax_pc): Use `memset' to
initialize `hi_reloc'.

MIPS/BFD: Correctly report unsupported `.reginfo' section size

Report an error when an unsupported `.reginfo' section size is found in
`_bfd_mips_elf_section_processing', removing an assertion that triggers
at elfxx-mips.c:7105 in GAS when assembling input like:

.section .reginfo
.word 0xdeadbeef

and in `objcopy --rename-section' when renaming an incorrectly sized
section to `.reginfo'.

bfd/
* elfxx-mips.c (_bfd_mips_elf_section_processing): For
SHT_MIPS_REGINFO sections don't assert the correct size and
report an error instead.

binutils/
* testsuite/binutils-all/mips/mips-reginfo.d: New test.
* testsuite/binutils-all/mips/mips-reginfo-n32.d: New test.
* testsuite/binutils-all/mips/mips-reginfo.s: New test source.
* testsuite/binutils-all/mips/mips.exp: Run the new tests.

gas/
* testsuite/gas/mips/reginfo-2.d: New test.
* testsuite/gas/mips/reginfo-2-n32.d: New test.
* testsuite/gas/mips/reginfo-2.l: New test stderr output.
* testsuite/gas/mips/reginfo-2.s: New test source.
* testsuite/gas/mips/mips.exp: Run the new tests.

ELF/BFD: Propagate the return status from backend section processing

Fix the issue of any failure from `->elf_backend_section_processing'
not being propagated by `_bfd_elf_write_object_contents'. The MIPS
`_bfd_mips_elf_section_processing' handler can actually potentially
fail, however the caller currently ignores that.

bfd/
* elf.c (_bfd_elf_write_object_contents): Propagate a failure
status from `->elf_backend_section_processing'.

binutils/testsuite: Support stderr options with `run_dump_test'

Add support for the `error', `error_output', `warning' and
`warning_output' options for `run_dump_test' input files, based on the
version of the procedure in ld/testsuite/lib/ld-lib.exp and providing
compatible semantics. These options apply to PROG under test and let
test cases specify output expected on stderr as well as express a
requirement for PROG to exit unsuccessfully. Messages to match against
can be supplied either inline or fetched from a named file. Update
procedure description in the introductory comment accordingly.

As the exit status from `remote_exec' is regrettably lost in our default
implementation of `binutils_run', which is user-overridable, avoid
changing this procedure's API and use a global `binutils_run_status'
variable to pass the status up to the caller, similarly to how
`binutils_run_failed' is handled. Document the new variable in the
respective introductory comments.

binutils/
* testsuite/config/default.exp (binutils_run): Document
`binutils_run_status'.
* testsuite/lib/utils-lib.exp (default_binutils_run): Likewise,
and set it.
(run_dump_test): Add `error', `error_output', `warning' and
`warning_output' options. Update documentation accordingly.

Updated Brazillian portuguese and Russian translation

Align natural-format register values to the same column

Currently, commands such as "info reg", "info all-reg", as well as register
window in the TUI print badly aligned columns, like here:

eax 0x1 1
ecx 0xffffd3e0 -11296
edx 0xffffd404 -11260
ebx 0xf7fa5ff4 -134586380
esp 0xffffd390 0xffffd390
ebp 0xffffd3c8 0xffffd3c8
esi 0x0 0
edi 0x0 0
eip 0x8048b60 0x8048b60 <main+16>
eflags 0x286 [ PF SF IF ]
cs 0x23 35
ss 0x2b 43
ds 0x2b 43
es 0x2b 43
fs 0x0 0
gs 0x63 99

After this patch, these commands print the third column values consistently
aligned one under another, provided the second column is not too long.
Originally, the third column was (attempted to be) aligned using a simple tab
character. This patch changes the alignment to spaces only. The tests checking
the output and expecting the single tab have been fixed in a previous patch, so
this change doesn't break any.

gdb/ChangeLog:

* infcmd.c (default_print_one_register_info): Align natural-format
column values consistently one under another.
(pad_to_column): New function.

Move comment in gdb/dwarf2read.c::dwarf2_physname

This commit just moves a comment right next to where it is actually
relevant. No actual code change.

gdb/ChangeLog:

* dwarf2read.c (dwarf2_physname): Move commment.

Tested by rebuilding GDB.

Automatic date update in version.in

ld/testsuite: Replace -Sw with -SW for readelf

-Sw is a typo. It should be -SW for readelf.

* testsuite/ld-i386/lea1d.d: Replace -Sw with -SW.
* testsuite/ld-i386/lea1e.d: Likewise.
* testsuite/ld-i386/lea1f.d: Likewise.
* testsuite/ld-x86-64/lea1g.d: Likewise.
* testsuite/ld-x86-64/lea1h.d: Likewise.
* testsuite/ld-x86-64/lea1i.d: Likewise.
* testsuite/ld-x86-64/lea1j.d: Likewise.
* testsuite/ld-x86-64/lea1k.d: Likewise.
* testsuite/ld-x86-64/lea1l.d: Likewise.

Automatic date update in version.in