Alan Modra [Tue, 19 Nov 2019 04:23:44 +0000 (14:53 +1030)]
PR25191, internal error in _bfd_elf_set_section_contents
This PR copies a fuzzed PE input file to ELF output, in the process
confusing the ELF backend by copying COFF-only section flags to the
output. SEC_COFF_SHARED has the same value as SEC_ELF_COMPRESS. One
approach to fixing this problem is of course not to reuse flag bits,
but we've run out. So this patch only copies section flags that are
in the bfd_applicable_section_flags set when changing the flavour of
the output file.
PR 25191
* objcopy.c (is_nondebug_keep_contents_section): Use bfd_get_flavour.
(copy_object): Likewise.
(setup_section): Likewise. If flavour of input and output files
differ, restrict section flags to the intersection of input and
output bfd_applicable_section_flags.
Alan Modra [Tue, 19 Nov 2019 01:08:36 +0000 (11:38 +1030)]
PR25197, assertion fail coffgen.c
The testcase in this PR triggered "BFD_ASSERT (p2->is_sym)" by
sneakily generating a C_FILE sym whose value pointed into auxents.
The fix then is in the last changed line of this patch, to check
p->is_sym as well as p->u.syment.n_sclass. The other changes fix
various overflow checks that weren't as solid as they could be.
PR 25197
* coffgen.c (coff_find_nearest_line_with_names): Check that C_FILE
u.syment.n_value does point at another C_FILE sym and not into
some auxent that happens to look like a C_FILE. Properly check
for integer overflow and avoid possible pointer wrap-around.
Simplify pr17512 checks.
Alan Modra [Mon, 18 Nov 2019 21:36:24 +0000 (08:06 +1030)]
Add space between program name and file for objcopy/strip/objdump messages
The GNU coding standard does indicate there should be no space in
messages like these, but we tend to put a space in all other
messages. This patch cures the inconsistency in:
Andrew Burgess [Tue, 15 Oct 2019 21:58:08 +0000 (22:58 +0100)]
gdb/testsuite: Merge whatis.exp and ctf-whatis.exp
The recently added gdb.base/ctf-whatis.exp test is a slightly modified
version of gdb.base/whatis.exp, with a few tests removed, and the
source compiled with different compiler options. This patch merges
the two tests together into a single test script.
I tested using a version of GCC with CTF support added.
gdb/testsuite/ChangeLog:
* gdb.base/ctf-whatis.c: Delete.
* gdb.base/ctf-whatis.exp: Delete.
* gdb.base/whatis.exp: Rewrite to compile as both dwarf and ctf.
Andrew Burgess [Tue, 8 Oct 2019 09:13:54 +0000 (10:13 +0100)]
gdb/testsuite: Merge cvexpr.exp and ctf-cvexpr.exp
The recently added gdb.base/ctf-cvexpr.exp is just a copy of
gdb.base/cvexpr.exp but compiled with different options. This patch
merges these two tests together into a single test script.
I tested this change using a version of GCC with CTF support added.
gdb/testsuite/ChangeLog:
* gdb.base/ctf-cvexpr.exp: Delete.
* gdb.base/cvexpr.exp: Rewrite to compile as both dwarf and ctf.
Andrew Burgess [Tue, 8 Oct 2019 09:09:31 +0000 (10:09 +0100)]
gdb/testsuite: Introduce skip_ctf_tests guard function
Most versions of GCC in the wild don't support CTF debug format right
now, so, rather than attempting to compile the tests and failing each
time, this patch introduces a guard function to check if the compiler
supports CTF. If we don't have CTF support then the CTF tests are
skipped.
This patch only updates 3 of the 4 CTF tests, the fourth will be
handled in the next patch.
gdb/testsuite/ChangeLog:
* gdb.base/ctf-constvars.exp: Skip test if CTF is not supported in
the compiler. Clean up header comment a little.
* gdb.base/ctf-ptype.exp: Likewise.
* gdb.base/ctf-whatis.exp: Likewise.
* lib/gdb.exp (skip_ctf_tests): New proc.
A segfault can happen in a specific scenario when using TUI + a
corefile, as explained in the bug mentioned above. The problem
happens when opening a corefile on GDB:
$ gdb ./core program
entering TUI (C-x a), and then issuing a "run" command. GDB segfaults
with the following stack trace:
(top-gdb) bt
#0 0x00000000004cd5da in target_ops::shortname (this=0x0) at ../../binutils-gdb/gdb/target.h:449
#1 0x0000000000ac08fb in target_shortname () at ../../binutils-gdb/gdb/target.h:1323
#2 0x0000000000ac09ae in tui_locator_window::make_status_line[abi:cxx11]() const (this=0x23e1fa0 <_locator>) at ../../binutils-gdb/gdb/tui/tui-stack.c:86
#3 0x0000000000ac1043 in tui_locator_window::rerender (this=0x23e1fa0 <_locator>) at ../../binutils-gdb/gdb/tui/tui-stack.c:231
#4 0x0000000000ac1632 in tui_show_locator_content () at ../../binutils-gdb/gdb/tui/tui-stack.c:369
#5 0x0000000000ac63b6 in tui_set_key_mode (mode=TUI_COMMAND_MODE) at ../../binutils-gdb/gdb/tui/tui.c:321
#6 0x0000000000aaf9be in tui_inferior_exit (inf=0x2d446a0) at ../../binutils-gdb/gdb/tui/tui-hooks.c:181
#7 0x000000000044cddf in std::_Function_handler<void (inferior*), void (*)(inferior*)>::_M_invoke(std::_Any_data const&, inferior*&&) (__functor=..., __args#0=@0x7fffffffd650: 0x2d446a0)
at /usr/include/c++/9/bits/std_function.h:300
#8 0x0000000000757db9 in std::function<void (inferior*)>::operator()(inferior*) const (this=0x2cf3168, __args#0=0x2d446a0) at /usr/include/c++/9/bits/std_function.h:690
#9 0x0000000000757876 in gdb::observers::observable<inferior*>::notify (this=0x23de0c0 <gdb::observers::inferior_exit>, args#0=0x2d446a0)
at ../../binutils-gdb/gdb/gdbsupport/observable.h:106
#10 0x000000000075532d in exit_inferior_1 (inftoex=0x2d446a0, silent=1) at ../../binutils-gdb/gdb/inferior.c:191
#11 0x0000000000755460 in exit_inferior_silent (inf=0x2d446a0) at ../../binutils-gdb/gdb/inferior.c:234
#12 0x000000000059f47c in core_target::close (this=0x2d68590) at ../../binutils-gdb/gdb/corelow.c:265
#13 0x0000000000a7688c in target_close (targ=0x2d68590) at ../../binutils-gdb/gdb/target.c:3293
#14 0x0000000000a63d74 in target_stack::push (this=0x23e1800 <g_target_stack>, t=0x23c38c8 <the_amd64_linux_nat_target>) at ../../binutils-gdb/gdb/target.c:568
#15 0x0000000000a63dbf in push_target (t=0x23c38c8 <the_amd64_linux_nat_target>) at ../../binutils-gdb/gdb/target.c:583
#16 0x0000000000748088 in inf_ptrace_target::create_inferior (this=0x23c38c8 <the_amd64_linux_nat_target>, exec_file=0x2d58d30 "/usr/bin/cat", allargs="", env=0x25f12b0, from_tty=1)
at ../../binutils-gdb/gdb/inf-ptrace.c:128
#17 0x0000000000795ccb in linux_nat_target::create_inferior (this=0x23c38c8 <the_amd64_linux_nat_target>, exec_file=0x2d58d30 "/usr/bin/cat", allargs="", env=0x25f12b0, from_tty=1)
at ../../binutils-gdb/gdb/linux-nat.c:1094
#18 0x000000000074eae9 in run_command_1 (args=0x0, from_tty=1, run_how=RUN_NORMAL) at ../../binutils-gdb/gdb/infcmd.c:639
...
The problem happens because 'tui_locator_window::make_status_line'
needs the value of 'target_shortname' in order to update the status
line. 'target_shortname' is a macro which expands to:
and, in our scenario, 'current_top_target ()' returns NULL, which
obviously causes a segfault. But why does it return NULL, since,
according to its comment on target.h, it should never do that?
What is happening is that we're being caught in the middle of a
"target switch". We had the 'core_target' on top, because we were
inspecting a corefile, but when the user decided to invoke "run" GDB
had to actually create the inferior, which ends up detecting that we
have a target already, and tries to close it (from target.c):
/* See target.h. */
void
target_stack::push (target_ops *t)
{
/* If there's already a target at this stratum, remove it. */
strata stratum = t->stratum ();
if (m_stack[stratum] != NULL)
{
target_ops *prev = m_stack[stratum];
m_stack[stratum] = NULL;
target_close (prev); // <-- here
}
...
When the current target ('core_target') is being closed, it checks for
possible observers registered with it and calls them. TUI is one of
those observers, it gets called, tries to update the status line, and
GDB crashes.
The real problem is that we are clearing 'm_stack[stratum]', but
forgetting to adjust 'm_top'. Interestingly, this scenario is covered
in 'target_stack::unpush', but Pedro said he forgot to call it here..
The fix, therefore, is to call '::unpush' if there's a target on the
stack.
This patch has been tested on the Buildbot and no regressions have
been found. I'm also submitting a testcase for it.
gdb/ChangeLog:
2019-11-18 Sergio Durigan Junior <sergiodj@redhat.com>
Pedro Alves <palves@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1765117
* target.c (target_stack::push): Call 'unpush' if there's a
target on top of the stack.
Alan Modra [Wed, 2 Oct 2019 01:11:01 +0000 (10:41 +0930)]
[GOLD] OSABI not set when STT_GNU_IFUNC or STB_GNU_UNIQUE symbols output
This patch arranges to have OSABI set to ELFOSABI_GNU (if not set to
some other non-zero value) when gold outputs an ifunc local or global
symbol, or a unique global symbol to either .dynsym or .symtab.
STT_GNU_IFUNC and STB_GNU_UNIQUE have values in the LOOS to HIOS range
and therefore require interpretation according to OSABI.
I'm not sure why parameters->target() is const Target& while
parameters->sized_target() is Sized_target*, but it's inconvenient to
use the latter in Symbol_table::finalize. So this patch adds another
const_cast complained about in layout.cc and gold.cc.
PR 24853
* symtab.h (set_has_gnu_output, has_gnu_output_): New.
* symtab.cc (Symbol_table::Symbol_table): Init has_gnu_output_.
(Symbol_table::finalize): Set ELFOSABI_GNU when has_gnu_output_.
(Symbol_table::set_dynsym_indexes, Symbol_table::sized_finalize):
Call set_has_gnu_output for STT_GNU_IFUNC and STB_GNU_UNIQUE globals.
* object.cc (Sized_relobj_file::do_finalize_local_symbols): Call
set_has_gnu_output when STT_GNU_IFUNC locals will be output.
Fix a bunch of python leaks due to missing calls to tp_free in *_dealloc functions.
valgrind reports leaks in many python tests, such as:
==17162== VALGRIND_GDB_ERROR_BEGIN
==17162== 8,208 (5,472 direct, 2,736 indirect) bytes in 57 blocks are definitely lost in loss record 7,551 of 7,679
==17162== at 0x4835753: malloc (vg_replace_malloc.c:307)
==17162== by 0x6EAFD1: _PyObject_New (object.c:279)
==17162== by 0x4720E6: blpy_iter(_object*) (py-block.c:92)
==17162== by 0x698772: PyObject_GetIter (abstract.c:2577)
==17162== by 0x2343BE: _PyEval_EvalFrameDefault (ceval.c:3159)
==17162== by 0x22E9E2: function_code_fastcall (call.c:283)
==17162== by 0x2340A8: _PyObject_Vectorcall (abstract.h:127)
==17162== by 0x2340A8: call_function (ceval.c:4987)
==17162== by 0x2340A8: _PyEval_EvalFrameDefault (ceval.c:3486)
==17162== by 0x22E9E2: function_code_fastcall (call.c:283)
==17162== by 0x82172B: _PyObject_Vectorcall (abstract.h:127)
==17162== by 0x82172B: method_vectorcall (classobject.c:67)
==17162== by 0x6AF474: _PyObject_Vectorcall (abstract.h:127)
==17162== by 0x6AF474: _PyObject_CallNoArg (abstract.h:153)
==17162== by 0x6AF474: _PyObject_CallFunctionVa (call.c:914)
==17162== by 0x6B0673: callmethod (call.c:1010)
==17162== by 0x6B0673: _PyObject_CallMethod_SizeT (call.c:1103)
==17162== by 0x477DFE: gdb_PyObject_CallMethod<> (python-internal.h:182)
==17162== by 0x477DFE: get_py_iter_from_func(_object*, char const*) (py-framefilter.c:272)
==17162== by 0x4791B4: py_print_args (py-framefilter.c:706)
==17162== by 0x4791B4: py_print_frame(_object*, enum_flags<frame_filter_flag>, ext_lang_frame_args, ui_out*, int, htab*) (py-framefilter.c:960)
==17162== by 0x47A130: gdbpy_apply_frame_filter(extension_language_defn const*, frame_info*, enum_flags<frame_filter_flag>, ext_lang_frame_args, ui_out*, int, int) (py-framefilter.c:1236)
==17162== by 0x369C39: apply_ext_lang_frame_filter(frame_info*, enum_flags<frame_filter_flag>, ext_lang_frame_args, ui_out*, int, int) (extension.c:563)
==17162== by 0x4EC9C9: backtrace_command_1 (stack.c:2031)
==17162== by 0x4EC9C9: backtrace_command(char const*, int) (stack.c:2183)
...
Most of the leaks in python tests are due to the fact that many
PyObject xxxxx_dealloc functions are missing the line to free self
or obj such as:
Py_TYPE (self)->tp_free (self);
or
Py_TYPE (obj)->tp_free (obj);
With this patch, the number of python tests leaking decreases from 52 to 12.
gdb/ChangeLog
2019-11-18 Philippe Waroquiers <philippe.waroquiers@skynet.be>
Don't use class-initialization for the owner union
As reported by PhilippeW, valgrind reports that symtab is uninitialized
when compiling with GCC 4.8.5, which is the default compiler on CentOS 7.
This is apparently a compiler bug fixed in later versions, but to keep
CentOS 7 working, this patch initializes the union explicitly instead of
using a class initializer.
gdb/ChangeLog:
2019-11-18 Christian Biesinger <cbiesinger@google.com>
* symtab.h (struct symbol) <owner>: Initialize explicitly in the
constructor instead of using a class initializer.
Alan Modra [Mon, 18 Nov 2019 06:39:40 +0000 (17:09 +1030)]
elf_backend_init_file_header
This patch renames elf_backend_post_process_headers and moves the
prep_headers code into the new function. Naming the backend functions
elf_backend_init_file_header and elf_backend_modify_headers makes it
clear which function is called first.
* elf-bfd.h (struct elf_backend_data <elf_backend_init_file_header>):
Rename from elf_backend_post_process_headers.
(_bfd_elf_post_process_headers): Delete.
(_bfd_elf_init_file_header): Declare.
* elf.c (_bfd_elf_compute_section_file_positions): Call new function
in place of prep_headers and elf_backend_post_process_headers.
(_bfd_elf_init_file_header): Renamed from prep_headers with
updated args and made global. Delete dead code.
(_bfd_elf_post_process_headers): Delete.
* elf32-arm.c (elf32_arm_init_file_header): Rename from
elf32_arm_post_process_headers and call _bfd_elf_init_file_header.
Return status.
(elf_backend_init_file_header): Define.
(elf_backend_post_process_headers): Don't define.
* elf32-i386.c (elf_i386_fbsd_init_file_header): Similarly.
* elf32-m68hc1x.c (elf32_m68hc11_init_file_header): Similarly.
* elf32-metag.c (elf_metag_init_file_header): Similarly.
* elf32-spu.c (spu_elf_init_file_header
* elf32-visium.c (visium_elf_init_file_header
* elf64-alpha.c (elf64_alpha_fbsd_init_file_header
* elf64-hppa.c (elf64_hppa_init_file_header
* elf64-ia64-vms.c (elf64_vms_init_file_header
* elfnn-aarch64.c (elfNN_aarch64_init_file_header
* elfnn-ia64.c (elfNN_hpux_init_file_header
* elfxx-mips.c (_bfd_mips_init_file_header
* elfxx-mips.h (_bfd_mips_post_process_headers): Delete.
(_bfd_mips_init_file_header): Declare.
(elf_backend_post_process_headers): Delete.
(elf_backend_init_file_header): Define.
* elfxx-target.h (elf_backend_post_process_headers): Delete.
(elf_backend_init_file_header): Define and use.
* elf32-m68hc12.c (elf_backend_init_file_header): Define.
(elf_backend_post_process_headers): Don't define.
* elf32-m68hc1x.h (elf32_m68hc11_post_process_headers): Delete.
(elf32_m68hc11_init_file_header): Declare.
* elf32-ppc.c (elf_backend_post_process_headers): Remove
unnecessary undef.
Alan Modra [Mon, 18 Nov 2019 06:39:01 +0000 (17:09 +1030)]
elf_backend_modify_headers
This patch renames elf_backend_modify_program_headers and moves the
elf.c code tweaking the ELF file header for -pie -Ttext-segment to a
new function, _bfd_elf_modify_headers, which then becomes the default
elf_backed_modify_headers and is called from any other target
elf_backed_modify_headers.
* elf-bfd.h (struct elf_backend_data <elf_backend_modify_headers>):
Rename from elf_backend_modify_program_headers.
(_bfd_elf_modify_headers): Declare.
* elf.c (assign_file_positions_except_relocs): Set
elf_program_header_size. Always call elf_backend_modify_headers.
Extract code modifying file header..
(_bfd_elf_modify_headers): ..to here. New function.
* elf32-arm.c (elf_backend_modify_headers): Renamed from
elf_backend_modify_program_headers.
* elf32-i386.c: Similarly.
* elf64-x86-64.c: Similarly.
* elfxx-target.h: Similarly. Default elf_backend_modify_headers
to _bfd_elf_modify_headers.
* elf-nacl.h (nacl_modify_headers): Rename from
nacl_modify_program_headers.
* elf-nacl.c (nacl_modify_headers): Rename from
nacl_modify_program_headers and call _bfd_elf_modify_headers.
* elf32-rx.c (elf32_rx_modify_headers): Similarly.
* elf32-spu.c (spu_elf_modify_headers): Similarly.
* elfnn-ia64.c (elfNN_ia64_modify_headers): Similarly.
* elf32-sh.c (elf_backend_modify_program_headers): Don't undef.
Alan Modra [Mon, 18 Nov 2019 02:01:55 +0000 (12:31 +1030)]
PR25196, abort in rewrite_elf_program_header
This patch introduces a new "sorry, cannot handle this file" bfd error
status. The idea is to use this error in cases where bfd hasn't found
a bfd_bad_value error, ie. an input file or set of options that are
invalid, but rather an input file that is simply too difficult to
process. Typically this might happen with fuzzed object files such as
the one in the PR, a wildly improbable core file. Some things are
just not worth wasting time over to fix "properly".
PR 25196
* bfd.c (bfd_error_type): Add bfd_error_sorry.
(bfd_errmsgs): Likewise.
* elf.c (rewrite_elf_program_header): Don't abort on confused
lma/alignment. Replace bfd_error_bad_value with bfd_error_sorry.
(_bfd_elf_validate_reloc): Use bfd_error_sorry.
(_bfd_elf_final_write_processing): Likewise.
* bfd-in2.h: Regenerate.
Andrew Burgess [Mon, 4 Nov 2019 12:27:45 +0000 (12:27 +0000)]
gas: Add --gdwarf-cie-version command line flag
Add a flag to control the version of CIE that is generated. By
default gas produces CIE version 1, and this continues to be the
default after this patch.
However, a user can now provide --gdwarf-cie-version=NUMBER to switch
to either version 3 or version 4 of CIE, version 2 was never released,
and so causes an error as does any number less than 1 or greater than
4.
Producing version 4 CIE requires two new fields to be added to the
CIE, an address size field, and an segment selector field. For a flat
address space the DWARF specification indicates that the segment
selector should be 0, and the address size fields just contains the
address size in bytes. For now we support 4 or 8 byte addresses, and
the segment selector is always produced as 0. At some future time we
might need to allow targets to override this.
gas/ChangeLog:
* as.c (parse_args): Parse --gdwarf-cie-version option.
(flag_dwarf_cie_version): New variable.
* as.h (flag_dwarf_cie_version): Declare.
* dw2gencfi.c (output_cie): Switch from DW_CIE_VERSION to
flag_dwarf_cie_version.
* doc/as.texi (Overview): Document --gdwarf-cie-version.
* NEWS: Likewise.
* testsuite/gas/cfi/cfi.exp: Add new tests.
* testsuite/gas/cfi/cie-version-0.d: New file.
* testsuite/gas/cfi/cie-version-1.d: New file.
* testsuite/gas/cfi/cie-version-2.d: New file.
* testsuite/gas/cfi/cie-version-3.d: New file.
* testsuite/gas/cfi/cie-version-4.d: New file.
* testsuite/gas/cfi/cie-version.s: New file.
There is no need to keep mingw-strerror around; we can just always use
the code from posix-strerror. The main reason we had that code, it
seems, is to handle winsock error codes, but gnulib's version
handles those.
Unfortunately the code can't be moved into common-utils.c because
libinproctrace.so uses common-utils but not gnulib.
gdb/ChangeLog:
2019-11-15 Christian Biesinger <cbiesinger@google.com>
Tom Tromey [Fri, 15 Nov 2019 20:48:27 +0000 (13:48 -0700)]
Add no-dist to gnulib configure
This adds the no-dist option to the gnulib configure script. gdb
doesn't use "make dist", so there's no need for this. Adding this
option makes the Makefiles less verbose.
gnulib/ChangeLog
2019-11-15 Tom Tromey <tromey@adacore.com>
Tom Tromey [Fri, 15 Nov 2019 20:40:53 +0000 (13:40 -0700)]
Minor updates to readline configury
Christian's recent patches to gnulib made me realize that readline
should be changed to use AC_CONFIG_MACRO_DIRS (ACLOCAL_AMFLAGS is
deprecated) and that it can put the automake options into
configure.ac. I also added no-define to the automake options. This
doesn't matter much (we don't generate a config.h here), but gnulib
does it, and it does make configure slightly smaller.
readline/ChangeLog
2019-11-15 Tom Tromey <tromey@adacore.com>
* configure, Makefile.in: Rebuild.
* configure.ac: Use AC_CONFIG_MACRO_DIRS. Pass options to
AM_INIT_AUTOMAKE.
* Makefile.am (AUTOMAKE_OPTIONS, ACLOCAL_AMFLAGS): Remove.
To make these calls threadsafe. localtime_r is provided by gnulib if
necessary, and for ctime_r we can just use it because it is in a linux-
specific file.
gdb/ChangeLog:
2019-11-15 Christian Biesinger <cbiesinger@google.com>
* maint.c (scoped_command_stats::print_time): Use localtime_r
instead of localtime (provided through gnulib if necessary).
* nat/linux-osdata.c (time_from_time_t): Use ctime_r instead
of ctime.
Import the strerror_r-posix module and use it in GDB.
Makes sure to assign the return value of strerror_r to an int,
so that we get a compile error if we accidentally get the
wrong version.
gdb/ChangeLog:
2019-11-15 Christian Biesinger <cbiesinger@google.com>
* config.in: Regenerate.
* configure: Regenerate.
* gdbsupport/common.m4: No longer check for strerror_r.
* gdbsupport/posix-strerror.c (safe_strerror): Always call the
POSIX version of strerror_r, now that gnulib provides it if
necessary.
gdb/gdbserver/ChangeLog:
2019-11-15 Christian Biesinger <cbiesinger@google.com>
* config.in: Regenerate.
* configure: Regenerate.
gnulib/ChangeLog:
2019-11-15 Christian Biesinger <cbiesinger@google.com>
* Makefile.in: Regenerate.
* aclocal.m4: Regenerate.
* config.in: Regenerate.
* configure: Regenerate.
* import/Makefile.am: Update.
* import/Makefile.in: Regenerate.
* import/extra/config.rpath: New file.
* import/glthread/lock.c: New file.
* import/glthread/lock.h: New file.
* import/glthread/threadlib.c: New file.
* import/m4/gnulib-cache.m4: Update.
* import/m4/gnulib-comp.m4: Update.
* import/m4/lib-ld.m4: New file.
* import/m4/lib-link.m4: New file.
* import/m4/lib-prefix.m4: New file.
* import/m4/lock.m4: New file.
* import/m4/strerror_r.m4: New file.
* import/m4/threadlib.m4: New file.
* import/strerror_r.c: New file.
* update-gnulib.sh: Import strerror_r-posix.
Generate gnulib's toplevel Makefile.in using automake
This is a lot simpler and as a side-effect this will correctly
regenerate import/Makefile and config.h during rebuilds if
necessary.
gnulib/ChangeLog:
2019-11-15 Christian Biesinger <cbiesinger@google.com>
* Makefile.am: New file.
* Makefile.in: Replace with generated file.
* aclocal-m4-deps.mk: Remove.
* configure.ac: Use the foreign option for automake and specify
the aclocal search path here.
* update-gnulib.sh: Don't generate aclocal-m4-deps.mk anymore.
Also don't specify the aclocal include path here, now that it
is in configure.ac.
Tom Tromey [Tue, 5 Nov 2019 17:12:27 +0000 (10:12 -0700)]
Allow re-assigning to convenience variables
A customer reported somewhat odd gdb behavior, where re-assigning an
array or string to a convenience variable would yield "Too many array
elements". A test case is:
(gdb) p $x = "x"
(gdb) p $x = "xyz"
This patch fixes the problem by making a special case in the evaluator
for assignment to convenience variables, which seems like the correct
behavior.
Note that a previous patch implemented this for Ada, see commit f411722cb ("Allow re-assigning to convenience variables").
gdb/ChangeLog
2019-11-14 Tom Tromey <tromey@adacore.com>
* eval.c (evaluate_subexp_standard) <BINOP_ASSIGN>: Do not pass an
expected type for the RHS if the LHS is a convenience variable.
gdb/testsuite/ChangeLog
2019-11-14 Tom Tromey <tromey@adacore.com>
Jan Beulich [Thu, 14 Nov 2019 07:48:22 +0000 (08:48 +0100)]
x86: drop redundant SYSCALL/SYSRET templates
The Cpu64 forms are no different in their attributes except for the CPU
flags; there's no need to key these off of anything other than
CpuSYSCALL even for the 64-bit forms. Dropping these improves the
diagnostic on SYSRETQ used in 32-bit code from "unsupported instruction
`sysret'" to "invalid instruction suffix for `sysret'".
Jan Beulich [Thu, 14 Nov 2019 07:46:19 +0000 (08:46 +0100)]
x86: make AnySize an insn attribute
... instead of an operand one. Which operand it applies to can be
determined from other operand properties, but as it turns out the only
place it is actually used at doesn't even need further qualification.
Jan Beulich [Thu, 14 Nov 2019 07:45:26 +0000 (08:45 +0100)]
x86/Intel: correct CMPSD test cases' regexp closing paren placement
The CMPS test case derivation from their MOVS counterparts I did in d241b91073 ("x86/Intel: correct MOVSD and CMPSD handling") ended up
with misplaced closing parentheses in som regexps. Correct this.
valgrind reports a leak when a breakpoint is created then deleted:
==1313== 40 bytes in 1 blocks are definitely lost in loss record 1,115 of 8,596
==1313== at 0x4835753: malloc (vg_replace_malloc.c:307)
==1313== by 0x6E05BC: _PyObject_New (object.c:255)
==1313== by 0x470E4B: gdbpy_breakpoint_created(breakpoint*) (py-breakpoint.c:1023)
==1313== by 0x2946D9: operator() (std_function.h:687)
==1313== by 0x2946D9: notify (observable.h:106)
==1313== by 0x2946D9: install_breakpoint(int, std::unique_ptr<breakpoint, std::default_delete<breakpoint> >&&, int) (breakpoint.c:8136)
==1313== by 0x295BCA: create_breakpoint_sal (breakpoint.c:8878)
==1313== by 0x295BCA: create_breakpoints_sal (breakpoint.c:8919)
==1313== by 0x295BCA: create_breakpoints_sal_default (breakpoint.c:13671)
...
The leak is due to a superfluous Py_INCREF when the python object
is allocated inside gdbpy_breakpoint_created, when the python object
is allocated locally: this object has already a refcount of 1, and
the only reference is the reference from the C breakpoint object.
The Py_INCREF is however needed when the python object was created from
python: the python object was stored in bppy_pending_object, and
gdbpy_breakpoint_created creates a new reference to this object.
Solve the leak by calling 'Py_INCREF (newbp);' only in the bppy_pending_object
case.
Regression tested on debian/amd64 natively and under valgrind on centos/amd64.
Before the patch, 795 tests have a definite leak.
After the patch, 197 have a definite leak.
Thanks to Tom, that helped on irc with the python refcount logic.
gdb/ChangeLog
2019-11-14 Philippe Waroquiers <philippe.waroquiers@skynet.be>
* python/py-finishbreakpoint.c (gdbpy_breakpoint_created):
only call Py_INCREF (newbp) in the bppy_pending_object case.
Tom Tromey [Mon, 11 Nov 2019 14:43:13 +0000 (07:43 -0700)]
Remove symbol-related static asserts
commit 3573abe1d added static asserts to ensure that symbol sizes
don't vary. However, this failed to build on Windows, on at least one
ARM platform (see PR build/25182) and internally at AdaCore for PPC.
So, I think it is probably best to just remove these assertions,
effectively reverting 3573abe1d.
gdb/ChangeLog
2019-11-13 Tom Tromey <tromey@adacore.com>
Jim Wilson [Wed, 13 Nov 2019 00:13:00 +0000 (16:13 -0800)]
RISC-V: Support the INSN_CLASS.*F.* classes for .insn directive.
We have to enable the f extension through -march or ELF attribute if we use the
FPR in .insn directive. The behavior is same as the riscv_opcodes.
2019-11-12 Nelson Chu <nelson.chu@sifive.com>
opcodes/
* riscv-opc.c (riscv_insn_types): Replace the INSN_CLASS_I with
INSN_CLASS_F and the INSN_CLASS_C with INSN_CLASS_F_AND_C if we
use the floating point register (FPR).
gas/
* testsuite/gas/riscv/insn.d: Add the f extension to -march option.
Jim Wilson [Tue, 12 Nov 2019 23:50:48 +0000 (15:50 -0800)]
RISC-V: Fix ld relax failure with calls and align directives.
Make _bfd_riscv_relax_call handle section alignment padding same as
the _bfd_riscv_relax_lui and _bfd_riscv_relax_pc functions already
do. Use the max section alignment if section boundaries are crossed,
otherwise the alignment of the containing section.
bfd/
PR 25181
* elfnn-riscv.c (_bfd_riscv_relax_call): Always add max_alignment to
foff. If sym_sec->output_section and sec->output_section are the same
and not *ABS* then set max_alignment to that section's alignment.
ld/
PR 25181
* testsuite/ld-riscv-elf/call-relax-0.s: New file.
* testsuite/ld-riscv-elf/call-relax-1.s: New file.
* testsuite/ld-riscv-elf/call-relax-2.s: New file.
* testsuite/ld-riscv-elf/call-relax-3.s: New file.
* testsuite/ld-riscv-elf/call-relax.d: New test.
* testsuite/ld-riscv-elf/ld-riscv-elf.exp: Run call-relax test.
Andrew Burgess [Tue, 5 Nov 2019 14:24:17 +0000 (14:24 +0000)]
gdb: Support printf 'z' size modifier
The gdb format mechanism doesn't currently support the 'z' size
modifier, there are a few places in GDB where this is used. Instead
of removing these uses lets just add support to GDB for using 'z'.
I found this issue when trying to use some of the debug output.
Before this commit:
(gdb) set debug dwarf-line 9
(gdb) file test
Reading symbols from test...
Unrecognized format specifier 'z' in printf
(No debugging symbols found in test)
(gdb)
After this commit:
(gdb) set debug dwarf-line 9
(gdb) file test
Reading symbols from test...
Adding dir 1: /usr/include
Adding file 1: test.c
Adding file 2: stdc-predef.h
Processing actual line 3: file 1, address 0x4004a0, is_stmt 1, discrim 0
Processing actual line 4: file 1, address 0x4004a0, is_stmt 1, discrim 0
.... lots of debug output ...
Processing actual line 10: file 1, address 0x4003b7, is_stmt 0, discrim 0
(gdb)
I've added a self test to cover the integer format size modifiers,
including the 'z' modifier.
Make struct symbol inherit from general_symbol_info
Since this is now no longer a POD, also give it a constructor that
initializes all fields. (I have considered overloading operator new
to zero-initialize the memory instead; let me know if you prefer that)
gdb/ChangeLog:
2019-11-12 Christian Biesinger <cbiesinger@google.com>
Tom Tromey [Fri, 16 Aug 2019 21:16:12 +0000 (15:16 -0600)]
Make TUI resizing tests more robust
As Sergio pointed out, the TUI resizing tests are flaky. Debugging
this showed three main problems.
1. expect's "stty" command processes its arguments one-by-one. So,
rather than requesting a single resize, it sends two separate resize
requests (one for rows and one for columns). This means gdb sees two
SIGWINCH signals and resizes the terminal twice.
I consider this a bug in expect, but I couldn't readily see how to
report a bug; and anyway the fix wouldn't propagate very quickly.
This patch works around this problem by explicitly doing two separate
resizes (so it will be robust if expect ever does change); and then by
waiting for each resize to complete before continuing.
2. gdb uses curses to drive the console rendering. Currently the test
suite looks for terminal text insertion sequences to decide when a
command has completed. However, it turns out that, sometimes, curses
can output things in non-obvious ways. I didn't debug into curses but
I guess this can happen due to output optimizations. No matter the
reason, sometimes the current approach of only tracking text
insertions is not enough to detect that gdb has finished rendering.
This patch fixes this problem by arranging to detect the termination
output after any curses command, not just insertion.
3. Detecting when a resize has completed is tricky. In fact, I could
not find a way to reliably do this.
This patch fixes this problem by adding a special maint
"tui-resize-message" setting to gdb. When this is enabled, gdb will
print a message after each SIGWINCH has been fully processed. The
test suite enables this mode and then waits for the message in order
to know when control can be returned to the calling test.
This patch also adds a timeout, to avoid the situation where the
terminal code fails to notice a change for some reason. This lets the
test at least try to continue.
gdb/ChangeLog
2019-11-12 Tom Tromey <tom@tromey.com>
* tui/tui-win.c (resize_message): New global.
(show_tui_resize_message): New function.
(tui_async_resize_screen): Print message if requested.
(_initialize_tui_win): Add tui-resize-message setting.
* NEWS: Add entry for new commands.
gdb/doc/ChangeLog
2019-11-12 Tom Tromey <tom@tromey.com>
* gdb.texinfo (Maintenance Commands): Document new command.
gdb/testsuite/ChangeLog
2019-11-12 Tom Tromey <tom@tromey.com>
* lib/tuiterm.exp (_accept): Add wait_for parameter. Check output
after any command. Expect prompt after WAIT_FOR is seen.
(enter_tui): Enable resize messages.
(command): Expect command in output.
(get_line): Avoid error when cursor appears to be off-screen.
(dump_screen): Include screen size in title.
(_do_resize): New proc, from "resize".
(resize): Rewrite. Do resize in two steps.
* gdb.tui/empty.exp (layouts): Fix entries.
(check_boxes): Remove xfail.
(check_text): Dump screen on failure.
Mihail Ionescu [Tue, 12 Nov 2019 13:57:20 +0000 (13:57 +0000)]
[gas][arm] Enable VLDM, VSTM, VPUSH, VPOP for MVE
This patch enables a few instructions for Armv8.1-M MVE. Currently VLDM,
VSTM, VSTR, VLDR, VPUSH and VPOP are enabled only when the Armv8-M
Floating-point Extension is enabled. According to the ARMv8.1-M ARM,
section A.1.4.2[1], they can be enabled by having "Armv8-M Floating-point
Extension and/or Armv8.1-M MVE".
2019-11-12 Mihail Ionescu <mihail.ionescu@arm.com>
* config/tc-arm.c (do_vfp_nsyn_push): Move in order to enable it for
both fpu_vfp_ext_v1xd and mve_ext and add call to the aliased vstm
instruction for mve_ext.
(do_vfp_nsyn_pop): Move in order to enable it for both
fpu_vfp_ext_v1xd and mve_ext and add call to the aliased vldm
instruction for mve_ext.
(do_neon_ldm_stm): Add fpu_vfp_ext_v1 and mve_ext checks.
(insns): Enable vldm, vldmia, vldmdb, vstm, vstmia, vstmdb, vpop,
vpush, and fldd, fstd, flds, fsts for arm_ext_v6t2 instead
of fpu_vfp_ext_v1xd.
* testsuite/gas/arm/v8_1m-mve.s: New.
* testsuite/gas/arm/v8_1m-mve.d: New.
Mihail Ionescu [Tue, 12 Nov 2019 13:55:37 +0000 (13:55 +0000)]
[binutils][arm] Update the decoding of MVE VMOV, VMVN
This patch updates the decoding of the VMOV and VMVN instructions which depend on cmode.
Previously VMOV and VMVN with cmode 1101 were not allowed.
The cmode changes also required updating of the MVE conflict checking.
Now instructions with opcodes 0xef800d50 and 0xef800e70 correctly get decoded as VMOV
and VMVN, respectively.
2019-11-12 Mihail Ionescu <mihail.ionescu@arm.com>
* opcodes/arm-dis.c (mve_opcodes): Enable VMOV imm to vec with
cmode 1101.
(is_mve_encoding_conflict): Update cmode conflict checks for
MVE_VMVN_IMM.
2019-11-12 Mihail Ionescu <mihail.ionescu@arm.com>
Mihail Ionescu [Tue, 12 Nov 2019 13:53:06 +0000 (13:53 +0000)]
[gas][arm] Make .fpu reset the FPU/Coprocessor feature bits
This patch is fixes the '.fpu' behaviour.
Currently, using '.fpu' resets the previously selected '.fpu' options (by overwriting them),
but does not reset previous FPU options selected by other means (ie. when using
'.arch_extension fp' in conjunction with '.fpu <x>', the FPU is not reset).
Example:
.arch armv8-a @ SET BASE
.arch_extension fp @ ADD FP-ARMV8
.fpu vfpv2 @ ADD (already existing bits, does not reset)
vfms.f32 s0, s1, s2 @ OK
Jan Beulich [Tue, 12 Nov 2019 08:09:31 +0000 (09:09 +0100)]
x86: fold EsSeg into IsString
EsSeg (a per-operand bit) is used with IsString (a per-insn attribute)
only. Extend the attribute to 2 bits, thus allowing to encode
- not a string insn,
- string insn with neither operand requiring use of %es:,
- string insn with 1st operand requiring use of %es:,
- string insn with 2nd operand requiring use of %es:,
which covers all possible cases, allowing to drop EsSeg.
The (transient) need to comment out the OTUnused #define did uncover an
oversight in the earlier OTMax -> OTNum conversion, which is being taken
care of here.
Jan Beulich [Tue, 12 Nov 2019 08:07:34 +0000 (09:07 +0100)]
x86: introduce operand type "instance"
Special register "class" instances can't be combined with one another
(neither in templates nor in register entries), and hence it is not a
good use of resources (memory as well as execution time) to represent
them as individual bits of a bit field.
Furthermore the generalization becoming possible will allow
improvements to the handling of insns accepting only individual
registers as their operands.
Tom Tromey [Mon, 11 Nov 2019 19:45:35 +0000 (12:45 -0700)]
Fix typo in vFile:pwrite documentation
A user on irc noticed that the remote protocol documentation mentioned
"vFile:write" -- but this is a typo, there is only "vFile:pwrite".
This patch fixes the bug. Tested by rebuilding, committing as
obvious.
gdb/doc/ChangeLog
2019-11-11 Tom Tromey <tromey@adacore.com>
* gdb.texinfo (Host I/O Packets): Fix typo in "vFile:pwrite".
Jan Beulich [Mon, 11 Nov 2019 12:27:47 +0000 (13:27 +0100)]
Arm64: fix build with old glibc
Some old glibc versions have string.h surface "index", which some
compilers then warn about if shadowed by a local variable. Re-use an
existing variable instead.
Andrew Burgess [Tue, 15 Oct 2019 15:18:26 +0000 (16:18 +0100)]
gdb/python: Introduce gdb.lookup_static_symbols
If gdb.lookup_static_symbol is going to return a single symbol then it
makes sense (I think) for it to return a context sensitive choice of
symbol, that is the global static symbol that would be visible to the
program at that point.
However, if the user of the python API wants to instead get a
consistent set of global static symbols, no matter where they stop,
then they have to instead consider all global static symbols with a
given name - there could be many. That is what this new API function
offers, it returns a list (possibly empty) of all global static
symbols matching a given name (and optionally a given symbol domain).
Andrew Burgess [Mon, 23 Sep 2019 15:59:08 +0000 (16:59 +0100)]
gdb/python: smarter symbol lookup for gdb.lookup_static_symbol
When using gdb.lookup_static_symbol I think that GDB should find
static symbols (global symbol with static linkage) from the current
object file ahead of static symbols from other object files.
This means that if we have two source files f1.c and f2.c, and both
files contains 'static int foo;', then when we are stopped in f1.c a
call to 'gdb.lookup_static_symbol ("foo")' will find f1.c::foo, and if
we are stopped in f2.c we would find 'f2.c::foo'.
Given that gdb.lookup_static_symbol always returns a single symbol,
but there can be multiple static symbols with the same name GDB is
always making a choice about which symbols to return. I think that it
makes sense for the choice GDB makes in this case to match what a user
would get on the command line if they asked to 'print foo'.
gdb/testsuite/ChangeLog:
* gdb.python/py-symbol.c: Declare and call function from new
py-symbol-2.c file.
* gdb.python/py-symbol.exp: Compile both source files, and add new
tests for gdb.lookup_static_symbol.
* gdb.python/py-symbol-2.c: New file.
gdb/doc/ChangeLog:
* python.texi (Symbols In Python): Extend documentation for
gdb.lookup_static_symbol.
gdb/ChangeLog:
* python/py-symbol.c (gdbpy_lookup_static_symbol): Lookup in
static block of current object file first. Also fix typo in
header comment.
Andrew Burgess [Fri, 8 Nov 2019 16:18:43 +0000 (16:18 +0000)]
gdb: Add a class to track last display symtab and line information
In stack.c we currently have a set of static global variables to track
the last displayed symtab and line. This commit moves all of these
into a class and adds an instance of the class to track the same
information.
The API into stack.c is unchanged after this cleanup.
There should be no user visible changes after this commit.
gdb/ChangeLog:
* stack.c (set_last_displayed_sal): Delete.
(last_displayed_sal_valid): Delete.
(last_displayed_pspace): Delete.
(last_displayed_addr): Delete.
(last_displayed_symtab): Delete.
(last_displayed_line): Delete.
(class last_displayed_symtab_info_type): New.
(last_displayed_symtab_info): New static global variable.
(print_frame_info): Call methods on last_displayed_symtab_info.
(clear_last_displayed_sal): Update header comment, and make use of
last_displayed_symtab_info.
(last_displayed_sal_is_valid): Likewise.
(get_last_displayed_pspace): Likewise.
(get_last_displayed_addr): Likewise.
(get_last_displayed_symtab): Likewise.
(get_last_displayed_line): Likewise.
(get_last_displayed_sal): Likewise.
* stack.h (clear_last_displayed_sal): Update header comment.
(last_displayed_sal_is_valid): Likewise.
(get_last_displayed_pspace): Likewise.
(get_last_displayed_addr): Likewise.
(get_last_displayed_symtab): Likewise.
(get_last_displayed_line): Likewise.
(get_last_displayed_sal): Likewise.
Andrew Burgess [Fri, 8 Nov 2019 23:39:14 +0000 (23:39 +0000)]
gdb_vecs.h: Avoid self move assign
While working on another patch I ran into an issue with
unordered_remove (in gdb_vecs.h), where removing the last item of the
vector can cause a self move assign.
When compiling the C++ standard library in debug mode (with
-D_GLIBCXX_DEBUG=1) this causes an error to trigger.
I've fixed the issue in this patch and provided a unit test.
The provided unit test includes an assignment operator which checks
for self move assign, this removes the need to compile with
-D_GLIBCXX_DEBUG=1 in order to spot the bug. If you're keen to see
the error reported from the C++ standard library then remove operator=
from the unit test and recompile GDB with -D_GLIBCXX_DEBUG=1.
gdb/ChangeLog:
* Makefile.in (SUBDIR_UNITTESTS_SRCS): Add new file to the list.
* unittests/vec-utils-selftests.c: New file.
* gdbsupport/gdb_vecs.h (unordered_remove): Avoid self move assign.
Tom Tromey [Sun, 10 Nov 2019 17:33:07 +0000 (10:33 -0700)]
Remove can_highlight from TUI windows
Each TUI window has a "can_highlight" member. However, this has the
same meaning as "can_box" -- a window can be highlighted if and only
if it can be boxed. So, this patch removes can_highlight in favor of
simply using can_box.
gdb/ChangeLog
2019-11-10 Tom Tromey <tom@tromey.com>
H.J. Lu [Fri, 8 Nov 2019 17:31:06 +0000 (09:31 -0800)]
i386: Only check suffix in instruction mnemonic
We should check suffix in instruction mnemonic when matching instruction.
In Intel syntax, normally we check for memory operand size. But the same
mnemonic with 2 different encodings can have the same memory operand
size and i.suffix is set to LONG_DOUBLE_MNEM_SUFFIX from memory operand
size in Intel syntax to distinguish them. When there is no suffix in
mnemonic, we check LONG_DOUBLE_MNEM_SUFFIX in i.suffix for mnemonic
suffix.
gas/
PR gas/25167
* config/tc-i386.c (match_template): Don't check instruction
suffix set from operand.
* testsuite/gas/i386/code16.d: New file.
* testsuite/gas/i386/code16.s: Likewise.
* testsuite/gas/i386/i386.exp: Run code16.
* testsuite/gas/i386/x86-64-branch-4.l: Updated.
opcodes/
PR gas/25167
* i386-opc.tbl: Remove IgnoreSize from cmpsd and movsd.
* i386-tbl.h: Regenerated.
Tom Tromey [Tue, 22 Oct 2019 19:32:39 +0000 (13:32 -0600)]
Constify command_line_input
This changes command_line_input to return a "const char *", which is
appropriate because the memory is owned by command_line_input. Then
it fixes up the users.
I looked at making command_line_input transfer ownership to its caller
instead, but this is complicated due to the way read_next_line is
called, so I decided against it.
Tested by rebuilding.
gdb/ChangeLog
2019-11-08 Tom Tromey <tromey@adacore.com>
* top.c (read_command_file): Update.
(command_line_input): Make return type const.
* python/py-gdb-readline.c: Update.
* linespec.c (decode_line_2): Update.
* defs.h (command_line_input): Make return type const.
* cli/cli-script.c (read_next_line): Make return type const.
* ada-lang.c (get_selections): Update.
Jan Beulich [Fri, 8 Nov 2019 08:03:23 +0000 (09:03 +0100)]
x86: introduce operand type "class"
Many operand types, in particular the various kinds of registers, can't
be combined with one another (neither in templates nor in register
entries), and hence it is not a good use of resources (memory as well as
execution time) to represent them as individual bits of a bit field.
[gas][aarch64] Add the v8.6 Data Gathering Hint mnemonic [10/X]
Hi,
This patch is part of a series that adds support for Armv8.6-A
to binutils.
In this last patch, the new Data Gathering Hint mnemonic is introduced.
Committed on behalf of Mihail Ionescu.
gas/ChangeLog:
2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com>
* testsuite/gas/aarch64/dgh.s: New test.
* testsuite/gas/aarch64/dgh.d: New test.
opcodes/ChangeLog:
2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com>
* opcodes/aarch64-tbl.h (V8_6_INSN): New macro for v8.6 instructions.
(aarch64_opcode_table): Add data gathering hint mnemonic.
* opcodes/aarch64-dis-2.c: Account for new instruction.
This patch is part of a series that adds support for Armv8.6-A
(Matrix Multiply and BFloat16 extensions) to binutils.
This patch introduces the Matrix Multiply (Int8, F32, F64) extensions
to the aarch64 backend.
The following instructions are added: {s/u}mmla, usmmla, {us/su}dot,
fmmla, ld1rob, ld1roh, d1row, ld1rod, uzip{1/2}, trn{1/2}.
Committed on behalf of Mihail Ionescu.
gas/ChangeLog:
2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com>
* config/tc-aarch64.c: Add new arch fetures to suppport the mm extension.
(parse_operands): Add new operand.
* testsuite/gas/aarch64/i8mm.s: New test.
* testsuite/gas/aarch64/i8mm.d: New test.
* testsuite/gas/aarch64/f32mm.s: New test.
* testsuite/gas/aarch64/f32mm.d: New test.
* testsuite/gas/aarch64/f64mm.s: New test.
* testsuite/gas/aarch64/f64mm.d: New test.
* testsuite/gas/aarch64/sve-movprfx-mm.s: New test.
* testsuite/gas/aarch64/sve-movprfx-mm.d: New test.
include/ChangeLog:
2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com>
* opcode/aarch64.h (AARCH64_FEATURE_I8MM): New.
(AARCH64_FEATURE_F32MM): New.
(AARCH64_FEATURE_F64MM): New.
(AARCH64_OPND_SVE_ADDR_RI_S4x32): New.
(enum aarch64_insn_class): Add new instruction class "aarch64_misc" for
instructions that do not require special handling.
opcodes/ChangeLog:
2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com>
* aarch64-tbl.h (aarch64_feature_i8mm_sve, aarch64_feature_f32mm_sve,
aarch64_feature_f64mm_sve, aarch64_feature_i8mm, aarch64_feature_f32mm,
aarch64_feature_f64mm): New feature sets.
(INT8MATMUL_INSN, F64MATMUL_SVE_INSN, F64MATMUL_INSN,
F32MATMUL_SVE_INSN, F32MATMUL_INSN): New macros to define matrix multiply
instructions.
(I8MM_SVE, F32MM_SVE, F64MM_SVE, I8MM, F32MM, F64MM): New feature set
macros.
(QL_MMLA64, OP_SVE_SBB): New qualifiers.
(OP_SVE_QQQ): New qualifier.
(INT8MATMUL_SVE_INSNC, F64MATMUL_SVE_INSNC,
F32MATMUL_SVE_INSNC): New feature set for bfloat16 instructions to support
the movprfx constraint.
(aarch64_opcode_table): Support for SVE_ADDR_RI_S4x32.
(aarch64_opcode_table): Define new instructions smmla,
ummla, usmmla, usdot, sudot, fmmla, ld1rob, ld1roh, ld1row, ld1rod
uzip{1/2}, trn{1/2}.
* aarch64-opc.c (operand_general_constraint_met_p): Handle
AARCH64_OPND_SVE_ADDR_RI_S4x32.
(aarch64_print_operand): Handle AARCH64_OPND_SVE_ADDR_RI_S4x32.
* aarch64-dis-2.c (aarch64_opcode_lookup_1, aarch64_find_next_opcode):
Account for new instructions.
* opcodes/aarch64-asm-2.c (aarch64_insert_operand): Support the new
S4x32 operand.
* aarch64-opc-2.c (aarch64_operands): Support the new S4x32 operand.
[Patch][binutils][aarch64] .bfloat16 directive for AArch64 [7/10]
Hi,
This patch is part of a series that adds support for Armv8.6-A
(Matrix Multiply and BFloat16 extensions) to binutils.
This patch implements the '.bfloat' directive for the AArch64 backend.
The syntax for the directive is:
.bfloat16 <0-n numbers>
e.g.
.bfloat16 12.0
.bfloat16 0.123, 1.0, NaN, 5
This is implemented by utilizing the ieee_atof_detail function in order
to encode the slightly
different bfloat16 format.
Added testcases to verify the correct encoding for various bfloat16
values (NaN, Infinity (+ & -), normals, subnormals etc...).
Cross compiled and tested on aarch64-none-elf and aarch64-none-linux-gnu
with no issues.
Committed on behalf of Mihail Ionescu.
gas/ChangeLog:
2019-10-29 Mihail Ionescu <mihail.ionescu@arm.com>
2019-10-29 Barnaby Wilks <barnaby.wilks@arm.com>
* config/tc-aarch64.c (md_atof): Add encoding for the bfloat16 format.
* testsuite/gas/aarch64/bfloat16-directive-le.d: New test.
* testsuite/gas/aarch64/bfloat16-directive-be.d: New test.
* testsuite/gas/aarch64/bfloat16-directive.s: New test.
[Patch][binutils][arm] .bfloat16 directive for Arm [6/X]
Hi,
This patch is part of a series that adds support for Armv8.6-A
(Matrix Multiply and BFloat16 extensions) to binutils.
This patch implements the '.bfloat16' directive for the Arm backend.
The syntax for the directive is:
.bfloat16 <0-n numbers>
e.g.
.bfloat16 12.0
.bfloat16 0.123, 1.0, NaN, 5
This is implemented by utilizing the ieee_atof_detail function (included
in the previous patch) in order to encode the slightly different
bfloat16 format.
Added testcases to verify the correct encoding for various bfloat16
values (NaN, Infinity (+ & -), normals, subnormals etc...).
Cross compiled and tested on arm-none-eabi and arm-none-linux-gnueabihf
with no issues.
Committed on behalf of Mihail Ionescu.
gas/ChangeLog:
2019-10-21 Mihail Ionescu <mihail.ionescu@arm.com>
2019-10-21 Barnaby Wilks <barnaby.wilks@arm.com>
* config/tc-arm.c (md_atof): Add encoding for bfloat16
* testsuite/gas/arm/bfloat16-directive-le.d: New test.
* testsuite/gas/arm/bfloat16-directive-be.d: New test.
* testsuite/gas/arm/bfloat16-directive.s: New test.
[Patch][binutils] Generic support for parsing numbers in bfloat16 format [5/X]
Hi,
This patch is part of a series that adds support for Armv8.6-A
(Matrix Multiply and BFloat16 extensions).
This patch contains some general refactoring of the atof_ieee
function, exposing a function that allows a higher level of control
over the format of IEEE-like floating point numbers.
This has been done in order to be able to add a directive for assembling
floating point literals in the bfloat16 format in the following patches.
Committed on behalf of Mihail Ionescu.
Tested on arm-none-eabi, arm-none-linux-gnueabihf, aarch64-none-elf
and aarch64-none-linux-gnuwith no issues.
gas/ChangeLog:
2019-10-21 Mihail Ionescu <mihail.ionescu@arm.com>
2019-10-21 Barnaby Wilks <barnaby.wilks@arm.com>
* as.h (atof_ieee_detail): Add prototype for atof_ieee_detail function.
(atof_ieee): Move some code into the atof_ieee_detail function.
(atof_ieee_detail): Add function that provides a higher level of control over generating
IEEE-like numbers.
This patch is part of a series that adds support for Armv8.6-A
(Matrix Multiply and BFloat16 extensions) to binutils.
This patch introduces BFloat16 instructions to the arm backend.
The following BFloat16 instructions are added: vdot, vfma{l/t},
vmmla, vfmal{t/b}, vcvt, vcvt{t/b}.
gas/ChangeLog:
2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com>
2019-11-07 Matthew Malcomson <matthew.malcomson@arm.com>
* config/tc-arm.c (arm_archs): Add armv8.6-a option.
(cpu_arch_ver): Add TAG_CPU_ARCH_V8 tag for Armv8.6-a.
* doc/c-arm.texi (-march): New armv8.6-a arch.
* config/tc-arm.c (arm_ext_bf16): New feature set.
(enum neon_el_type): Add NT_bfloat value.
(B_MNEM_vfmat, B_MNEM_vfmab): New bfloat16 encoder
helpers.
(BAD_BF16): New message.
(parse_neon_type): Add bf16 type specifier.
(enum neon_type_mask): Add N_BF16 type.
(type_chk_of_el_type): Account for NT_bfloat.
(el_type_of_type_chk): Account for N_BF16.
(neon_three_args): Split out from neon_three_same.
(neon_three_same): Part split out into neon_three_args.
(CVT_FLAVOUR_VAR): Add bf16_f32 cvt flavour.
(do_neon_cvt_1): Account for vcvt.bf16.f32.
(do_bfloat_vmla): New.
(do_mve_vfma): New function to deal with the mnemonic clash between the BF16
vfmat and the MVE vfma in a VPT block with a 't'rue condition.
(do_neon_cvttb_1): Account for vcvt{t,b}.bf16.f32.
(do_vdot): New
(do_vmmla): New
(insns): Add vdot and vmmla mnemonics.
(arm_extensions): Add "bf16" extension.
* doc/c-arm.texi: Document "bf16" extension.
* testsuite/gas/arm/attr-march-armv8_6-a.d: New test.
* testsuite/gas/arm/bfloat16-bad.d: New test.
* testsuite/gas/arm/bfloat16-bad.l: New test.
* testsuite/gas/arm/bfloat16-bad.s: New test.
* testsuite/gas/arm/bfloat16-cmdline-bad-2.d: New test.
* testsuite/gas/arm/bfloat16-cmdline-bad-3.d: New test.
* testsuite/gas/arm/bfloat16-cmdline-bad.d: New test.
* testsuite/gas/arm/bfloat16-neon.s: New test.
* testsuite/gas/arm/bfloat16-non-neon.s: New test.
* testsuite/gas/arm/bfloat16-thumb-bad.d: New test.
* testsuite/gas/arm/bfloat16-thumb-bad.l: New test.
* testsuite/gas/arm/bfloat16-thumb.d: New test.
* testsuite/gas/arm/bfloat16-vfp.d: New test.
* testsuite/gas/arm/bfloat16.d: New test.
* testsuite/gas/arm/bfloat16.s: New test.
include/ChangeLog:
2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com>
2019-11-07 Matthew Malcomson <matthew.malcomson@arm.com>
* opcode/arm.h (ARM_EXT2_V8_6A, ARM_AEXT2_V8_6A,
ARM_ARCH_V8_6A): New.
* opcode/arm.h (ARM_EXT2_BF16): New feature macro.
(ARM_AEXT2_V8_6A): Include above macro in definition.
opcodes/ChangeLog:
2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com>
2019-11-07 Matthew Malcomson <matthew.malcomson@arm.com>
* arm-dis.c (select_arm_features): Update bfd_march_arm_8 with
Armv8.6-A.
(coprocessor_opcodes): Add bfloat16 vcvt{t,b}.
(neon_opcodes): Add bfloat SIMD instructions.
(print_insn_coprocessor): Add new control character %b to print
condition code without checking cp_num.
(print_insn_neon): Account for BFloat16 instructions that have no
special top-byte handling.
[Patch][binutils][arm] Create a new generic coprocessor array [3/10]
Hi,
This patch is part of a series that adds support for Armv8.6-A
(Matrix Multiply and BFloat16 extensions) to binutils.
Some generic instructions match a large range of encoding space (e.g.
stc, mcr, mrc).
Currently these instructions are in the coprocessor_opcodes array, which
means they are checked before many other instructions when disassembling
arm and thumb32 codes.
This patch moves the generic instructions into a separate array to be
checked later on.
This is done in order to avoid instruction conflict between the generic
instructions and newer ones -- this has already been seen with MVE, and
is also a problem with BFloat16.
One way to avoid the conflict could be to swap the search order between
coprocessor_opcodes and neon_opcodes.
We avoid this since it's a larger change that may introduce extra bugs
(that aren't caught by the testsuite).
We have decided against searching the generic array after searching the
arm specific and thumb32 specific arrays with a similar reasoning about
keeping the change small.
Regression tested with arm-none-linux-gnueabihf.
Committed on behalf of Mihail Ionescu.
opcodes/ChangeLog:
2019-10-29 Mihail Ionescu <mihail.ionescu@arm.com>
2019-10-29 Matthew Malcomson <matthew.malcomson@arm.com>
* arm-dis.c (print_insn_coprocessor,
print_insn_generic_coprocessor): Create wrapper functions around
the implementation of the print_insn_coprocessor control codes.
(print_insn_coprocessor_1): Original print_insn_coprocessor
function that now takes which array to look at as an argument.
(print_insn_arm): Use both print_insn_coprocessor and
print_insn_generic_coprocessor.
(print_insn_thumb32): As above.
This patch is part of a series that adds support for Armv8.6-A
(Matrix Multiply and BFloat16 extensions) to binutils.
This patch introduces the following BFloat16 instructions to the
aarch64 backend: bfdot, bfmmla, bfcvt, bfcvtnt, bfmlal[t/b],
bfcvtn2.
Committed on behalf of Mihail Ionescu.
gas/ChangeLog:
2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com>
2019-11-07 Matthew Malcomson <matthew.malcomson@arm.com>
* config/tc-aarch64.c (vectype_to_qualifier): Special case the
S_2H operand qualifier.
* doc/c-aarch64.texi: Document bf16 and bf16mmla4 extensions.
* testsuite/gas/aarch64/bfloat16.d: New test.
* testsuite/gas/aarch64/bfloat16.s: New test.
* testsuite/gas/aarch64/illegal-bfloat16.d: New test.
* testsuite/gas/aarch64/illegal-bfloat16.l: New test.
* testsuite/gas/aarch64/illegal-bfloat16.s: New test.
* testsuite/gas/aarch64/sve-bfloat-movprfx.s: New test.
* testsuite/gas/aarch64/sve-bfloat-movprfx.d: New test.
include/ChangeLog:
2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com>
2019-11-07 Matthew Malcomson <matthew.malcomson@arm.com>
* opcode/aarch64.h (AARCH64_FEATURE_BFLOAT16): New feature macros.
(AARCH64_ARCH_V8_6): Include BFloat16 feature macros.
(enum aarch64_opnd_qualifier): Introduce new operand qualifier
AARCH64_OPND_QLF_S_2H.
(enum aarch64_insn_class): Introduce new class "bfloat16".
(BFLOAT16_SVE_INSNC): New feature set for bfloat16
instructions to support the movprfx constraint.
opcodes/ChangeLog:
2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com>
2019-11-07 Matthew Malcomson <matthew.malcomson@arm.com>
* aarch64-asm.c (aarch64_ins_reglane): Use AARCH64_OPND_QLF_S_2H
in reglane special case.
* aarch64-dis-2.c (aarch64_opcode_lookup_1,
aarch64_find_next_opcode): Account for new instructions.
* aarch64-dis.c (aarch64_ext_reglane): Use AARCH64_OPND_QLF_S_2H
in reglane special case.
* aarch64-opc.c (struct operand_qualifier_data): Add data for
new AARCH64_OPND_QLF_S_2H qualifier.
* aarch64-tbl.h (QL_BFDOT QL_BFDOT64, QL_BFDOT64I, QL_BFMMLA2,
QL_BFCVT64, QL_BFCVTN64, QL_BFCVTN2_64): New qualifiers.
(aarch64_feature_bfloat16, aarch64_feature_bfloat16_sve,
aarch64_feature_bfloat16_bfmmla4): New feature sets.
(BFLOAT_SVE, BFLOAT): New feature set macros.
(BFLOAT_SVE_INSN, BFLOAT_BFMMLA4_INSN, BFLOAT_INSN): New macros
to define BFloat16 instructions.
(aarch64_opcode_table): Define new instructions bfdot,
bfmmla, bfcvt, bfcvtnt, bfdot, bfdot, bfcvtn, bfmlal[b/t]
bfcvtn2, bfcvt.
This patch is part of a series that adds support for Armv8.6-A
to binutils.
This first patch adds the Armv8.6-A flag to binutils.
No instructions are behind it at the moment.
Commited on behalf of Mihail Ionescu.
gas/ChangeLog:
2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com>
2019-11-07 Matthew Malcomson <matthew.malcomson@arm.com>
* config/tc-aarch64.c (armv8.6-a): New arch.
* doc/c-aarch64.texi (armv8.6-a): Document new arch.
include/ChangeLog:
2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com>
2019-11-07 Matthew Malcomson <matthew.malcomson@arm.com>
Nick Clifton [Thu, 7 Nov 2019 11:56:54 +0000 (11:56 +0000)]
Allow the --output option of the "ar" prorgam to extract files to locations outside of the current directory.
* ar.c (open_output_file): Check for filename validity before
prefixing with output directory.
Display the constructed output filename if in verbose mode.
(extract_file): Let open_output_file display the filename.
Tom de Vries [Thu, 7 Nov 2019 09:49:56 +0000 (10:49 +0100)]
[gdb/contrib] Add words.sh script
Add a script that takes a list of files as arguments and output a list of
words from the C comments with their frequencies.
For:
...
$ ./gdb/contrib/words.sh $(find gdb -type f -name "*.c" -o -name "*.h")
...
it generates a list of ~15000 words prefixed with frequency.
This could be used to generate a dictionary that is kept as part of the
sources, against which new code can be checked, generating a warning or
error. The hope is that misspellings would trigger this frequently, and rare
words rarely, otherwise the burden of updating the dictionary would be too
much.
And for:
...
$ ./gdb/contrib/words.sh -f 1 $(find gdb -type f -name "*.c" -o -name "*.h")
...
it generates a list of ~5000 words with frequency 1.
This can be used to scan for misspellings manually.