Tom Tromey [Tue, 4 Jan 2022 15:02:24 +0000 (08:02 -0700)]
Change how Python architecture and language are handled
Currently, gdb's Python layer captures the current architecture and
language when "entering" Python code. This has some undesirable
effects, and so this series changes how this is handled.
This is incorrect, because both of these are NULL when not otherwise
assigned. This can cause crashes in some cases -- I've added one to
the test suite. (Note that this crasher is just an example, other
ones along the same lines are possible.)
Second, when the language is captured in this way, it means that
Python code cannot affect the current language for its own purposes.
It's reasonable to want to write code like this:
gdb.execute('set language mumble')
... stuff using the current language
gdb.execute('set language previous-value')
However, this won't actually work, because the language is captured on
entry. I've added a test to show this as well.
This patch changes gdb to try to avoid capturing the current values.
The Python concept of the current gdbarch is only set in those few
cases where a non-default value is computed or needed; and the
language is not captured at all -- instead, in the cases where it's
required, the current language is temporarily changed.
H.J. Lu [Fri, 14 Jan 2022 21:48:36 +0000 (13:48 -0800)]
ld: Rewrite lang_size_relro_segment_1
1. Compute the desired PT_GNU_RELRO segment base and find the maximum
section alignment of sections starting from the PT_GNU_RELRO segment.
2. Find the first preceding load section.
3. Don't add the 1-page gap between the first preceding load section and
the relro segment if the maximum page size >= the maximum section
alignment. Align the PT_GNU_RELRO segment first. Subtract the maximum
page size if therer is still a 1-page gap.
PR ld/28743
PR ld/28819
* ldlang.c (lang_size_relro_segment_1): Rewrite.
* testsuite/ld-x86-64/pr28743-1.d: New file.
* testsuite/ld-x86-64/pr28743-1.s: Likewise.
* testsuite/ld-x86-64/x86-64.exp: Run pr28743-1.
Tom Tromey [Thu, 30 Dec 2021 19:13:52 +0000 (12:13 -0700)]
Always print the "host libthread-db" message to stdout
linux-thread-db.c has a bit of unusual code that unconditionally
prints a message, but decides whether to print to gdb_stdout or
gdb_stdlog based on a debug flag. It seems better to me to simply
always print this; and this is the only spot in gdb where we
conditionally pass gdb_stdout to one of the f*_unfiltered functions.
Tom Tromey [Fri, 31 Dec 2021 00:31:41 +0000 (17:31 -0700)]
Reduce explicit use of gdb_stdout
In an earlier version of the pager rewrite series, it was important to
audit unfiltered output calls to see which were truly necessary.
This is no longer necessary, but it still seems like a decent cleanup
to change calls to avoid explicitly passing gdb_stdout. That is,
rather than using something like fprintf_unfiltered with gdb_stdout,
the code ought to use plain printf_unfiltered instead.
This patch makes this change. I went ahead and converted all the
_filtered calls I could find, as well, for the same clarity.
Tom Tromey [Wed, 29 Dec 2021 21:16:02 +0000 (14:16 -0700)]
Sent timing stats to gdb_stdlog
This changes the time / space / symtab per-command statistics code to
send its output to gdb_stdlog rather than gdb_stdout. This seems
slightly more correct to me.
Mike Frysinger [Sun, 23 Jan 2022 17:53:22 +0000 (12:53 -0500)]
gold: drop old cygnus install hack
The gold subdir doesn't actually have a manual, so this hack doesn't
do anything. Plus the automake cygnus option was removed years ago
by Simon in d0ac1c44885daf68f631befa37e ("Bump to autoconf 2.69 and
automake 1.15.1"). So delete it here.
Mike Frysinger [Sun, 23 Jan 2022 17:44:24 +0000 (12:44 -0500)]
gas: drop old cygnus install hack
This was needed when gas was using the automake cygnus option, but
this was removed years ago by Simon in d0ac1c44885daf68f631befa37e
("Bump to autoconf 2.69 and automake 1.15.1"). So delete it here.
The info pages are already & still installed by default w/out it.
Tom Tromey [Sun, 23 Jan 2022 19:48:38 +0000 (12:48 -0700)]
Simplify some Rust expression-evaluation code
A few Rust operations do a bit of work in their 'evaluate' functions
and then call another function -- but are also the only caller. This
patch simplifies this code by removing the extra layer.
H.J. Lu [Sun, 23 Jan 2022 15:29:27 +0000 (07:29 -0800)]
bfd: Properly install library and header files
Rename bfdlib_LTLIBRARIES and bfdinclude_HEADERS to lib_LTLIBRARIES and
include_HEADERS to fix the missing installed library and header files in
bfd caused by
Mike Frysinger [Mon, 20 Dec 2021 07:05:31 +0000 (02:05 -0500)]
bfd: rename core.texi to corefile.texi
This is a generated file name from a correspondingly named C file.
Rename it to avoid unique build rules since there's no difference
to the generated manual.
Simon Marchi [Fri, 21 Jan 2022 16:33:29 +0000 (11:33 -0500)]
gdb: include gdbsupport/buildargv.h in ser-mingw.c
Fixes:
CXX ser-mingw.o
/home/simark/src/binutils-gdb/gdb/ser-mingw.c: In function ‘int pipe_windows_open(serial*, const char*)’:
/home/simark/src/binutils-gdb/gdb/ser-mingw.c:870:3: error: ‘gdb_argv’ was not declared in this scope
870 | gdb_argv argv (name);
| ^~~~~~~~
Andrew Burgess [Fri, 21 Jan 2022 12:49:48 +0000 (12:49 +0000)]
gdb/doc: fill in two missing @r
I noticed two places in the docs where we appear to be missing @r.
makeinfo seems to do the correct things despite these being
missing (at least, I couldn't see any difference in the pdf or info
output), but it doesn't hurt to have the @r in place.
Tom Tromey [Thu, 13 Jan 2022 16:48:18 +0000 (09:48 -0700)]
Avoid bad breakpoints with --gc-sections
We found a case where --gc-sections can cause gdb to set an invalid
breakpoint. In the included test case, gdb will set a breakpoint with
two locations, one of which is 0x0.
The code in lnp_state_machine::check_line_address is intended to
filter out this sort of problem, but in this case, the entire CU is
empty, causing unrelocated_lowpc==0x0 -- which circumvents the check.
It seems to me that if a CU is empty like this, then it is ok to
simply ignore the line table, as there won't be any locations anyway.
Respect `set print array-indexes' with Fortran arrays
Add `set print array-indexes' handling for Fortran arrays. Currently
the setting is ignored and indices are never shown.
Keep track of the most recent index handled so that any outstanding
repeated elements printed when the limit set by `set print elements' is
hit have the correct index shown.
Add `set print repeats' tests for C/C++ arrays, complementing one for
Fortran arrays and covering the different interpretation of the `set
print elements' setting in particular where the per-dimension count of
the elements handled is matched against the trigger rather than the
total element count as with Fortran arrays.
Implement `set print repeats' handling for Fortran arrays. Currently
the setting is ignored and always treated as if no limit was set.
Unlike the generic array walker implemented decades ago the Fortran one
is a proper C++ class. Rather than trying to mimic the old walker then,
which turned out a bit of a challenge where interacting with the `set
print elements' setting, write it entirely from scratch, by adding an
extra specialization handler method for processing dimensions other than
the innermost one and letting the specialization class call the `walk_1'
method from the handler as it sees fit. This way repeats can be tracked
and the next inner dimension recursed into as a need arises only, or
unconditionally in the base class.
Keep track of the dimension number being handled in the class rather as
a parameter to the walker so that it does not have to be passed across
by the specialization class.
Use per-dimension element count tracking, needed to terminate processing
early when the limit set by `set print elements' is hit. This requires
extra care too where the limit triggers exactly where another element
that is a subarray begins. In that case rather than recursing we need
to terminate processing or lone `(...)' would be printed. Additionally
if the skipped element is the last one in the current dimension we need
to print `...' by hand, because `continue_walking' won't print it at the
upper level, because it can see the last element has already been taken
care of.
Preserve the existing semantics of `set print elements' where the total
count of the elements handled is matched against the trigger level which
is unlike with the C/C++ array printer where the per-dimension element
count is used instead.
Amend existing test cases accordingly that rely on the current incorrect
behavior and explicitly request that there be no limit for printing
repeated elements there.
Add suitable test cases as well covering sliced arrays in particular.
Co-Authored-By: Andrew Burgess <andrew.burgess@embecosm.com>
Alan Modra [Wed, 19 Jan 2022 02:49:51 +0000 (13:19 +1030)]
PowerPC64 DT_RELR ELFv1
More fun with R_PPC64_NONE found in .opd. Fixed by the
allocate_dynrelocs and ppc64_elf_size_dynamic_sections changes, and
since we are doing ifunc, opd and SYMBOL_REFERENCES_LOCAL tests later,
don't duplicate that work in check_relocs.
* elf64-ppc.c (ppc64_elf_check_relocs): Remove opd and ifunc
conditions for rel_count.
(dec_dynrel_count): Likewise.
(allocate_dynrelocs): Test for opd and ifunc when allocating
relative relocs.
(ppc64_elf_size_dynamic_sections): Likewise.
Alan Modra [Wed, 19 Jan 2022 00:23:53 +0000 (10:53 +1030)]
PowerPC64 DT_RELR local GOT
Fixes another case where we end up with superfluous R_PPC64_NONE.
* elf64-ppc.c (ppc64_elf_size_dynamic_sections): Don't allocate
space for GOT relocs against non-TLS local syms when enable_dt_relr.
(ppc64_elf_layout_multitoc): Likewise.
Alan Modra [Tue, 18 Jan 2022 23:05:00 +0000 (09:35 +1030)]
Re: PowerPC64 DT_RELR
HJ: "There are 238 R_PPC64_NONEs in libc.so.6 alone."
Indeed, let's make them go away. I had the SYMBOL_REFERENCES_LOCAL
test in the wrong place. check_relocs is too early to know whether a
symbol is dynamic in a shared library. Lots of glibc symbols are made
local by version script, but that doesn't happen until
size_dynamic_sections.
* elf64-ppc.c (ppc64_elf_check_relocs): Don't count relative relocs
here depending on SYMBOL_REFERENCES_LOCAL.
(dec_dynrel_count): Likewise.
(allocate_dynrelocs): Do so here instead.
Add the threads_debug_printf and THREADS_SCOPED_DEBUG_ENTER_EXIT, which
use the logging infrastructure from gdbsupport/common-debug.h. Replace
all debug_print uses that are predicated by debug_threads with
threads_dethreads_debug_printf. Replace uses of the debug_enter and
debug_exit macros with THREADS_SCOPED_DEBUG_ENTER_EXIT, which serves
essentially the same purpose, but allows showing what comes between the
enter and the exit in an indented form.
Note that "threads" debug is currently used for a bit of everything in
GDBserver, not only threads related stuff. It should ideally be cleaned
up and separated logically as is done in GDB, but that's out of the
scope of this patch.
Simon Marchi [Mon, 17 Jan 2022 02:21:24 +0000 (21:21 -0500)]
gdbserver: turn debug_threads into a boolean
debug_threads is always used as a boolean. Except in ax.cc and
tracepoint.cc. These files have their own macros that use
debug_threads, and have a concept of verbosity level. But they both
have a single level, so it's just a boolean in the end.
Remove this concept of level. If we ever want to re-introduce it, I
think it will be better implemented in a more common location.
Change debug_threads to bool and adjust some users that were treating it
as an int.
Tom Tromey [Thu, 13 Jan 2022 01:11:53 +0000 (18:11 -0700)]
Simplify Ada catchpoints
All the Ada catchpoints use the same breakpoint_ops contents, because
the catchpoint itself records its kind. This patch simplifies the
code by removing the redundant ops structures.
Tom Tromey [Wed, 12 Jan 2022 23:13:23 +0000 (16:13 -0700)]
Move "catch exec" to a new file
The "catch exec" code is reasonably self-contained, and so this patch
moves it out of breakpoint.c (the second largest source file in gdb)
and into a new file, break-catch-exec.c.
Tom Tromey [Wed, 12 Jan 2022 23:03:33 +0000 (16:03 -0700)]
Move "catch fork" to a new file
The "catch fork" code is reasonably self-contained, and so this patch
moves it out of breakpoint.c (the second largest source file in gdb)
and into a new file, break-catch-fork.c.
Tom Tromey [Fri, 14 Jan 2022 15:29:42 +0000 (08:29 -0700)]
Split event_location into subclasses
event_location uses the old C-style discriminated union approach.
However, it's better to use subclassing, as this makes the code
clearer and removes some chances for error. This also enables future
cleanups to avoid manual memory management and copies.
Tom Tromey [Fri, 14 Jan 2022 14:33:33 +0000 (07:33 -0700)]
Remove EL_* macros from location.c
This patch removes the old-style EL_* macros from location.c. This
cleans up the code by itself, IMO, but also enables further cleanups
in subsequent patches.
Tom Tromey [Fri, 14 Jan 2022 14:24:18 +0000 (07:24 -0700)]
Remove a use of xfree in location.c
This small cleanup removes a use of xfree from location.c, by
switching to unique_xmalloc_ptr. One function is only used in
location.c, so it is made static. And, another function is changed to
avoid a copy.
Andrew Burgess [Wed, 12 Jan 2022 17:13:38 +0000 (17:13 +0000)]
gdb: preserve `|` in connection details string
Consider this GDB session:
$ gdb -q
(gdb) target remote | gdbserver - ~/tmp/hello.x
Remote debugging using | gdbserver - ~/tmp/hello.x
... snip ...
(gdb) info connections
Num What Description
* 1 remote gdbserver - ~/tmp/hello.x Remote target using gdb-specific protocol
(gdb) python conn = gdb.selected_inferior().connection
(gdb) python print(conn.details)
gdbserver - ~/tmp/hello.x
(gdb)
I think there are two things wrong here, first in the "What" column of
the 'info connections' output, I think the text should be:
remote | gdbserver - ~/tmp/hello.x
to correctly show the user how the connection was established. And in
a similar fashion, I think that the `details` string of the
gdb.TargetConnection object should be:
| gdbserver - ~/tmp/hello.x
This commit makes this change. Currently the '|' is detected and
removed in gdb/serial.c. The string passed to the pipe_ops
structure (from gdb/ser-pipe.c), doesn't then, contain the `|`, this
is instead implied by the fact that it is a pipes based implementation
of the serial_ops interface.
After this commit we still detect the `|` in gdb/serial.c, but we now
store the full string (including the `|`) in the serial::name member
variable.
For pipe based serial connections, this name is only used for
displaying the two fields I mention above, and in pipe_open (from
gdb/ser-pipe.c), and in pipe_open, we now know to skip over the `|`.
The benefit I see from this change is that GDB's output now more
accurately reflects the commands used to start a target, thus making
it easier for a user to understand what is going on.
Tiezhu Yang [Tue, 18 Jan 2022 07:17:57 +0000 (15:17 +0800)]
gdb: testsuite: print explicit test result for gdb.base/dfp-test.exp
In the current code, if decimal floating point is not supported for
this target, there is no binary file dfp-test, and also there is no
test result after execute the following commands:
$ make check-gdb TESTS="gdb.base/dfp-test.exp"
$ grep error gdb/testsuite/gdb.log
/home/loongson/gdb.git/gdb/testsuite/gdb.base/dfp-test.c:39:1: error: decimal floating point not supported for this target
[...]
$ cat gdb/testsuite/gdb.sum
[...]
Running target unix
Running /home/loongson/gdb.git/gdb/testsuite/gdb.base/dfp-test.exp ...
=== gdb Summary ===
[...]
With this patch:
$ make check-gdb TESTS="gdb.base/dfp-test.exp"
$ cat gdb/testsuite/gdb.sum
[...]
Running target unix
Running /home/loongson/gdb.git/gdb/testsuite/gdb.base/dfp-test.exp ...
UNSUPPORTED: gdb.base/dfp-test.exp: decimal floating point not supported for this target.
=== gdb Summary ===
# of unsupported tests 1
[...]
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Simon Marchi [Sun, 16 Jan 2022 03:13:07 +0000 (22:13 -0500)]
bfd/elf64-ppc.c: fix clang -Wbitwise-instead-of-logical warning in ppc64_elf_check_init_fini
I see this error with clang-14:
CC elf64-ppc.lo
/home/smarchi/src/binutils-gdb/bfd/elf64-ppc.c:13131:11: error: use of bitwise '&' with boolean operands [-Werror,-Wbitwise-instead-of-logical]
return (check_pasted_section (info, ".init")
~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Fix by replacing & with &&. But given that the check_pasted_section
function has side-effects and we want to make sure both calls are made,
assign to temporary variables before evaluating the `&&`.
Alan Modra [Sun, 21 Nov 2021 23:46:01 +0000 (10:16 +1030)]
PR28029, debuginfod tests
binutils/NEWS says of the change in --process-links semantics:
If other debug section display options are also enabled (eg
--debug-dump=info) then the contents of matching sections in both the main
file and the separate debuginfo file *will* be displayed. This is because in
most cases the debug section will only be present in one of the files.
Implying that debug info is dumped without --process-links. Indeed
that appears to be the case for readelf. This does the same for
objdump.
PR 28029
* objdump.c (dump_bfd): Do not exit early when !is_mainfile
&& !processlinks, instead just exclude non-debug output.
(dump_dwarf): Add is_mainfile parameter and pass to
dump_dwarf_section.
(dump_dwarf_section): Only display debug sections when
!is_mainfile and !process_links.
Alan Modra [Thu, 6 Jan 2022 21:19:16 +0000 (07:49 +1030)]
PowerPC64 DT_RELR
PowerPC64 takes a more traditional approach to DT_RELR than x86. Count
relative relocs in check_relocs, allocate space for them and output in
the usual places but not doing so when enable_dt_relr. DT_RELR is
sized in the existing ppc stub relaxation machinery, run via the
linker's ldemul_after_allocation hook. DT_RELR is output in the same
function that writes ppc stubs, run via ldemul_finish.
This support should be considered experimental.
bfd/
* elf64-ppc.c (struct ppc_local_dyn_relocs): Renamed from
ppc_dyn_relocs. Add rel_count field. Update uses.
(struct ppc_dyn_relocs): New. Replace all uses of elf_dyn_relocs.
(struct ppc_link_hash_table): Add relr_alloc, relr_count and
relr_addr.
(ppc64_elf_copy_indirect_symbol): Merge rel_count.
(ppc64_elf_check_relocs): Init rel_count for global and local syms.
(dec_dynrel_count): Change r_info param to reloc pointer. Update
all callers. Handle decrementing rel_count.
(allocate_got): Don't allocate space for relative relocs when
enable_dt_relr.
(allocate_dynrelocs): Likewise.
(ppc64_elf_size_dynamic_sections): Likewise. Handle srelrdyn.
(ppc_build_one_stub): Don't emit relative relocs on .branch_lt.
(compare_relr_address, append_relr_off): New functions.
(got_and_plt_relr_for_local_syms, got_and_plt_relr): Likewise.
(ppc64_elf_size_stubs): Size .relr.syn.
(ppc64_elf_build_stubs): Emit .relr.dyn.
(build_global_entry_stubs_and_plt): Don't output relative relocs
when enable_dt_relr.
(write_plt_relocs_for_local_syms): Likewise.
(ppc64_elf_relocate_section): Likewise.
binutils/
* testsuite/lib/binutils-common.exp (supports_dt_relr): Add
powerpc64.
ld/
* emulparams/elf64ppc.sh: Source dt-relr.sh.
* testsuite/ld-elf/dt-relr-2b.d: Adjust for powerpc.
* testsuite/ld-elf/dt-relr-2c.d: Likewise.
* testsuite/ld-elf/dt-relr-2d.d: Likewise.
* testsuite/ld-elf/dt-relr-2e.d: Likewise.
Alan Modra [Wed, 12 Jan 2022 13:12:23 +0000 (23:42 +1030)]
tweak __ehdr_start visibility and flags for check_relocs
bfd/
* elf-bfd.h (UNDEFWEAK_NO_DYNAMIC_RELOC): Test linker_def.
ld/
* ldelf.c (ldelf_before_allocation): Don't force __ehdr_start
local and hidden here..
* ldlang.c (lang_symbol_tweaks): ..do so here instead and set
def_regular and linker_def for check_relocs. New function
extracted from lang_process.
Jan Beulich [Mon, 17 Jan 2022 09:26:43 +0000 (10:26 +0100)]
x86: adjust struct instr_info field types
Now that this lives on the stack, let's have it be a little less
wasteful in terms of space. Switch boolean fields to "bool" (also when
this doesn't change their size) and also limit the widths of "rex",
"rex_used", "op_ad", and "op_index". Do a little bit of re-ordering as
well to limit the number of padding holes.
Jan Beulich [Mon, 17 Jan 2022 09:25:40 +0000 (10:25 +0100)]
x86: fold variables in memory operand index handling
There's no real need for the pseudo-boolean "haveindex" or for separate
32-bit / 64-bit index pointers. Fold them into a single "indexes" and
set that uniformly to AT&T names, compensating by emitting the register
name via oappend_maybe_intel().
The commit imported the commit 0e32a5aa8bc9 ("libiberty: Add support for
D `typeof(*null)' types") from the gcc repository. That commit includes
an update to libiberty/testsuite/d-demangle-expected, which updates a
test for the exact same mangled name:
I don't know anything about D, but give that the change was made by Iain
Buclaw, the D language maintainer, I trust him on that.
Fix our test by updating the expected output in the same way.
Note: it's not really useful to have all these D demangling tests in the
GDB testsuite, since there are demangling tests in libiberty. We should
consider removing them, but we first need to make sure that everything
that is covered in gdb/testsuite/gdb.dlang/demangle.exp is also covered
in libiberty/testsuite/d-demangle-expected.
gdb/testsuite: enable __INTEL_LLVM_COMPILER preprocessor in get_compiler_info
Intel Next Gen compiler defines preprocessor __INTEL_LLVM_COMPILER and provides
version info in __clang_version__ e.g. value: 12.0.0 (icx 2020.10.0.1113).
gdb/testsuite/ChangeLog:
2020-12-07 Abdul Basit Ijaz <abdul.b.ijaz@intel.com>
* lib/compiler.c: Add Intel next gen compiler pre-processor check.
* lib/compiler.cc: Ditto.
* lib/fortran.exp (fortran_main): Check Intel next gen compiler in
test_compiler_info.
Alan Modra [Fri, 14 Jan 2022 11:25:51 +0000 (21:55 +1030)]
PR28751 mbind2a / mbind2b regressions on powerpc*-linux
include/
* bfdlink.h (struct bfd_link_info): Add commonpagesize_is_set.
ld/
PR 28751
* emultempl/elf.em (handle_option): Set commonpagesize_is_set.
* ldelf.c (ldelf_after_parse): Don't error when only one of
-z max-page-size or -z common-page-size is given, correct the
other value to make it sane.
* testsuite/ld-elf/elf.exp (mbind2a, mbind2b): Do not pass
-z max-page-size.
Jan Beulich [Fri, 14 Jan 2022 09:56:06 +0000 (10:56 +0100)]
x86: share yet more VEX table entries with EVEX decoding
On top of prior similar work more opportunities have appeared in the
meantime. Note that this also happens to address the prior lack of
decoding of EVEX.L'L for VMOV{L,H}P{S,D} and VMOV{LH,HL}PS.
Jan Beulich [Fri, 14 Jan 2022 09:55:42 +0000 (10:55 +0100)]
x86: consistently use scalar_mode for AVX512-FP16 scalar insns
For some reason the original AVFX512F insns were not taken as a basis
here, causing unnecessary divergence. While not an active issue, it is
still relevant to note that OP_XMM() has special treatment of e.g.
scalar_mode (marking broadcast as invalid). Such would better be
consistent for all sufficiently similar insns.
projects / thirdparty / binutils-gdb.git / log
