Tom de Vries [Fri, 16 Sep 2022 14:40:56 +0000 (16:40 +0200)]
[gdb/tdep] Fix PowerPC IEEE 128-bit format arg passing
On a powerpc system with gcc 12 built to default to 128-bit IEEE long double,
I run into:
...
(gdb) print find_max_long_double_real(4, ldc1, ldc2, ldc3, ldc4)^M
$8 = 0 + 0i^M
(gdb) FAIL: gdb.base/varargs.exp: print \
find_max_long_double_real(4, ldc1, ldc2, ldc3, ldc4)
...
This is due to incorrect handling of the argument in ppc64_sysv_abi_push_param.
Fix this and similar cases, and expand the test-case to test handling of
homogeneous aggregates.
Tested on ppc64le-linux, power 10.
Co-Authored-By: Ulrich Weigand <uweigand@de.ibm.com> Tested-by: Carl Love <cel@us.ibm.com>
Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=29543
Tom de Vries [Fri, 16 Sep 2022 14:34:13 +0000 (16:34 +0200)]
[gdb/testsuite] Fix gdb.dwarf2/dw2-dir-file-name.exp for aarch64
[ Another attempt at fixing the problem described in commit cd919f5533c
("[gdb/testsuite] Fix gdb.dwarf2/dw2-dir-file-name.exp"). ]
When running the test-case gdb.dwarf2/dw2-dir-file-name.exp with
aarch64-linux, we run into:
...
(gdb) continue^M
Continuing.^M
^M
Breakpoint 2, compdir_missing__ldir_missing__file_basename () at \
tmp-dw2-dir-file-name.c:999^M
(gdb) FAIL: gdb.dwarf2/dw2-dir-file-name.exp: \
compdir_missing__ldir_missing__file_basename: continue to breakpoint: \
compdir_missing__ldir_missing__file_basename
...
The breakpoint set at compdir_missing__ldir_missing__file_basename_label,
address 0x400608 starts at a line entry:
...
CU: tmp-dw2-dir-file-name.c:
File name Line number Starting address View Stmt
tmp-dw2-dir-file-name.c 999 0x400608 x
tmp-dw2-dir-file-name.c 1000 0x40062c x
tmp-dw2-dir-file-name.c - 0x40062c
...
and therefore the breakpoint is printed without instruction address.
In contrast, for x86_64-linux, we have the breakpoint printed with instruction
address:
...
(gdb) continue^M
Continuing.^M
^M
Breakpoint 2, 0x004004c1 in compdir_missing__ldir_missing__file_basename () \
at tmp-dw2-dir-file-name.c:999^M
(gdb) PASS: gdb.dwarf2/dw2-dir-file-name.exp: \
compdir_missing__ldir_missing__file_basename: continue to breakpoint: \
compdir_missing__ldir_missing__file_basename
...
The breakpoint set at compdir_missing__ldir_missing__file_basename_label,
address 0x004004c1 doesn't start at a line entry:
...
CU: tmp-dw2-dir-file-name.c:
File name Line number Starting address View Stmt
tmp-dw2-dir-file-name.c 999 0x4004bd x
tmp-dw2-dir-file-name.c 1000 0x4004d3 x
tmp-dw2-dir-file-name.c - 0x4004d3
...
Fix this by:
- unifying behaviour between the archs by adding an explicit line number entry
for the address compdir_missing__ldir_missing__file_basename_label, making
the FAIL reproducible on x86_64-linux.
- expecting the breakpoint to be printed without instruction address.
Tom de Vries [Fri, 16 Sep 2022 13:53:47 +0000 (15:53 +0200)]
[gdb] Handle pending ^C after rl_callback_read_char
In completion tests in various test-cases, we've been running into these
"clearing input line" timeouts:
...
(gdb) $cmd^GPASS: gdb.gdb/unittest.exp: tab complete "$cmd"
FAIL: gdb.gdb/unittest.exp: tab complete "$cmd" (clearing input line) (timeout)
...
where $cmd == "maintenance selftest name_that_does_not_exist".
AFAIU, the following scenario happens:
- expect sends "$cmd\t"
- gdb detects the stdin event, and calls rl_callback_read_char until it
comes to handle \t
- readline interprets the \t as completion, tries to complete, fails to do so,
outputs a bell (^G)
- expect sees the bell, and proceeds to send ^C
- readline is still in the call to rl_callback_read_char, and stores the
signal in _rl_caught_signal
- readline returns from the call to rl_callback_read_char, without having
handled _rl_caught_signal
- gdb goes to wait for the next event
- expect times out waiting for "Quit", the expected reaction for ^C
Fix this by handling pending signals after each call to rl_callback_read_char.
The fix is only available for readline 8.x, if --with-system-readline provides
an older version, then the fix is disabled due to missing function
rl_check_signals.
Alan Modra [Fri, 16 Sep 2022 08:38:44 +0000 (18:08 +0930)]
PowerPC64 pcrel got relocs against local symbols
Not that anyone would want to indirect via the GOT when an address can
be loaded directly with pla, the following:
pld 3,x@got@pcrel
x:
leads to "Internal error in md_apply_fix", because the generic parts
of assembler fixup handling convert the fx_pcrel fixup to one without
a symbol. Stop that happening.
* config/tc-ppc.c (ppc_force_relocation): Add PLT_PCREL34 and
assorted GOT_PCREL34 relocs.
Tsukasa OI [Thu, 15 Sep 2022 04:06:09 +0000 (04:06 +0000)]
bfd, binutils, gas: Remove/mark unused variables
Clang generates a warning on unused (technically, written but not read
thereafter) variables. By the default configuration (with "-Werror"), it
causes a build failure (unless "--disable-werror" is specified).
This commit adds ATTRIBUTE_UNUSED attribute to some of them, which means
they are *possibly* unused (can be used but no warnings occur when
unused) and removes others.
Alan Modra [Wed, 14 Sep 2022 07:27:42 +0000 (16:57 +0930)]
looping in bfd_mach_o_fat_openr_next_archived_file
mach-o.c doesn't sanity check mach-o-fat archives, making it easy for
fuzzers to create an archive with mach_o_fat_archentry headers that
point to the same offset. bfd_mach_o_fat_openr_next_archived_file
uses the previous element offset to find its header, and thus the next
element. If two offsets are the same, any tool reading the archive
will get stuck. This patch rejects such archives, and any with
overlapping elements.
* mach-o.c (overlap_previous): New function.
(bfd_mach_o_fat_archive_p): Sanity check that elements do not
overlap each other or the file and archive headers.
Tsukasa OI [Sat, 10 Sep 2022 17:08:13 +0000 (17:08 +0000)]
bfd: Stop using -Wstack-usage=262144 when built with Clang
Some components of GNU Binutils will pass "-Wstack-usage=262144" when
"GCC >= 5.0" is detected. However, Clang does not support "-Wstack-usage",
despite that related configuration part in bfd/warning.m4 handles the latest
Clang (15.0.0 as of this writing) as "GCC >= 5.0".
The option "-Wstack-usage" was ignored when the first version of Clang is
released but even this "ignoring" behavior is removed before Clang 4.0.0.
So, if we give Clang "-Wstack-usage=262144", it generates a warning, making
the build failure.
This commit checks "__clang__" macro to prevent adding the option if the
compiler is identified as Clang.
bfd/ChangeLog:
* warning.m4: Stop appending "-Wstack-usage=262144" option when
compiled with Clang.
* configure: Regenerate.
Alan Modra [Sat, 10 Sep 2022 03:28:44 +0000 (12:58 +0930)]
PR29540, R_PPC64_NONE in .rela.dyn when linking Linux vdso
PR 29540
* elf64-ppc.c (allocate_dynrelocs): Don't alloc space for relocs
against discarded sections.
(ppc64_elf_size_dynamic_sections): Use standard test for discarded
sections.
* elf32-ppc.c (allocate_dynrelocs): Don't alloc space for relocs
against discarded sections.
(ppc_elf_size_dynamic_sections): Use standard test for discarded
sections.
objdump: '-S' should trigger search for separate debuginfo.
Add with_source_code to the command line options that trigger
might_need_separate_debug_info and dump_any_debugging. This helps
'objdump -S' download missing files via debuginfod without the need for
specifying extra command line options like '-L'.
Bruno Larsen [Wed, 20 Jul 2022 19:44:37 +0000 (16:44 -0300)]
gdb/testsuite: Update gdb.base/so-impl-ld.exp
gdb.base/so-impl-ld.exp was setup assuming that the compiler would add
epilogue information and that GDB would stop in the } line. This would
make clang tests fail like so:
step^M
solib_main (arg=10000) at ../../../common/git-repos/binutils-gdb/gdb/testsuite/gdb.base/solib1.c:7^M
7|__ return arg*arg;|__|___/* HERE */^M
(gdb) PASS: gdb.base/so-impl-ld.exp: step into solib call
next^M
main () at ../../../common/git-repos/binutils-gdb/gdb/testsuite/gdb.base/so-impl-ld.c:22^M
22|_ return 0;^M
(gdb) FAIL: gdb.base/so-impl-ld.exp: step in solib call
next^M
0x00007ffff7cef560 in __libc_start_call_main () from /lib64/libc.so.6^M
(gdb) FAIL: gdb.base/so-impl-ld.exp: step out of solib call
This patch changes it so solib_main has 2 lines where GDB can stop
regardless of compiler choices, and updates the exp file to
generically deal with unknown number of steps until leaving that
function.
Bruno Larsen [Wed, 20 Jul 2022 19:44:26 +0000 (16:44 -0300)]
gdb/testsuite: introduce gdb_step_until
Currently, GDB's testsuite uses a set amount of step commands to exit
functions. This is a problem if a compiler emits different epilogue
information from gcc, or emits no epilogue information at all. It was
most noticeable if Clang was used to test GDB.
To fix this unreliability, this commit introduces a new proc that will
step the inferior until it is stopped at a line that matches the given
regexp, or until it steps too many times - defined as an optional
argument. If the line is found, it shows up as a single PASS in the
test, and if the line is not found, a single FAIL is emitted.
This patch only introduces this proc, but does not add it to any
existing tests, these will be introduced in the following commit.
Bruno Larsen [Wed, 20 Jul 2022 19:44:36 +0000 (16:44 -0300)]
explicitly test for stderr in gdb.base/dprintf.exp
Not all compilers add stderr debug information when compiling a
program. Clang, for instance, prefers to add nothing from standard
libraries and let an external debug package have this information.
Because of this, gdb.base/dprintf.exp was failing when GDB attempted to
use dprintf as a call to fprintf(stderrr, ...), like this:
(gdb) PASS: gdb.base/dprintf.exp: call: fprintf: set dprintf style to call
continue
Continuing.
kickoff 1234
also to stderr 1234
'stderr' has unknown type; cast it to its declared type
(gdb) FAIL: gdb.base/dprintf.exp: call: fprintf: 1st dprintf (timeout)
To avoid this false positive, we explicitly test to see if
the compiler has added information about stderr at all, and abort
testing dprintf as an fprintf call if it is unavailable.
Jiangshuai Li [Tue, 13 Sep 2022 03:20:54 +0000 (11:20 +0800)]
gdbserver/csky add csky gdbserver support
Add new files:
gdb/arch/csky.c
gdb/arch/csky.h
gdb/features/cskyv2-linux.c
gdbserver/linux-csky-low.cc
1. In gdb/arch/csky.c file, add function "csky_create_target_description()"
for csky_target::low_arch_setup(). later, it can be used for csky native gdb.
2. In gdb/features/cskyv2-linux.c file, create target_tdesc for csky, include
gprs, pc, hi, lo, float, vector and float control registers.
3. In gdbserver/linux-csky-low.cc file, using PTRACE_GET/SET_RGESET to
get/set registers. The main data structures in asm/ptrace.h are:
struct pt_regs {
unsigned long tls;
unsigned long lr;
unsigned long pc;
unsigned long sr;
unsigned long usp;
/*
* a0, a1, a2, a3:
* r0, r1, r2, r3
*/
unsigned long orig_a0;
unsigned long a0;
unsigned long a1;
unsigned long a2;
unsigned long a3;
/*
* r4 ~ r13
*/
unsigned long regs[10];
/* r16 ~ r30 */
unsigned long exregs[15];
unsigned long rhi;
unsigned long rlo;
unsigned long dcsr;
};
struct user_fp {
unsigned long vr[96];
unsigned long fcr;
unsigned long fesr;
unsigned long fid;
unsigned long reserved;
};
Tom Tromey [Wed, 17 Aug 2022 17:47:17 +0000 (11:47 -0600)]
Use checked_static_cast in more places
I went through all the uses of dynamic_cast<> in gdb, looking for ones
that could be replaced with checked_static_cast. This patch is the
result. Regression tested on x86-64 Fedora 34.
Peter Bergner [Mon, 12 Sep 2022 19:56:20 +0000 (14:56 -0500)]
ppc: Document the -mfuture and -Mfuture options and make them usable
The -mfuture and -Mfuture options which are used for adding potential
new ISA instructions were not documented. They also lacked a bitmask
so new instructions could not be enabled by those options. Fixed.
Bruno Larsen [Wed, 20 Jul 2022 19:44:34 +0000 (16:44 -0300)]
add xfails to gdb.base/complex-parts.exp when testing with clang
clang doesn't add encoding to the name of complex variables, only says
that the type name is complex, making the relevant tests fail.
This patch adds the xfails to the tests that expect the variable name to
include it.
These symbols are local static variables, and "main" is the name of
the function they are defined in. GCC instead appends a sequence
number to the linkage name:
$ nm -A call-ar-st.gcc | grep integer_
call-ar-st/call-ar-st:00000000000061a0 b integer_array.3968
$ nm -A call-ar-st.clang | grep integer_
call-ar-st:00000000004061a0 b main.integer_array
This commit changes the testcase to accept both outputs, as they are
functionally identical.
Co-Authored-By: Pedro Alves <pedro@palves.net>
Change-Id: Iaf2ccdb9d5996e0268ed12f595a6e04b368bfcb4
Bruno Larsen [Wed, 20 Jul 2022 19:44:31 +0000 (16:44 -0300)]
update gdb.base/info-program.exp to not fail with clang
The test specifically mentions that it doesn't care where the program
stops, however it was still testing for a specific location. The clang
compiler emits different line information for epilogue, so GDB reports a
different stopping location, depending on the used compiler. With this
patch the test works even with clang.
Bruno Larsen [Wed, 20 Jul 2022 19:44:30 +0000 (16:44 -0300)]
gdb/testsuite: change gdb.base/nodebug.exp to not fail with clang
Clang organizes the variables differently to gcc in the original version
of this code, leading to the following differences when testing
p (int*) &dataglobal + 1
gcc:
$16 = (int *) 0x404034 <datalocal>
clang:
$16 = (int *) 0x404034 <dataglobal8>
However, since the important part of this test doesn't seem to be which
symbol is linked, but rather if GDB is correctly increasing the
address. This test was changed to actually measure address changes,
instead of assuming the ordering and naming of symbols.
Co-Authored-By: Andrew Burgess <aburgess@redhat.com>
Martin Storsjö [Tue, 6 Sep 2022 15:39:07 +0000 (18:39 +0300)]
ld: pe: Apply review suggestions on the existing exports/imports arrays
Use a separate explicit max_exports/imports field, instead of
deducing it from the number of allocated elements. Use a named
constant for the incremental growth of the array.
Use bool instead of int for boolean values.
Remove an unnecessary if statement/scope in the def_file_free
function.
Add more verbose comments about parameters, and about insertion
into an array of structs.
Generally use unsigned integers for all array indices and sizes.
The num_exports/imports fields are kept as is as signed integers,
since changing them to unsigned would require a disproportionate
amount of changes ti pe-dll.c to avoid comparisons between signed
and unsigned.
Simply use xrealloc instead of a check and xmalloc/xrealloc;
xrealloc can take NULL as the first parameter (and does a similar
check internally). (This wasn't requested in review though,
but noticed while working on the code.)
Martin Storsjö [Fri, 2 Sep 2022 09:22:29 +0000 (12:22 +0300)]
ld: pe: Improve performance of object file exclude symbol directives
Store the list of excluded symbols in a sorted list, speeding up
checking for duplicates when inserting new entries.
This is done in the same way as is done for exports and imports
(while the previous implementation was done with a linked list,
based on the implementation for aligncomm).
When linking object files with excluded symbols, there can potentially
be very large numbers of excluded symbols (just like builds with
exports can have a large number of exported symbols).
This improves the link performance somewhat, when linking with large
numbers of excluded symbols.
The later actual use of the excluded symbols within pe-dll.c
handles them via an unordered linked list still, though.
Tom de Vries [Mon, 12 Sep 2022 08:05:18 +0000 (10:05 +0200)]
[gdb] Fix abort in selftest run_on_main_thread with ^C
When running selftest run_on_main_thread and pressing ^C, we can run into:
...
Running selftest run_on_main_thread.
terminate called without an active exception
Fatal signal: Aborted
...
The selftest function looks like this:
...
static void
run_tests ()
{
std::thread thread;
done = false;
{
gdb::block_signals blocker;
thread = std::thread (set_done);
}
while (!done && gdb_do_one_event () >= 0)
;
/* Actually the test will just hang, but we want to test
something. */
SELF_CHECK (done);
thread.join ();
}
...
The error message we see is due to the destructor of thread being called while
thread is joinable.
This is supposed to be taken care of by thread.join (), but the ^C prevents
that one from being called, while the destructor is still called.
Fix this by ensuring thread.join () is called (if indeed required) before the
destructor using SCOPE_EXIT.
Tom de Vries [Mon, 12 Sep 2022 08:05:18 +0000 (10:05 +0200)]
[gdb/testsuite] Fix gdb.dwarf2/dw2-dir-file-name.exp for ppc64le
In commit cd919f5533c ("[gdb/testsuite] Fix
gdb.dwarf2/dw2-dir-file-name.exp"), I made gdb.dwarf2/dw2-dir-file-name.exp
independent of prologue analyzers, using this change:
...
- gdb_breakpoint $func
+ gdb_breakpoint *$func
...
That however caused a regression on ppc64le. For PowerPC, as described in the
ELFv2 ABI, a function can have a global and local entry point.
Setting a breakpoint on *$func effectively creates a breakpoint for the global
entry point, so if the function is entered through the local entry point, the
breakpoint doesn't trigger.
Fix this by reverting commit cd919f5533c, and setting the breakpoint on
${func}_label instead.
Tom de Vries [Mon, 12 Sep 2022 08:05:18 +0000 (10:05 +0200)]
[gdb/testsuite] Fix gdb.dwarf2/dw2-dir-file-name.exp with clang
When running test-case gdb.dwarf2/dw2-dir-file-name.exp with clang, we run
into:
...
(gdb) break *compdir_missing__ldir_missing__file_basename^M
Breakpoint 2 at 0x400580^M
(gdb) continue^M
Continuing.^M
^M
Breakpoint 2, 0x0000000000400580 in \
compdir_missing.ldir_missing.file_basename ()^M
(gdb) FAIL: gdb.dwarf2/dw2-dir-file-name.exp: \
compdir_missing__ldir_missing__file_basename: continue to breakpoint: \
compdir_missing__ldir_missing__file_basename
...
The problem is that the test-case uses labels outside functions, which is know
to cause problem with clang, as documented in the comment for proc
function_range.
Fix this by using get_func_info instead.
Tested on x86_64-linux, with both gcc 7.5.0 and clang 13.0.0.
Jan Beulich [Mon, 12 Sep 2022 06:19:55 +0000 (08:19 +0200)]
x86: avoid i386_dis_printf()'s staging area for a fair part of output
While PR binutils/29483 has now been addressed differently, this
originally proposed change still has its merits: Avoiding vsnprintf()
for typically far more than half of the overall output results in a 2-3%
performance gain in my testing (with debug builds of objdump, libbfd,
and libopcodes).
With that part of output no longer using staging_area[], the array also
doesn't need to be quite as large anymore (the largest presently used
size is 27, from "64-bit address is disabled").
While limiting the scope of "res" it became apparent that
- no caller cares about the function's return value,
- the comment about the return value was wrong,
- a particular positive return value would have been meaningless to the
caller.
Therefore convert the function to return "void" at the same time.
Nelson Chu [Mon, 12 Sep 2022 01:26:52 +0000 (09:26 +0800)]
RISC-V: PR28509, the default visibility symbol cannot be referenced by R_RISCV_JAL.
When generating the shared object, the default visibility symbols may bind
externally, which means they will be exported to the dynamic symbol table,
and are preemptible by default. These symbols cannot be referenced by the
non-pic R_RISCV_JAL and R_RISCV_RVC_JUMP. However, consider that linker
may relax the R_RISCV_CALL relocations to R_RISCV_JAL or R_RISCV_RVC_JUMP,
if these relocations are relocated to the plt entries, then we won't report
error for them. Perhaps we also need the similar checks for the
R_RISCV_BRANCH and R_RISCV_RVC_BRANCH relocations.
After applying this patch, and revert the following glibc patch,
riscv: Fix incorrect jal with HIDDEN_JUMPTARGET
https://sourceware.org/git/?p=glibc.git;a=commit;h=68389203832ab39dd0dbaabbc4059e7fff51c29b
I get the expected errors as follows,
ld: relocation R_RISCV_RVC_JUMP against `__sigsetjmp' which may bind externally can not be used when making a shared object; recompile with -fPIC
ld: relocation R_RISCV_JAL against `exit' which may bind externally can not be used when making a shared object; recompile with -fPIC
Besides, we also have similar changes for libgcc,
RISC-V: jal cannot refer to a default visibility symbol for shared object
https://github.com/gcc-mirror/gcc/commit/45116f342057b7facecd3d05c2091ce3a77eda59
bfd/
pr 28509
* elfnn-riscv.c (riscv_elf_relocate_section): Report errors when
makeing a shard object, and the referenced symbols of R_RISCV_JAL
relocations are default visibility. Besides, we should handle most
of the cases here, so don't need the unresolvable check later for
R_RISCV_JAL and R_RISCV_RVC_JUMP.
ld/
pr 28509
* testsuite/ld-riscv-elf/ld-riscv-elf.exp: Updated.
* testsuite/ld-riscv-elf/lib-nopic-01a.s: Removed.
* testsuite/ld-riscv-elf/lib-nopic-01b.d: Likewise.
* testsuite/ld-riscv-elf/lib-nopic-01b.s: Likewise.
* testsuite/ld-riscv-elf/shared-lib-nopic-01.d: New testcase.
* testsuite/ld-riscv-elf/shared-lib-nopic-01.s: Likewise.
* testsuite/ld-riscv-elf/shared-lib-nopic-02.d: Likewise.
* testsuite/ld-riscv-elf/shared-lib-nopic-02.s: Likewise.
* testsuite/ld-riscv-elf/shared-lib-nopic-03.d: Likewise.
* testsuite/ld-riscv-elf/shared-lib-nopic-03.s: Likewise.
* testsuite/ld-riscv-elf/shared-lib-nopic-04.d: Likewise.
* testsuite/ld-riscv-elf/shared-lib-nopic-04.s: Likewise.
Tom de Vries [Sun, 11 Sep 2022 07:01:03 +0000 (09:01 +0200)]
[gdb/symtab] Fix handling of DW_TAG_unspecified_type
Currently, the test-case contained in this patch fails:
...
(gdb) p (int) foo ()^M
Invalid cast.^M
(gdb) FAIL: gdb.dwarf2/dw2-unspecified-type.exp: p (int) foo ()
...
because DW_TAG_unspecified_type is translated as void.
There's some code in read_unspecified_type that marks the type as stub, but
that's only active for ada:
...
if (cu->lang () == language_ada)
type->set_is_stub (true);
...
Fix this by:
- marking the type as a stub for all languages, and
- handling the stub return type case in call_function_by_hand_dummy.
Tsukasa OI [Mon, 5 Sep 2022 07:54:51 +0000 (07:54 +0000)]
RISC-V: Fix vector CSR requirements
Vector CSRs are also required on smaller vector subsets.
Not only that the most of vector CSRs are general purpose (and must be
accessible for every vector subsets), current minimum vector subset 'Zve32x'
requires fixed point arithmetic, making remaining non-general purpose
(fixed point arithmetic only) CSRs mandatory for such subsets.
So, those CSRs must be accessible from 'Zve32x', not just from 'V'.
This commit fixes this issue which caused CSR accessibility warnings.
gas/ChangeLog:
* config/tc-riscv.c (riscv_csr_address): Change vector CSR
requirement from 'V' to 'Zve32x'.
* testsuite/gas/riscv/csr-version-1p9p1.l: Change vector CSR
requirement from 'V' to 'Zve32x'.
* testsuite/gas/riscv/csr-version-1p10.l: Likewise.
* testsuite/gas/riscv/csr-version-1p11.l: Likewise.
* testsuite/gas/riscv/csr-version-1p12.l: Likewise.
Carl Love [Thu, 8 Sep 2022 15:13:03 +0000 (15:13 +0000)]
Fix hardware watchpoint check in test gdb.base/watchpoint-reuse-slot.exp
This test generates 48 failures on Power 9 when testing with HW watchpoints
enabled. Note HW watchpoint support is disabled on Power 9 due to a HW bug.
The skip_hw_watchpoint_tests proc must be used to correctly determine
if the processor supports HW watchpoints.
This patch replaces the [target_info exists gdb,no_hardware_watchpoints]
with the skip_hw_watchpoint_tests check.
This patch was tested on Power 9, Power 10 and X86-64 with no regressions.
Nick Clifton [Thu, 8 Sep 2022 11:43:33 +0000 (12:43 +0100)]
Gas generated incorrect debug info (top-level DW_TAG_unspecified_type DIE)
PR 29559
* dwarf2dbg.c (out_debug_info): Place DW_TAG_unspecified_type at
the end of the list of children, not at the start of the CU
information.
* testsuite/gas/elf/dwarf-3-func.d: Update expected output.
* testsuite/gas/elf/dwarf-5-func-global.d: Likewise.
* testsuite/gas/elf/dwarf-5-func-local.d: Likewise.
* testsuite/gas/elf/dwarf-5-func.d: Likewise.
Tsukasa OI [Thu, 8 Sep 2022 02:54:15 +0000 (02:54 +0000)]
gdbsupport: Fix config.status dependency
Commit 171fba11ab27 ("Make GDBserver abort on internal error in development mode")
created a new substitution CONFIG_STATUS_DEPENDENCIES but this is used by
Makefile.in (which is not regenerated by that commit). After regenerating
it, it is found that CONFIG_STATUS_DEPENDENCIES value is not valid, making
gdbsupport fail to build.
Since the CONFIG_STATUS_DEPENDENCIES value is used in the Makefile, macro
substitution must have a Makefile format but commit 171fba11ab27 used shell
format "$srcdir/../bfd/development.sh".
This commit fixes this issue by substituting "$srcdir" (shell format) to
"$(srcdir)" (Makefile format). It preserves the dependency as Pedro
intended and fixes the build problem.
It also regenerates corresponding files with the maintainer mode.
Tom de Vries [Wed, 7 Sep 2022 17:14:17 +0000 (19:14 +0200)]
[gdb/testsuite] xfail gdb.ada/O2_float_param.exp for aarch64 and gcc 7.5.0
On aarch64-linux, with gcc 7.5.0, we run into:
...
(gdb) frame^M
#0 callee.increment (val=99.0, val@entry=9.18340949e-41, msg=...) at \
callee.adb:21^M
21 if Val > 200.0 then^M
(gdb) FAIL: gdb.ada/O2_float_param.exp: scenario=all: frame
...
The problem is a GCC bug, filed as "PR98148 - [AArch64] Wrong location
expression for function entry values" (
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98148 ).
Tom de Vries [Wed, 7 Sep 2022 09:29:11 +0000 (11:29 +0200)]
[gdb/testsuite] Fix gdb.ada/access_tagged_param.exp for aarch64
On aarch64-linux, I run into:
...
Breakpoint 2, pck.inspect (obj=0x430eb0 \
<system.pool_global.global_pool_object>, <objL>=0) at pck.adb:17^M
17 procedure Inspect (Obj: access Top_T'Class) is^M
(gdb) FAIL: gdb.ada/access_tagged_param.exp: continue
...
while on x86_64-linux, I see:
...
Breakpoint 2, pck.inspect (obj=0x62b2a0, <objL>=2) at pck.adb:19^M
19 null;^M
(gdb) PASS: gdb.ada/access_tagged_param.exp: continue
...
Note the different line numbers, 17 vs 19.
The difference comes from the gdbarch_skip_prologue implementation.
The amd64_skip_prologue implementation doesn't use gcc line numbers, and falls
back to the architecture-specific prologue analyzer, which correctly skips
past the prologue, to address 0x4022f7:
... 00000000004022ec <pck__inspect>:
4022ec: 55 push %rbp
4022ed: 48 89 e5 mov %rsp,%rbp
4022f0: 48 89 7d f8 mov %rdi,-0x8(%rbp)
4022f4: 89 75 f4 mov %esi,-0xc(%rbp)
4022f7: 90 nop
4022f8: 90 nop
4022f9: 5d pop %rbp
4022fa: c3 ret
...
The aarch64_skip_prologue implementation does use gcc line numbers, which are:
...
File name Line number Starting address View Stmt
pck.adb 17 0x402580 x
pck.adb 17 0x402580 1 x
pck.adb 19 0x40258c x
pck.adb 20 0x402590 x
...
and which are represented like this internally in gdb:
...
INDEX LINE ADDRESS IS-STMT PROLOGUE-END
0 17 0x0000000000402580 Y
1 17 0x0000000000402580 Y
2 19 0x000000000040258c Y
3 20 0x0000000000402590 Y
4 END 0x00000000004025a0 Y
...
The second entry is interpreted as end-of-prologue, so 0x402580 is used, while
the actual end of the prologue is at 0x40258c:
... 0000000000402580 <pck__inspect>:
402580: d10043ff sub sp, sp, #0x10
402584: f90007e0 str x0, [sp, #8]
402588: b90007e1 str w1, [sp, #4]
40258c: d503201f nop
402590: d503201f nop
402594: 910043ff add sp, sp, #0x10
402598: d65f03c0 ret
40259c: d503201f nop
...
Note that the architecture-specific prologue analyzer would have gotten this
right:
...
(gdb) p /x aarch64_analyze_prologue (gdbarch, pc, pc + 128, 0)
$2 = 0x40258c
...
Fix the FAIL by making the test-case more robust against problems in prologue
skipping, by setting the breakpoint on line 19 instead.
Luis Machado [Wed, 3 Aug 2022 23:00:26 +0000 (00:00 +0100)]
Fix endianness handling for arm record self tests
v2:
- Add 32-bit Arm instruction selftest
- Refactored abstract memory reader into abstract instruction reader
- Adjusted code to use templated type and to use host endianness as
opposed to target endianness.
The arm record tests handle 16-bit and 32-bit thumb instructions, but the
code is laid out in a way that handles the 32-bit thumb instructions as
two 16-bit parts.
This is fine, but it is prone to host-endianness issues given how the two
16-bit parts are stored and how they are accessed later on. Arm is
little-endian by default, so running this test with a GDB built with
--enable-targets=all and on a big endian host will run into the following:
Running selftest arm-record.
Process record and replay target doesn't support syscall number -2036195
Process record does not support instruction 0x7f70ee1d at address 0x0.
Self test failed: self-test failed at ../../binutils-gdb/gdb/arm-tdep.c:14482
It turns out the abstract memory reader class is more generic than it needs to
be, and we can simplify the code a bit by assuming we have a simple instruction
reader that only reads up to 4 bytes, which is the length of a 32-bit
instruction.
Instead of returning a bool, we return instead the instruction that has been
read. This way we avoid having to deal with the endianness conversion, and use
the host endianness instead. The Arm selftests can be executed on non-Arm
hosts.
While at it, Tom suggested adding a 32-bit Arm instruction selftest to increase
the coverage of the selftests.
Tom de Vries [Wed, 7 Sep 2022 07:59:12 +0000 (09:59 +0200)]
[gdb/testsuite] Use prototype to call libc functions
On openSUSE Tumbleweed (using glibc 2.36), I run into:
...
(gdb) print /d (int) munmap (4198400, 4096)^M
Invalid cast.^M
(gdb) FAIL: gdb.base/break-main-file-remove-fail.exp: cmdline: \
get integer valueof "(int) munmap (4198400, 4096)"
...
The problem is that after starting the executable, the symbol has type
"void (*) (void)":
...
(gdb) p munmap
$1 = {<text variable, no debug info>} 0x401030 <munmap@plt>
(gdb) start
...
(gdb) p munmap
$2 = {void (void)} 0x7ffff7feb9a0 <__GI_munmap>
...
which causes the "Invalid cast" error.
Looking at the debug info for glibc for symbol __GI_munmap:
...
<0><189683>: Abbrev Number: 1 (DW_TAG_compile_unit)
<189691> DW_AT_name : ../sysdeps/unix/syscall-template.S
<189699> DW_AT_producer : GNU AS 2.39.0
<1><1896ae>: Abbrev Number: 2 (DW_TAG_subprogram)
<1896af> DW_AT_name : __GI___munmap
<1896b3> DW_AT_external : 1
<1896b4> DW_AT_low_pc : 0x10cad0
<1896bc> DW_AT_high_pc : 37
...
that's probably caused by this bit (or similar bits for other munmap aliases).
This is fixed in gas on trunk by commit 5578fbf672e ("GAS: Add a return type
tag to DWARF DIEs generated for function symbols").
Work around this (for say gas 2.39) by explicitly specifying the prototype for
munmap.
Likewise for getpid in a couple of other test-cases.
Currently objdump -S is not able to make use files downloaded from debuginfod.
This is due to bfd_find_nearest_line_discriminator being unable to locate any
separate debuginfo files in the debuginfod cache. Additionally objdump lacked
a call to debuginfod_find_source in order to download missing source files.
Fix this by using bfd_find_nearest_line_with_alt instead of
bfd_find_nearest_line_discriminator. Also add a call to
debuginfod_find_source in order to download missing source files.
bfd_find_nearest_line_with_alt functions like bfd_find_nearest_line with
the addition of a parameter for specifying the filename of a supplementary
debug file such as one referenced by .gnu_debugaltlink or .debug_sup.
This patch focuses on implementing bfd_find_nearest_line_with_alt
support for ELF/DWARF2 .gnu_debugaltlink. For other targets this
function simply sets the invalid_operation bfd_error.
Tsukasa OI [Tue, 30 Aug 2022 12:21:13 +0000 (12:21 +0000)]
gdb: Add non-enum disassembler options
This is paired with "opcodes: Add non-enum disassembler options".
There is a portable mechanism for disassembler options and used on some
architectures:
- ARC
- Arm
- MIPS
- PowerPC
- RISC-V
- S/390
However, it only supports following forms:
- [NAME]
- [NAME]=[ENUM_VALUE]
Valid values for [ENUM_VALUE] must be predefined in
disasm_option_arg_t.values. For instance, for -M cpu=[CPU] in ARC
architecture, opcodes/arc-dis.c builds valid CPU model list from
include/elf/arc-cpu.def.
Tom de Vries [Tue, 6 Sep 2022 08:15:01 +0000 (10:15 +0200)]
[gdb/symtab] Support .debug_names section with TUs in .debug_info
When running test-case gdb.cp/cpexprs-debug-types.exp on target board
cc-with-debug-names/gdb:debug_flags=-gdwarf-5, we get an executable with
a .debug_names section, but no .debug_types section. For dwarf-5, the TUs
are no longer put in a separate unit, but instead they're put in the
.debug_info section.
When loading the executable, the .debug_names section is silently ignored
because of this check in dwarf2_read_debug_names:
...
if (map->tu_count != 0)
{
/* We can only handle a single .debug_types when we have an
index. */
if (per_bfd->types.size () != 1)
return false;
...
which triggers because per_bfd->types.size () == 0.
The intention of the check is to make sure we don't have more that one
.debug_types section, as can happen in a object file (see PR12984):
...
$ grep "\.debug_types" 11.s
.section .debug_types,"G",@progbits,wt.75c042c23a9a07ee,comdat
.section .debug_types,"G",@progbits,wt.c59c413bf50a4607,comdat
...
Fix this by:
- changing the check condition to "per_bfd->types.size () > 1", and
- handling per_bfd->types.size () == 0.
Tom de Vries [Tue, 6 Sep 2022 08:15:00 +0000 (10:15 +0200)]
[gdb/testsuite] Add gdb.dwarf2/debug-names-tu.exp
Add a test-case gdb.dwarf2/debug-names-tu.exp, that uses the dwarf assembler
to specify a .debug_names index with the TU list referring to a TU from the
.debug_types section.
This is intended to produce something similar to:
...
$ gcc -g -fdebug-types-section ~/hello.c -gdwarf-4
$ gdb-add-index -dwarf-5 a.out
...
Tsukasa OI [Tue, 30 Aug 2022 12:20:30 +0000 (12:20 +0000)]
opcodes: Add non-enum disassembler options
This is paired with "gdb: Add non-enum disassembler options".
There is a portable mechanism for disassembler options and used on some
architectures:
- ARC
- Arm
- MIPS
- PowerPC
- RISC-V
- S/390
However, it only supports following forms:
- [NAME]
- [NAME]=[ENUM_VALUE]
Valid values for [ENUM_VALUE] must be predefined in
disasm_option_arg_t.values. For instance, for -M cpu=[CPU] in ARC
architecture, opcodes/arc-dis.c builds valid CPU model list from
include/elf/arc-cpu.def.
changes were made to the simulator source to handle the new libopcodes
disassembler styling API.
Unfortunately, these changes broke building GDB with the erc32 (sparc)
simulator, like this:
../src/configure --target=sparc-linux
make all-gdb
....
/usr/bin/ld: ../sim/erc32/libsim.a(interf.o): in function `sim_open':
/tmp/build/sim/../../src/sim/erc32/interf.c:247: undefined reference to `fprintf_styled'
collect2: error: ld returned 1 exit status
The problem is that in commit 7b01c1cc1d11 the fprintf_styled function
was added into sis.c. This file is only used when building the 'run'
binary, that is, the standalone simulator, and is not included in the
libsim.a library.
Now, the obvious fix would be to move fprintf_styled into libsim.a,
however, that turns out to be tricky.
The erc32 simulator currently has two copies of the function run_sim,
one in sis.c, and one in interf.c, both of these copies are global.
Currently, the 'run' binary links fine, though I suspect this might be
pure luck. When I tried moving fprintf_styled into interf.c, I ran
into multiple-definition (of run_sim) errors. I suspect that by
requiring the linker to pull in fprintf_styled from libsim.a I was
changing the order in which symbols were loaded, and the linker was
now seeing both copies of run_sim, while currently we only see one
copy.
The ideal solution of course, would be to merge the two similar, but
slightly different copies of run_sim, and just use the one copy. Then
we could safely move fprintf_styled into interf.c too, and all would
be good.
But I don't have time right now to start debugging the erc32
simulator, so I wanted a solution that fixes the build without
introducing multiple definition errors.
The easiest solution I think is to just have two copies of
fprintf_styled, one in sis.c, and one in interf.c. Unlike run_sim,
these two copies are both static, so we will not run into multiple
definition issues with this function. The functions themselves are
not very big, so it's not a huge amount of duplicate code.
I am very aware that this is not an ideal solution, and I would
welcome anyone who wants to take on fixing the run_sim problem
properly, and then cleanup the fprintf_styled duplication.
Max Filippov [Tue, 23 Aug 2022 12:25:55 +0000 (05:25 -0700)]
xtensa: bfd: fix TLS relocations generated for PIE
When generating TLS dynamic relocations the existing xtensa BFD code
treats linking to a PIE exactly as linking to a shared object, resulting
in generation of wrong relocations for TLS entries. Fix that and add
tests.
bfd/
* elf32-xtensa.c (elf_xtensa_check_relocs): Use bfd_link_dll
instead of bfd_link_pic. Add elf_xtensa_dynamic_symbol_p test
when generating GOT entries.
(elf_xtensa_relocate_section): Use bfd_link_dll instead of
bfd_link_pic.
ld/
* testsuite/ld-xtensa/tlspie.dd: New file.
* testsuite/ld-xtensa/tlspie.rd: New file.
* testsuite/ld-xtensa/tlspie.sd: New file.
* testsuite/ld-xtensa/tlspie.td: New file.
* testsuite/ld-xtensa/xtensa-linux.exp (TLS PIE transitions):
New test.
Max Filippov [Thu, 18 Aug 2022 09:46:38 +0000 (02:46 -0700)]
xtensa: adjust expected output in ld TLS tests
objdump output for l32r opcode was changed in commit b3ea76397a07
("opcodes: xtensa: display loaded literal value"), but xtensa linker TLS
relaxation tests weren't adjusted accordingly.
readelf output was changed in commit 23356397449a ("Adjust readelf's
output so that section symbols without a name as shown with their
section name."), but xtensa linker TLS relaxation tests weren't adjusted
accordingly.
Fix expected output changes in xtensa ld TLS relaxation tests.
ld/
* testsuite/ld-xtensa/tlsbin.dd: Adjust expected output for l32r
opcodes.
* testsuite/ld-xtensa/tlsbin.rd: Adjust expected output to allow
for named section symbols.
* testsuite/ld-xtensa/tlspic.dd: Adjust expected output for l32r
opcodes.
* testsuite/ld-xtensa/tlspic.rd: Adjust expected output to allow
for named section symbols.
Tsukasa OI [Sat, 27 Aug 2022 00:11:01 +0000 (00:11 +0000)]
RISC-V: Print highest address (-1) on the disassembler
This patch makes possible to print the highest address (-1) and the addresses
related to gp which value is -1. This is particularly useful if the highest
address space is used for I/O registers and corresponding symbols are defined.
Besides, despite that it is very rare to have GP the highest address, it would
be nice because we enabled highest address printing on regular cases.
gas/ChangeLog:
* testsuite/gas/riscv/dis-addr-topaddr.s: New test for the top
address (-1) printing.
* testsuite/gas/riscv/dis-addr-topaddr-32.d: Likewise.
* testsuite/gas/riscv/dis-addr-topaddr-64.d: Likewise.
* testsuite/gas/riscv/dis-addr-topaddr-gp.s: New test for
GP-relative addressing when GP is the highest address (-1).
* testsuite/gas/riscv/dis-addr-topaddr-gp-32.d: Likewise.
* testsuite/gas/riscv/dis-addr-topaddr-gp-64.d: Likewise.
opcodes/ChangeLog:
* riscv-dis.c (struct riscv_private_data): Add `to_print_addr' to
enable printing the highest address.
(maybe_print_address): Utilize `to_print_addr'.
(riscv_disassemble_insn): Likewise.
If either the base register is `zero', `tp' or `gp' and XLEN is 32, an
incorrectly sign-extended address is produced when printing. This commit
fixes this by fitting an address into a 32-bit value on RV32.
Besides, H. Peter Anvin discovered that we have wrong address computation
for JALR instruction (the initial bug is back in 2018). This commit also
fixes that based on the idea of Palmer Dabbelt.
gas/
pr29342
* testsuite/gas/riscv/lla32.d: Reflect RV32 address computation fix.
* testsuite/gas/riscv/dis-addr-overflow.s: New testcase.
* testsuite/gas/riscv/dis-addr-overflow-32.d: Likewise.
* testsuite/gas/riscv/dis-addr-overflow-64.d: Likewise.
opcodes/
pr29342
* riscv-dis.c (maybe_print_address): Fit address into 32-bit on RV32.
(print_insn_args): Fix JALR address by adding EXTRACT_ITYPE_IMM.
Tsukasa OI [Sat, 27 Aug 2022 00:10:58 +0000 (00:10 +0000)]
RISC-V: Add address printer tests with ADDIW
Address sequences involving ADDIW/C.ADDIW instructions require special
handling to sign-extend lower 32-bits of the original result.
This commit tests whether this sign-extension works.
gas/ChangeLog:
* testsuite/gas/riscv/dis-addr-addiw.s: New to test the address
computation with sign extension as used in ADDIW/C.ADDIW.
* testsuite/gas/riscv/dis-addr-addiw-a.d: Test PC sign bit 0.
* testsuite/gas/riscv/dis-addr-addiw-b.d: Test PC sign bit 1.
gas/ChangeLog:
* testsuite/gas/riscv/dis-addr-addiw-a.d: New test.
* testsuite/gas/riscv/dis-addr-addiw-b.d: New test.
* testsuite/gas/riscv/dis-addr-addiw.s: New test.
Tsukasa OI [Sat, 27 Aug 2022 02:12:09 +0000 (02:12 +0000)]
sim: Update mailing list address
The commit bf1102165389 "* MAINTAINERS: Perform some obvious fixups."
back in 2009 changed the mailing list address gdb-patches@sources.redhat.com
to gdb-patches@sourceware.org.
When running the gdb/configure script on ubuntu 22.04 with
python-3.10.4, I see:
checking for python... no
checking for python3... /usr/bin/python3
[...]/gdb/python/python-config.py:7: DeprecationWarning: The distutils package is deprecated and slated for removal in Python 3.12. Use setuptools or check PEP 632 for potential alternatives
from distutils import sysconfig
[...]/gdb/python/python-config.py:7: DeprecationWarning: The distutils.sysconfig module is deprecated, use sysconfig instead
from distutils import sysconfig
[...]/gdb/python/python-config.py:7: DeprecationWarning: The distutils package is deprecated and slated for removal in Python 3.12. Use setuptools or check PEP 632 for potential alternatives
from distutils import sysconfig
[...]/gdb/python/python-config.py:7: DeprecationWarning: The distutils.sysconfig module is deprecated, use sysconfig instead
from distutils import sysconfig
[...]/gdb/python/python-config.py:7: DeprecationWarning: The distutils package is deprecated and slated for removal in Python 3.12. Use setuptools or check PEP 632 for potential alternatives
from distutils import sysconfig
[...]/gdb/python/python-config.py:7: DeprecationWarning: The distutils.sysconfig module is deprecated, use sysconfig instead
from distutils import sysconfig
checking for python... yes
The distutils module is deprecated as per the PEP 632[1] and will be
removed in python-3.12.
This patch migrates gdb/python/python-config.py from distutils.sysconfig
to the sysconfig module[2].
The sysconfig module has has been introduced in the standard library in
python 3.2. Given that support for python < 3.2 has been removed by edae3fd6600f: "gdb/python: remove Python 2 support", this patch does not
need to support both implementations for backward compatibility.
Tom Tromey [Fri, 12 Aug 2022 19:15:01 +0000 (13:15 -0600)]
Fix interpreter-exec crash
PR mi/10347 points out that using interpreter-exec inside of a
"define" command will crash gdb. The bug here is that
gdb_setup_readline doesn't check for the case where instream==nullptr.
Tom Tromey [Fri, 12 Aug 2022 19:50:35 +0000 (13:50 -0600)]
Fix "source" with interpreter-exec
PR mi/15811 points out that "source"ing a file that uses
interpreter-exec will put gdb in a weird state, where the CLI stops
working. The bug is that tui_interp::suspend does not unregister the
event file descriptor.
Tom Tromey [Thu, 11 Aug 2022 19:16:46 +0000 (13:16 -0600)]
Remove a ui-related memory leak
gdb_setup_readline makes new streams and assigns to the various stream
members of struct ui. However, these assignments cause the previous
values to leak. As far as I can, this code is simply unnecessary and
can be removed -- with the exception of the assignment to gdb_stdtarg,
which is not initialized anywhere else.
projects / thirdparty / binutils-gdb.git / log
