Pedro Alves [Wed, 12 Nov 2014 10:10:49 +0000 (10:10 +0000)]
make "permanent breakpoints" per location and disableable
"permanent"-ness is currently a property of the breakpoint. But, it
should actually be an implementation detail of a _location_. Consider
this bit in infrun.c:
/* Normally, by the time we reach `resume', the breakpoints are either
removed or inserted, as appropriate. The exception is if we're sitting
at a permanent breakpoint; we need to step over it, but permanent
breakpoints can't be removed. So we have to test for it here. */
if (breakpoint_here_p (aspace, pc) == permanent_breakpoint_here)
{
if (gdbarch_skip_permanent_breakpoint_p (gdbarch))
gdbarch_skip_permanent_breakpoint (gdbarch, regcache);
else
error (_("\
The program is stopped at a permanent breakpoint, but GDB does not know\n\
how to step past a permanent breakpoint on this architecture. Try using\n\
a command like `return' or `jump' to continue execution."));
}
This will wrongly skip a non-breakpoint instruction if we have a
multiple location breakpoint where the whole breakpoint was set to
"permanent" because one of the locations happened to be permanent,
even if the one GDB is resuming from is not.
Related, because the permanent breakpoints are only marked as such in
init_breakpoint_sal, we currently miss marking momentary breakpoints
as permanent. A test added by a following patch trips on that.
Making permanent-ness be per-location, and marking locations as such
in add_location_to_breakpoint, the natural place to do this, fixes
this issue...
... and then exposes a latent issue with mark_breakpoints_out. It's
clearing the inserted flag of permanent breakpoints. This results in
assertions failing like this:
Breakpoint 1, main () at testsuite/gdb.base/callexit.c:32
32 return 0;
(gdb) call callexit()
[Inferior 1 (process 15849) exited normally]
gdb/breakpoint.c:12854: internal-error: allegedly permanent breakpoint is not actually inserted
A problem internal to GDB has been detected,
further debugging may prove unreliable.
The call dummy breakpoint, which is a momentary breakpoint, is set on
top of a manually inserted breakpoint instruction, and so is now
rightfully marked as a permanent breakpoint. See "Write a legitimate
instruction at the point where the infcall breakpoint is going to be
inserted." comment in infcall.c.
Re. make_breakpoint_permanent. That's only called by solib-pa64.c.
Permanent breakpoints were actually originally invented for HP-UX [1].
I believe that that call (the only one in the tree) is unnecessary
nowadays, given that nowadays the core breakpoints code analyzes the
instruction under the breakpoint to automatically detect whether it's
setting a breakpoint on top of a breakpoint instruction in the
program. I know close to nothing about HP-PA/HP-UX, though.
[1] https://sourceware.org/ml/gdb-patches/1999-q3/msg00245.html, and
https://sourceware.org/ml/gdb-patches/1999-q3/msg00242.html
In addition to the per-location issue, "permanent breakpoints" are
currently always displayed as enabled=='n':
(gdb) b main
Breakpoint 3 at 0x40053c: file ../../../src/gdb/testsuite/gdb.arch/i386-permbkpt.S, line 29.
(gdb) info breakpoints
Num Type Disp Enb Address What
3 breakpoint keep n 0x000000000040053c ../../../src/gdb/testsuite/gdb.arch/i386-permbkpt.S:29
But OTOH they're always enabled; there's no way to disable them...
In turn, this means that if one adds commands to such a breakpoint,
they're _always_ run:
(gdb) start
Starting program: /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.arch/i386-permbkpt
...
Temporary breakpoint 1, main () at ../../../src/gdb/testsuite/gdb.arch/i386-permbkpt.S:29
29 int3
(gdb) b main
Breakpoint 2 at 0x40053c: file ../../../src/gdb/testsuite/gdb.arch/i386-permbkpt.S, line 29.
(gdb) info breakpoints
Num Type Disp Enb Address What
2 breakpoint keep n 0x000000000040053c ../../../src/gdb/testsuite/gdb.arch/i386-permbkpt.S:29
(gdb) commands
Type commands for breakpoint(s) 2, one per line.
End with a line saying just "end".
>echo "hello!"
>end
(gdb) disable 2
(gdb) start
The program being debugged has been started already.
Start it from the beginning? (y or n) y
Temporary breakpoint 3 at 0x40053c: file ../../../src/gdb/testsuite/gdb.arch/i386-permbkpt.S, line 29.
Starting program: /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.arch/i386-permbkpt
Breakpoint 2, main () at ../../../src/gdb/testsuite/gdb.arch/i386-permbkpt.S:29
29 int3
"hello!"(gdb)
IMO, one should be able to disable such a breakpoint, and GDB should
then behave just like if the user hadn't created the breakpoint in the
first place (that is, report a SIGTRAP).
By making permanent-ness a property of the location, and eliminating
the bp_permanent enum enable_state state ends up fixing that as well.
No tests are added for these changes yet; they'll be added in a follow
up patch, as skipping permanent breakpoints is currently broken and
trips on an assertion in infrun.
Mark locations as permanent, not the whole breakpoint.
* breakpoint.c (remove_breakpoint_1, remove_breakpoint): Adjust.
(mark_breakpoints_out): Don't mark permanent breakpoints as
uninserted.
(breakpoint_init_inferior): Use mark_breakpoints_out.
(breakpoint_here_p): Adjust.
(bpstat_stop_status, describe_other_breakpoints): Remove handling
of permanent breakpoints.
(make_breakpoint_permanent): Mark each location as permanent,
instead of marking the breakpoint.
(add_location_to_breakpoint): If the location is permanent, mark
it as such, and as inserted.
(init_breakpoint_sal): Don't make the breakpoint permanent here.
(bp_location_compare, update_global_location_list): Adjust.
(update_breakpoint_locations): Don't make the breakpoint permanent
here.
(disable_breakpoint, enable_breakpoint_disp): Don't skip permanent
breakpoints.
* breakpoint.h (enum enable_state) <bp_permanent>: Delete field.
(struct bp_location) <permanent>: New field.
* guile/scm-breakpoint.c (bpscm_enable_state_to_string): Remove
reference to bp_permanent.
So I think we should default the gdbarch_skip_permanent_breakpoint
hook to advancing the PC by the length of the breakpoint instruction,
as determined by gdbarch_breakpoint_from_pc. I believe that simple
implementation does the right thing for most architectures. If
there's an oddball architecture where that doesn't work, then it
should override the hook, just like it should be overriding the hook
if there was no default anyway.
The only two implementation of skip_permanent_breakpoint are
i386_skip_permanent_breakpoint, for x86, and
hppa_skip_permanent_breakpoint, for PA-RISC/HP-UX
The x86 implementation is trivial, and can clearly be replaced by the
new default.
I don't know about the HP-UX one though, I know almost nothing about
PA. It may well be advancing the PC ends up being equivalent.
Otherwise, it must be that "jump $pc_after_bp" doesn't work either...
Nick Clifton [Tue, 11 Nov 2014 20:50:03 +0000 (20:50 +0000)]
Fix invalid memory accesses for more corrupt binary files.
PR binutils/17531
* binutils/readelf.c (dynamic_nent): Change type to size_t.
(slurp_rela_relocs): Use size_t type for nrelas.
(slurp_rel_relocs): Likewise.
(get_program_headers): Improve out of memory error message.
(get_32bit_section_headers): Likewise.
(get_32bit_section_headers): Likewise.
(get_64bit_section_headers): Likewise.
(get_32bit_elf_symbols): Likewise.
(get_64bit_elf_symbols): Likewise.
(process_section_groups): Likewise.
(get_32bit_dynamic_section): Likewise.
(get_64bit_dynamic_section): Likewise.
(process_dynamic_section): Likewise.
(process_version_sections): Likewise.
(get_symbol_index_type): Likewise.
(process_mips_specific): Likewise.
(process_corefile_note_segment): Likewise.
(process_version_sections): Use size_t type for total.
(get_dynamic_data): Change type of number parameter to size_t.
Improve out of memory error messages.
(process_symbol_table): Change type of nbuckets and nchains to
size_t. Skip processing of sections headers if there are none.
Improve out of memory error messages.
Nick Clifton [Tue, 11 Nov 2014 15:34:27 +0000 (15:34 +0000)]
More fixes for invalid memory accesses, uncovered by valgrind and binary fuzzers.
PR binutils/17512
* coffcode.h (coff_slurp_line_table): Initialise the parts of the
line number cache that would not be initialised by the copy from
the new line number table.
(coff_classify_symbol): Allow for _bfd_coff_internal_syment_name
returning NULL.
* coffgen.c (coff_get_normalized_symbols): Get the external
symbols before allocating space for the internal symbols, in case
the get fails.
* elf.c (_bfd_elf_slurp_version_tables): Only allocate a verref
array if one is needed. Likewise with the verdef array.
* peXXigen.c (_bfd_XXi_swap_sym_in): Replace abort()'s with error
messages.
(_bfd_XXi_swap_aux_in): Make sure that all fields of the aux
structure are initialised.
(pe_print_edata): Avoid reading off the end of the data buffer.
Linux supports multiple "PID namespaces". Processes in different PID
namespaces have different views of the system process list. Sometimes,
a single process can appear in more than one PID namespace, but with a
different PID in each. When GDB and its target are in different PID
namespaces, various features can break due to the mismatch between
what the target believes its PID to be and what GDB believes its PID
to be. The most visible broken functionality is thread enumeration
silently failing.
This patch explicitly warns users against trying to debug across PID
namespaces.
The patch introduced no new failures in my test suite run on an x86_64
installation of Ubuntu 14.10. It doesn't include a test: writing an
automated test that exercises this code would be very involved because
CLONE_NEWNS requires CAP_SYS_ADMIN; the easier way to reproduce the
problem is to start a new lxc container.
Warn about cross-PID-namespace debugging.
* nat/linux-procfs.h (linux_proc_pid_get_ns): New prototype.
* nat/linux-procfs.c (linux_proc_pid_get_ns): New function.
* linux-thread-db.c (check_pid_namespace_match): New function.
(thread_db_inferior_created): Call it.
Alan Modra [Tue, 11 Nov 2014 11:06:37 +0000 (21:36 +1030)]
Avoid coff OOM
bfd_zalloc/bfd_zmalloc to fix uninitialized memory reads is too big a
hammer, when the size allocated depends on user input. A typical
bfd_alloc, bfd_seek, bfd_bread sequence will give an error or warning
at the point the file read fails when some enormous item as described
by headers is not actually present in the file. Nice operating system
allow memory overcommit. But not if you write to the memory. So
bfd_zalloc can cause an OOM, thrashing, or system hangs.
The patch also fixes a recently introduced endless loop on bad input.
PR binutils/17512
* coffcode.h (coff_slurp_line_table): Don't bfd_zalloc, just
memset the particular bits we need. Update src after hitting loop
"continue". Don't count lineno omitted due to invalid symbols in
nbr_func, and update lineno_count. Init entire terminating
lineno. Don't both allocating terminator in n_lineno_cache.
Redirect sym->lineno pointer to where n_lineno_cache will be
copied, and free n_lineno_cache.
* pe-mips.c (NUM_HOWTOS): Typo fix.
Alan Modra [Tue, 11 Nov 2014 09:43:03 +0000 (20:13 +1030)]
ld -r abort in _bfd_elf_write_section_eh_frame
Turning on .eh_frame processing for ld -r resulted in systemtap
tickling a ld bug. Triggered by the zero terminator not being added
to .eh_frame in a separate file as it usually is (crtend.o), but
instead being present in the last .eh_frame section along with CIEs
and FDEs. The 4-byte terminator makes the section size check fail
on 64-bit targets.
* elf-eh-frame (_bfd_elf_write_section_eh_frame): Adjust section
size check to account for possible zero terminator.
Doug Evans [Mon, 10 Nov 2014 23:48:49 +0000 (15:48 -0800)]
PR 17564: Fix objfile search order for static symbols.
When searching static symbols, gdb would search over all
expanded symtabs of all objfiles, and if that fails only then
would it search all partial/gdb_index tables of all objfiles.
This means that the user could get a random instance of the
symbol depending on what symtabs have been previously expanded.
Now the search is consistent, searching each objfile completely
before proceeding to the next one.
gdb/ChangeLog:
PR symtab/17564
* symtab.c (lookup_symbol_in_all_objfiles): Delete.
(lookup_static_symbol): Move definition to new location and rewrite.
(lookup_symbol_in_objfile): New function.
(lookup_symbol_global_iterator_cb): Call it.
gdb/testsuite/ChangeLog:
PR symtab/17564
* gdb.base/symtab-search-order.exp: New file.
* gdb.base/symtab-search-order.c: New file.
* gdb.base/symtab-search-order-1.c: New file.
* gdb.base/symtab-search-order-shlib-1.c: New file.
Nick Clifton [Mon, 10 Nov 2014 16:32:32 +0000 (16:32 +0000)]
More fixes for assertion failures and out-of-bounds reads by readelf.
PR binutils/17531
* (ia64_process_unwind): Replace assertion with an error message.
Add range checking for group section indicies.
(hppa_process_unwind): Replace assertion with an error message.
(process_syminfo): Likewise.
(decode_arm_unwind_bytecode): Add range checking.
(dump_section_as_strings): Add more string range checking.
(display_tag_value): Likewise.
(display_arm_attribute): Likewise.
(display_gnu_attribute): Likewise.
(display_tic6x_attribute): Likewise.
(display_msp430x_attribute): Likewise.
Nick Clifton [Mon, 10 Nov 2014 14:18:45 +0000 (14:18 +0000)]
More fixes for problems exposed by valgrind and the address sanitizer
when displaying the contents of corrupt files.
PR binutils/17521
* coff-i386.c (NUM_HOWTOS): New define.
(RTYPE2HOWTO): Use it.
(coff_i386_rtype_to_howto): Likewise.
(coff_i386_reloc_name_lookup): Likewise.
(CALC_ADDEND): Check that reloc r_type field is valid.
* coff-x86_64.c (NUM_HOWTOS): New define.
(RTYPE2HOWTO): Use it.
(coff_amd64_rtype_to_howto): Likewise.
(coff_amd64_reloc_name_lookup): Likewise.
(CALC_ADDEND): Check that reloc r_type field is valid.
* coffcode.h (coff_slurp_line_table): Check for symbol table
indexing underflow.
(coff_slurp_symbol_table): Use zalloc to ensure that all table
entries are initialised.
* coffgen.c (_bfd_coff_read_string_table): Initialise unused bits
in the string table. Also ensure that the table is 0 terminated.
(coff_get_normalized_symtab): Check for symbol table indexing
underflow.
* opncls.c (bfd_alloc): Catch the case where a small negative size
can result in only 1 byte being allocated.
(bfd_alloc2): Use bfd_alloc.
* pe-mips.c (NUM_HOWTOS): New define.
(coff_mips_reloc_name_lookup): Use it.
(CALC_ADDEND): Check that reloc r_type field is valid.
* peXXigen.c (_bfd_XXi_swap_aouthdr_in): Initialise unused entries
in the DataDirectory.
(pe_print_idata): Avoid reading beyond the end of the data block
wen printing strings.
(pe_print_edata): Likewise.
Check for table indexing underflow.
* peicode.h (pe_mkobject): Initialise the pe_opthdr field.
(pe_bfd_object_p): Allocate and initialize enough space to hold a
PEAOUTHDR, even if the opt_hdr field specified less.
Ulrich Weigand [Mon, 10 Nov 2014 14:11:44 +0000 (15:11 +0100)]
Work around GCC bug 63748
A recent change to eval.c triggered a GCC bug that causes a false positive
"may be used uninitialized" warning in evaluate_subexp_standard. This seems
to be triggered by a specific CFG constructed via setjmp and gotos.
While the GCC bug is in the process of being fixed, there are released
compiler versions (in particular GCC 4.9) in the field that show this
problem. In order to allow compiling GDB with one of those compilers,
this commit slightly reworks the CFG (in an equivalent way) of the
affected function, so that the GCC bug is no longer triggered.
gdb/ChangeLog:
* eval.c (evaluate_subexp_standard): Work around GCC bug 63748.
H.J. Lu [Fri, 7 Nov 2014 20:22:53 +0000 (12:22 -0800)]
X32: Add REX prefix to encode R_X86_64_GOTTPOFF
Structions with R_X86_64_GOTTPOFF relocation must be encoded with REX
prefix even if it isn't required by destination register. Otherwise
linker can't safely perform IE -> LE optimization.
bfd/
PR ld/17482
* elf64-x86-64.c (elf_x86_64_relocate_section): Update comments
for IE->LE transition.
gas/
PR ld/17482
* config/tc-i386.c (output_insn): Add a dummy REX_OPCODE prefix
for structions with R_X86_64_GOTTPOFF relocation for x32 if needed.
gas/testsuite/
PR ld/17482
* gas/i386/ilp32/x32-tls.d: New file.
* gas/i386/ilp32/x32-tls.s: Likewise.
Pedro Alves [Fri, 7 Nov 2014 15:20:47 +0000 (15:20 +0000)]
gdb.base/sigstep.exp: xfail gdb/17511 on i?86 Linux
Running gdb.base/sigstep.exp with --target=i686-pc-linux-gnu on a
64-bit kernel naturally trips on PR gdb/17511 as well, given this is a
kernel bug.
I haven't really tested a real 32-bit kernel/machine, but given the
code in question in the kernel is shared between 32-bit and 64-bit,
I'm quite sure the bug triggers in those cases as well.
Pedro Alves [Fri, 7 Nov 2014 13:53:01 +0000 (13:53 +0000)]
Revert old nexti prologue check and eliminate in_prologue
The in_prologue check in the nexti code is obsolete; this commit
removes that, and then removes the in_prologue function as nothing
else uses it.
Looking at the code in GDB that makes use in_prologue, all we find is
this one caller:
if ((ecs->event_thread->control.step_over_calls == STEP_OVER_NONE)
|| ((ecs->event_thread->control.step_range_end == 1)
&& in_prologue (gdbarch, ecs->event_thread->prev_pc,
ecs->stop_func_start)))
{
/* I presume that step_over_calls is only 0 when we're
supposed to be stepping at the assembly language level
("stepi"). Just stop. */
/* Also, maybe we just did a "nexti" inside a prolog, so we
thought it was a subroutine call but it was not. Stop as
well. FENN */
/* And this works the same backward as frontward. MVS */
end_stepping_range (ecs);
return;
}
Not much discussion there, and no test, but looking at the code around
what was patched in that revision, we see that the checks that detect
whether the program has just stepped into a subroutine didn't rely on
the unwinders at all back then.
if ((step_over_calls == STEP_OVER_NONE)
|| ((step_range_end == 1)
&& in_prologue (prev_pc, ecs->stop_func_start)))
{
/* I presume that step_over_calls is only 0 when we're
supposed to be stepping at the assembly language level
("stepi"). Just stop. */
/* Also, maybe we just did a "nexti" inside a prolog,
so we thought it was a subroutine call but it was not.
Stop as well. FENN */
stop_step = 1;
print_stop_reason (END_STEPPING_RANGE, 0);
stop_stepping (ecs);
return;
}
Stripping the IN_SOLIB_RETURN_TRAMPOLINE checks for simplicity, we had:
if (stop_pc == ecs->stop_func_start /* Quick test */
|| in_prologue (stop_pc, ecs->stop_func_start)
|| ecs->stop_func_name == 0)
{
/* It's a subroutine call. */
That is, detecting a subroutine call was based on prologue detection
back then. So the in_prologue check in the current tree only made
sense back then as it was undoing a bad decision the in_prologue check
that used to exist above did.
Today, the check for a subroutine call relies on frame ids instead,
which are stable throughout the function. So we can just remove the
in_prologue check for nexti, and the whole in_prologue function along
with it.
Tested on x86_64 Fedora 20, and also by nexti-ing manually a prologue.
* infrun.c (process_event_stop_test) <subroutine check>: Don't
check if we did a "nexti" inside a prologue.
* symtab.c (in_prologue): Delete function.
* symtab.h (in_prologue): Delete declaration.
Nick Clifton [Fri, 7 Nov 2014 13:39:45 +0000 (13:39 +0000)]
Add more fixes for inavlid memory accesses triggered by corrupt files.
PR binutils/17531
* readelf.c (get_data): Avoid allocating memory when we know that
the read will fail.
(find_section_by_type): New function.
(get_unwind_section_word): Check for invalid symbol indicies.
Check for invalid reloc types.
(get_32bit_dynamic_section): Add range checks.
(get_64bit_dynamic_section): Add range checks.
(process_dynamic_section): Check for a corrupt time value.
(process_symbol_table): Add range checks.
(dump_section_as_strings): Add string length range checks.
(display_tag_value): Likewise.
(display_arm_attribute): Likewise.
(display_gnu_attribute): Likewise.
(display_tic6x_attribute): Likewise.
(display_msp430x_attribute): Likewise.
(process_mips_specific): Add range check.
Alan Modra [Fri, 7 Nov 2014 09:59:43 +0000 (20:29 +1030)]
tekhex buffer management and symbol types
Dramatically reduces memory consumption and processing time for large
all-zero data segments. Allows multiple symbol types attached to a
given segment to survive objcopy.
* tekhex.c (CHUNK_SPAN): Define.
(struct data_struct <chunk_init>): Use one byte per span, update
all code accessing this field.
(find_chunk): Add create param, don't create new entry unless set.
(insert_byte): Don't save zeros.
(first_phase): Set section SEC_CODE or SEC_DATA flag depending
on symbol type. Create an alternate section if both types of
symbol are given. Attach type '2' and '6' symbols to absolute
section.
(move_section_contents): Fix caching of chunk. Don't create chunk
when reading, or for writing zeros.
(tekhex_set_section_contents): Don't create initial chunks.
(tekhex_write_object_contents): Use CHUNK_SPAN.
Alan Modra [Fri, 7 Nov 2014 09:48:25 +0000 (20:18 +1030)]
aoutx.h tidy
Save a multiplication, and any concern that the buffer allocation
might be smaller than the amount read (as it could be if the header
size isn't a multiple of EXTERNAL_NLIST_SIZE).
* aoutx.h (aout_get_external_symbols): Tidy allocation of symbol buffer.
Doug Evans [Fri, 7 Nov 2014 07:29:49 +0000 (23:29 -0800)]
Rename some "aux" functions.
"aux" doesn't contribute anything to the name, and it makes the
reader wonder what it's supposed to mean.
gdb/ChangeLog:
* symtab.c (lookup_local_symbol): Renamed from lookup_symbol_aux_local.
All callers updated.
(lookup_symbol_in_all_objfiles): Renamed from
lookup_symbol_aux_symtabs. All callers updated.
(lookup_symbol_via_quick_fns): Renamed from lookup_symbol_aux_quick.
All callers updated.
(lookup_symbol_in_objfile_symtabs): Renamed from
lookup_symbol_aux_objfile. All callers updated.
Doug Evans [Fri, 7 Nov 2014 06:56:46 +0000 (22:56 -0800)]
Rename lookup_symbol_static to lookup_symbol_in_static_block,
and lookup_static_symbol_aux to lookup_static_symbol.
gdb/ChangeLog:
* symtab.c (lookup_static_symbol): Renamed from
lookup_static_symbol_aux. All callers updated.
(lookup_symbol_in_static_block): Renamed from lookup_symbol_static.
All callers updated.
Doug Evans [Fri, 7 Nov 2014 06:50:12 +0000 (22:50 -0800)]
New macro ALL_BLOCK_SYMBOLS_WITH_NAME.
gdb/ChangeLog:
* block.h (ALL_BLOCK_SYMBOLS_WITH_NAME): New macro.
* block.c (block_lookup_symbol): Use it.
* cp-support.c (make_symbol_overload_list_block): Use it.
* symtab.c (iterate_over_symbols): Use it.
Doug Evans [Fri, 7 Nov 2014 06:32:25 +0000 (22:32 -0800)]
Move lookup_block_symbol to block.c, rename to block_lookup_symbol.
There is another function, lookup_symbol_aux_block, and
the names lookup_block_symbol and lookup_symbol_aux_block don't
convey any real difference between them.
The difference is that lookup_block_symbol lives in the lower level
block API, and lookup_symbol_aux_block lives in the higher level symtab API.
This patch makes this distinction clear.
gdb/ChangeLog:
* symtab.c (lookup_block_symbol): Moved to ...
* block.c (block_lookup_symbol): ... here and renamed.
All callers updated.
* block.h (block_lookup_symbol): Declare.
* symtab.h (lookup_block_symbol): Delete.
Nick Clifton [Thu, 6 Nov 2014 14:49:10 +0000 (14:49 +0000)]
Prevent archive memebers with illegal pathnames from being extracted from an archive.
PR binutils/17552, binutils/17533
* bucomm.c (is_valid_archive_path): New function. Returns false
for absolute pathnames and pathnames that include /../.
* bucomm.h (is_valid_archive_path): Add prototype.
* ar.c (extract_file): Use new function to check for valid
pathnames when extracting files from an archive.
* objcopy.c (copy_archive): Likewise.
* doc/binutils.texi: Update documentation to mention the
limitation on pathname of archive members.
Nick Clifton [Wed, 5 Nov 2014 17:57:54 +0000 (17:57 +0000)]
More fixes for memory problems uncovered by file fuzzers.
PR binutils/17512
* coffcode.h (handle_COMDAT): Replace abort with BFD_ASSERT.
Replace another abort with an error message.
(coff_slurp_line_table): Add more range checking.
* peXXigen.c (pe_print_debugdata): Add range checking.
Nick Clifton [Wed, 5 Nov 2014 16:19:03 +0000 (16:19 +0000)]
More fixes for reading corrupt ELF files.
PR binutils/15731
* readelf.c (printable_section_name): New function.
(printable_section_name_from_index): New function.
(dump_relocations): Use new function.
(process_program_headers, get_32bit_elf_symbols,
(get_64bit_elf_symbols, process_section_headers,
(process_section_groups, process_relocs, ia64_process_unwind,
(hppa_process_unwind, get_unwind_section_word, decode_arm_unwind,
(arm_process_unwind, process_version_sections,
(process_symbol_table, apply_relocations, get_section_contents,
(dump_section_as_strings, dump_section_as_bytes,
(display_debug_section, process_attributes, process_mips_specific,
(process_mips_specific process_gnu_liblist): Likewise.
(get_unwind_section_word): Check for a missing symbol table.
Replace aborts with error messages.
(arm_process_unwind): Check for a missing string table.
(process_attributes): Check for an attribute length that is too
small.
(process_mips_specific): Check for a corrupt GOT symbol offset.
Nick Clifton [Wed, 5 Nov 2014 10:13:16 +0000 (10:13 +0000)]
More fixes for processing corrupt files.
PR binutils/17512
* coffcode.h (coff_set_alignment_hook): Warn if the file lies
about the number of relocations it contains.
(coff_sort_func_alent): Return 0 if the pointers are NULL.
(coff_slurp_line_table): Add more range checks. Do not free new
tables created when sorting line numbers.
* peXXigen.c (pe_print_idata): Add range checks.
(pe_print_edata): Likewise.
(rsrc_print_resource_entries): Likewise. Avoid printing control
characters. Terminate priniting if corruption is detected.
(rsrc_print_resource_directory): Terminate printing if an unknown
directory type is encountered.
(pe_print_debugdata): Fix off-by-one error.
(rsrc_count_entries): Add range checking.
(rsrc_parse_entry): Likewise.
Alan Modra [Tue, 4 Nov 2014 22:48:27 +0000 (09:18 +1030)]
Cast result of obstack_next_free
obstack_next_free is supposed to return a void*, rather than a char*
as it does currently. Avoid warning on void* arithmetic when
obstack_next_free gets it proper return type.
* cp-valprint.c (cp_print_value_fields): Cast obstack_next_free
to char* before doing pointer arithmetic.
Simon Marchi [Tue, 4 Nov 2014 13:27:06 +0000 (08:27 -0500)]
tui: Fix newterm call for older ncurses
Older versions of ncurses' newterm can't take NULL for their ofp and ifp
parameters. Newer versions can, and they fall back on stdout/stdin if
that is the case.
This patch explicitly passes stdout/stdin to the call to newterm to
avoid segfaulting with older ncurses.
Nick Clifton [Tue, 4 Nov 2014 15:29:03 +0000 (15:29 +0000)]
More fixes for memory corruption when readelf processes corrupt files.
PR binutils/17531
(get_32bit_program_headers): Verify program header entry size
before reading in the program headers.
(get_64bit_program_headers): Likewise.
(get_unwind_section_word): Do nothing if no section was provided.
Fail if the offset is outside of the section.
(print_dynamic_symbol): Catch out of range symbol indicies.
(process_mips_specific): Likewise.
(process_attributes): Make sure that there is enough space left in
the section before attempting to read the length of the next
attribute.
Alan Modra [Tue, 4 Nov 2014 04:11:00 +0000 (14:41 +1030)]
Use frag_now_fix_octets in gas d10v, d30v
obstack_next_free is supposed to return a void* rather than the char*
it does currently, so expressions involving pointer arithmetic need
a cast. Avoid the issue.
* config/tc-d10v.c (find_opcode): Call frag_now_fix_octets rather
than equivalent obstack_next_free expression.
* config/tc-d30v.c (find_format): Likewise.
Alan Modra [Tue, 4 Nov 2014 03:09:38 +0000 (13:39 +1030)]
Provide stat function for spu overlay manager iovec
Commit f54498b4 broke spu-elf, specifically the change "Do not try to
load a string table bigger than the file", because bfd_get_size
returns zero for the spu built-in overlay manager bfd.
* elf32-spu.c (ovl_mgr_stat): New function.
(spu_elf_open_builtin_lib): Pass to bfd_openr_iovec.
Nick Clifton [Tue, 4 Nov 2014 11:58:16 +0000 (11:58 +0000)]
Fixes for crashes running readelf.
PR binutils/17531
* readelf.c (get_data): If the reason parameter is null, do not
print any error messages.
(get_32bit_section_headers): Verify section header entry size
before reading in the section headers.
(get_64bit_section_headers): Likewise.
(process_section_headers): Pass FALSE to get_section_headers.
(get_file_header): Pass TRUE to get_section_headers.
(process_dynamic_section): Change an assert to an error message.
(process_symbol_table): Handle corrupt histograms.
Siva Chandra [Sat, 18 Oct 2014 13:14:00 +0000 (06:14 -0700)]
Fix evaluation of method calls under EVAL_SKIP.
When evaluating method calls under EVAL_SKIP, the "object" and the
arguments to the method should also be evaluated under EVAL_SKIP,
instead of skipping to evaluate them as was being done previously.
gdb/ChangeLog:
PR c++/17494
* eval.c (evaluate_subexp_standard): Evaluate the "object" and
the method args also under EVAL_SKIP when evaluating method
calls under EVAL_SKIP.
gdb/testsuite/ChangeLog:
PR c++/17494
* gdb.cp/pr17494.cc: New file.
* gdb.cp/pr17494.exp: New file.
Andrew Burgess [Mon, 27 Oct 2014 10:45:18 +0000 (10:45 +0000)]
When relaxing, update size of symbols.
When performing linker relaxation, reduce the size of symbols that span
the deleted bytes. This ensures that, for example, function symbols
will have the correct size.
bfd/ChangeLog:
* elf32-avr.c (elf32_avr_relax_delete_bytes): During linker
relaxation, reduce the size of symbols that span the deleted
bytes.
ld/ChangeLog:
* testsuite/ld-avr/relax-02.d: Update to check size of symbols has
changed.
* testsuite/ld-avr/relax-03.d: Likewise.
Andrew Burgess [Sat, 25 Oct 2014 14:08:14 +0000 (15:08 +0100)]
When relaxing, update symbols at the very end of the section.
Symbols at the very end of a section were not being updated correctly
when linker relaxation takes place due to the use of '<' instead of
'<='. Added a couple of tests to cover this behaviour.
bfd/ChangeLog:
* elf32-avr.c (elf32_avr_relax_delete_bytes): Modify symbols
located at the very end of the section.
ld/ChangeLog:
* ld/testsuite/ld-avr/relax-02.d: New file.
* ld/testsuite/ld-avr/relax-02.s: New file.
* ld/testsuite/ld-avr/relax-03.d: New file.
* ld/testsuite/ld-avr/relax-03.s: New file.
Nick Clifton [Mon, 3 Nov 2014 17:44:00 +0000 (17:44 +0000)]
More fixes for buffer overruns instigated by corrupt binaries.
PR binutils/17512
* objdump.c (slurp_symtab): Fail gracefully if the table could not
be read.
(dump_relocs_in_section): Likewise.
* aoutx.h (slurp_symbol_table): Check that computed table size is
not bigger than the file from which is it being read.
(slurp_reloc_table): Likewise.
* coffcode.h (coff_slurp_line_table): Remove unneeded local
'warned'. Do not try to print the details of a symbol with an
invalid index.
* coffgen.c (make_a_sectiobn_from_file): Check computed string
index against length of string table.
(bfd_coff_internal_syment_name): Check read in string offset
against length of string table.
(build_debug_section): Return a pointer to the section used.
(_bfd_coff_read_string_table): Store the length of the string
table in the coff_tdata structure.
(bfd_coff_free_symbols): Set the length of the string table to
zero when it is freed.
(coff_get_normalized_symtab): Check offsets against string table
or data table lengths as appropriate.
* cofflink.c (_bfd_coff_link_input_bfd): Check offset against
length of string table.
* compress.c (bfd_get_full_section_contents): Check computed size
against the size of the file.
* libcoff-in.h (obj_coff_strings_len): Define.
(struct coff_tdata): Add strings_len field.
* libcoff.h: Regenerate.
* peXXigen.c (pe_print_debugdata): Do not attempt to print the
data if the debug section is too small.
* xcofflink.c (xcoff_link_input_bfd): Check offset against
length of string table.
Victor Kamensky [Sun, 2 Nov 2014 21:28:35 +0000 (13:28 -0800)]
read_pieced_value do big endian processing only in case of valid gdb_regnum
During armv7b testing gdb.base/store.exp test was failling with
'GDB internal error' with the following message:
Temporary breakpoint 1, wack_double (u=
../../binutils-gdb/gdb/regcache.c:177: internal-error: register_size: Assertion `regnum >= 0 && regnum < (gdbarch_num_regs (gdbarch) + gdbarch_num_pseudo_regs (gdbarch))' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
It turns out that compiler generated DWARF with non-existent
register numbers. The compiler issue is present in both little endian
(armv7) and big endian (armv7b) (it is separate issue). Here is
example for one of formal parameters of wack_double function:
In both big and little endian cases gdbarch_dwarf2_reg_to_regnum
returns -1 which is stored into gdb_regnum. But it causes severe
problem only in big endian case because in read_pieced_value and
write_pieced_value functions BFD_ENDIAN_BIG related processing
happen regardless of gdb_regnum value, for example register_size
function is called and in case of gdb_regnum=-1, it cause
'GDB internal error' and crash.
Solution is to move BFD_ENDIAN_BIG related processing under
(gdb_regnum != -1) branch of processing.
Victor Kamensky [Sun, 2 Nov 2014 21:28:35 +0000 (13:28 -0800)]
ARM: arm_breakpoint should be little endian form in case for arm BE8
tdep->arm_breakpoint, tdep->thumb_breakpoint, tdep->thumb2_breakpoint
should be set le_ variants in case of arm BE8 code. Those instruciton
sequences are writen to target with simple write_memory, without
regarding gdbarch_byte_order_for_code. But in BE8 case even data
memory is in big endian form, instructions are still in little endian
form.
Because of this issue there are many issues while running gdb test
case in armv7b mode. For example gdb.arch/arm-disp-step.exp test fails
because it gets SIGILL when displaced instrucion sequence reaches
break instruction, which is in wrong byte order.
Solution is to set tdep->xxx_breakpoint sequences in BE8 case (i.e
when gdbarch_byte_order_for_code is BFD_ENDIAN_BIG.
Victor Kamensky [Sun, 2 Nov 2014 21:28:35 +0000 (13:28 -0800)]
ARM: extract_arm_insn function need to read instrs correctly in be8 case
extract_arm_insn function needs to read instructions in
gdbarch_byte_order_for_code byte order, because in case armv7b,
even data is big endian, instructions are still little endian.
Currently function uses gdbarch_byte_order which would be
big endian in armv7b case.
Because of this issue pretty much all gdb.reverse/ tests are
failing with 'Process record does not support instruction' message.
Fix is to change gdbarch_byte_order to gdbarch_byte_order_for_code,
when passed to extract_unsigned_integer that reads instruction.
Yao Qi [Sun, 2 Nov 2014 13:08:06 +0000 (21:08 +0800)]
Match the working directory on remote host
The test in gdb.python/python.exp tests "extended-prompt" and expects
working directory is printed. However, working directory on remote
host doesn't have "gdb/testsuite", so the test fails on remote host
like:
set extended-prompt \w ^M
^M
/home/yao FAIL: gdb.python/python.exp: set extended prompt working directory (timeout)
This patch is to get the working directory first, and use it to match
the output of "set extended-prompt \\w ". It works for remote host
and non remote host.
projects / binutils.git / log