A following patch will want to factor out a bit of
evaluate_subexp_standard, and it'd be handy to reuse the code under the
"nosideret:" label there too. This commits moves it to a separate
function as preparation for that.
gdb/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* eval.c (eval_skip_value): New function.
(evaluate_subexp_standard): Use it.
The patch to make GDB stop assuming functions return int left GDB with
an inconsistency. While with normal expression evaluation the
"unknown return type" error shows the name of the function that misses
debug info:
(gdb) p getenv ("PATH")
'getenv' has unknown return type; cast the call to its declared return type
^^^^^^
which is handy in more complicated expressions, "ptype" does not:
(gdb) ptype getenv ("PATH")
function has unknown return type; cast the call to its declared return type
^^^^^^^^
This commit builds on the new OP_VAR_MSYM_VALUE to fix it, by making
OP_FUNCALL extract the function name from the symbol stored in
OP_VAR_VALUE/OP_VAR_MSYM_VALUE. We now get the same error in "print"
vs "ptype":
(gdb) ptype getenv()
'getenv' has unknown return type; cast the call to its declared return type
(gdb) p getenv()
'getenv' has unknown return type; cast the call to its declared return type
gdb/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* eval.c (evaluate_subexp_standard): <OP_FUNCALL>: Extract
function name from symbol/minsym and pass it to
error_call_unknown_return_type.
gdb/testsuite/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* gdb.base/nodebug.exp: Test that ptype's error about functions
with unknown return type includes the function name too.
The previous patch left GDB with an inconsistency. While with normal
expression evaluation the "unknown return type" error shows the name
of the function that misses debug info:
(gdb) p getenv ("PATH")
'getenv' has unknown return type; cast the call to its declared return type
^^^^^^
which can by handy in more complicated expressions, "ptype" does not:
(gdb) ptype getenv ("PATH")
function has unknown return type; cast the call to its declared return type
^^^^^^^^
This commit is a step toward fixing it.
The problem is that while evaluating the expression above, we have no
reference to the minimal symbol where we could extract the name from.
This is because the resulting expression tree has no reference to the
minsym at all. During parsing, the type and address of the minsym are
extracted and an UNOP_MEMVAL / UNOP_MEMVAL_TLS operator is generated
(see write_exp_elt_msym). With "set debug expression", here's what
you see:
0 OP_FUNCALL Number of args: 0
3 UNOP_MEMVAL Type @0x565334a51930 (<text variable, no debug info>)
6 OP_LONG Type @0x565334a51c60 (__CORE_ADDR), value 140737345035648 (0x7ffff7751d80)
The "print" case finds the function name, because
call_function_by_hand looks up the function by address again.
However, for "ptype", we don't reach that code, because obviously we
don't really call the function.
Unlike minsym references, references to variables with debug info have
a pointer to the variable's symbol in the expression tree, with
OP_VAR_VALUE:
(gdb) ptype main()
...
0 OP_FUNCALL Number of args: 0
3 OP_VAR_VALUE Block @0x0, symbol @0x559bbbd9b358 (main(int, char**))
...
so I don't see why do minsyms need to be different. So to prepare for
fixing the missing function name issue, this commit adds a new
OP_VAR_MSYM_VALUE operator that mimics OP_VAR_VALUE, except that it's
for minsyms instead of debug symbols. For infcalls, we now get
expressions like these:
0 OP_FUNCALL Number of args: 0
3 OP_VAR_MSYM_VALUE Objfile @0x1e41bf0, msymbol @0x7fffe599b000 (getenv)
In the following patch, we'll make OP_FUNCALL extract the function
name from the symbol stored in OP_VAR_VALUE/OP_VAR_MSYM_VALUE.
OP_VAR_MSYM_VALUE will be used more in a later patch in the series
too.
gdb/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* ada-lang.c (resolve_subexp): Handle OP_VAR_MSYM_VALUE.
* ax-gdb.c (gen_msym_var_ref): New function.
(gen_expr): Handle OP_VAR_MSYM_VALUE.
* eval.c (evaluate_var_msym_value): New function.
* eval.c (evaluate_subexp_standard): Handle OP_VAR_MSYM_VALUE.
<OP_FUNCALL>: Extract function name from symbol/minsym and pass it
to call_function_by_hand.
* expprint.c (print_subexp_standard, dump_subexp_body_standard):
Handle OP_VAR_MSYM_VALUE.
(union exp_element) <msymbol>: New field.
* minsyms.h (struct type): Forward declare.
(find_minsym_type_and_address): Declare.
* parse.c (write_exp_elt_msym): New function.
(write_exp_msymbol): Delete, refactored as ...
(find_minsym_type_and_address): ... this new function.
(write_exp_msymbol): Reimplement using OP_VAR_MSYM_VALUE.
(operator_length_standard, operator_check_standard): Handle
OP_VAR_MSYM_VALUE.
* std-operator.def (OP_VAR_MSYM_VALUE): New.
The fact that GDB defaults to assuming that functions return int, when
it has no debug info for the function has been a recurring source of
user confusion. Recently this came up on the errno pretty printer
discussions. Shortly after, it came up again on IRC, with someone
wondering why does getenv() in GDB return a negative int:
(gdb) p getenv("PATH")
$1 = -6185
This question (with s/getenv/random-other-C-runtime-function) is a FAQ
on IRC.
The reason for the above is:
(gdb) p getenv
$2 = {<text variable, no debug info>} 0x7ffff7751d80 <getenv>
(gdb) ptype getenv
type = int ()
... which means that GDB truncated the 64-bit pointer that is actually
returned from getent to 32-bit, and then sign-extended it:
(gdb) p /x -6185
$6 = 0xffffe7d7
The workaround is to cast the function to the right type, like:
(gdb) p ((char *(*) (const char *)) getenv) ("PATH")
$3 = 0x7fffffffe7d7 "/usr/local/bin:/"...
IMO, we should do better than this.
I see the "assume-int" issue the same way I see printing bogus values
for optimized-out variables instead of "<optimized out>" -- I'd much
rather that the debugger tells me "I don't know" and tells me how to
fix it than showing me bogus misleading results, making me go around
tilting at windmills.
If GDB prints a signed integer when you're expecting a pointer or
aggregate, you at least have some sense that something is off, but
consider the case of the function actually returning a 64-bit integer.
For example, compile this without debug info:
unsigned long long
function ()
{
return 0x7fffffffffffffff;
}
Currently, with pristine GDB, you get:
(gdb) p function ()
$1 = -1 # incorrect
(gdb) p /x function ()
$2 = 0xffffffff # incorrect
maybe after spending a few hours debugging you suspect something is
wrong with that -1, and do:
(gdb) ptype function
type = int ()
and maybe, just maybe, you realize that the function actually returns
unsigned long long. And you try to fix it with:
(gdb) p /x (unsigned long long) function ()
$3 = 0xffffffffffffffff # incorrect
... which still produces the wrong result, because GDB simply applied
int to unsigned long long conversion. Meaning, it sign-extended the
integer that it extracted from the return of the function, to 64-bits.
and then maybe, after asking around on IRC, you realize you have to
cast the function to a pointer of the right type, and call that. It
won't be easy, but after a few missteps, you'll get to it:
..... (gdb) p /x ((unsigned long long(*) ()) function) ()
$666 = 0x7fffffffffffffff # finally! :-)
So to improve on the user experience, this patch does the following
(interrelated) things:
- makes no-debug-info functions no longer default to "int" as return
type. Instead, they're left with NULL/"<unknown return type>"
return type.
(gdb) ptype getenv
type = <unknown return type> ()
- makes calling a function with unknown return type an error.
(gdb) p getenv ("PATH")
'getenv' has unknown return type; cast the call to its declared return type
- and then to make it easier to call the function, makes it possible
to _only_ cast the return of the function to the right type,
instead of having to cast the function to a function pointer:
(gdb) p (char *) getenv ("PATH") # now Just Works
$3 = 0x7fffffffe7d7 "/usr/local/bin:/"...
(gdb) p ((char *(*) (const char *)) getenv) ("PATH") # continues working
$4 = 0x7fffffffe7d7 "/usr/local/bin:/"...
I.e., it makes GDB default the function's return type to the type
of the cast, and the function's parameters to the type of the
arguments passed down.
After this patch, here's what you'll get for the "unsigned long long"
example above:
(gdb) p function ()
'function' has unknown return type; cast the call to its declared return type
(gdb) p /x (unsigned long long) function ()
$4 = 0x7fffffffffffffff # correct!
Note that while with "print" GDB shows the name of the function that
has the problem:
(gdb) p getenv ("PATH")
'getenv' has unknown return type; cast the call to its declared return type
which can by handy in more complicated expressions, "ptype" does not:
(gdb) ptype getenv ("PATH")
function has unknown return type; cast the call to its declared return type
This will be fixed in the next patch.
gdb/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* ada-lang.c (ada_evaluate_subexp) <TYPE_CODE_FUNC>: Don't handle
TYPE_GNU_IFUNC specially here. Throw error if return type is
unknown.
* ada-typeprint.c (print_func_type): Handle functions with unknown
return type.
* c-typeprint.c (c_type_print_base): Handle functions and methods
with unknown return type.
* compile/compile-c-symbols.c (convert_symbol_bmsym)
<mst_text_gnu_ifunc>: Use nodebug_text_gnu_ifunc_symbol.
* compile/compile-c-types.c: Include "objfiles.h".
(convert_func): For functions with unknown return type, warn and
default to int.
* compile/compile-object-run.c (compile_object_run): Adjust call
to call_function_by_hand_dummy.
* elfread.c (elf_gnu_ifunc_resolve_addr): Adjust call to
call_function_by_hand.
* eval.c (evaluate_subexp_standard): Adjust calls to
call_function_by_hand. Handle functions and methods with unknown
return type. Pass expect_type to call_function_by_hand.
* f-typeprint.c (f_type_print_base): Handle functions with unknown
return type.
* gcore.c (call_target_sbrk): Adjust call to
call_function_by_hand.
* gdbtypes.c (objfile_type): Leave nodebug text symbol with NULL
return type instead of int. Make nodebug_text_gnu_ifunc_symbol be
an integer address type instead of nodebug.
* guile/scm-value.c (gdbscm_value_call): Adjust call to
call_function_by_hand.
* infcall.c (error_call_unknown_return_type): New function.
(call_function_by_hand): New "default_return_type" parameter.
Pass it down.
(call_function_by_hand_dummy): New "default_return_type"
parameter. Use it instead of defaulting to int. If there's no
default and the return type is unknown, throw an error. If
there's a default return type, and the called function has no
debug info, then assume the function is prototyped.
* infcall.h (call_function_by_hand, call_function_by_hand_dummy):
New "default_return_type" parameter.
(error_call_unknown_return_type): New declaration.
* linux-fork.c (call_lseek): Cast return type of lseek.
(inferior_call_waitpid, checkpoint_command): Adjust calls to
call_function_by_hand.
* linux-tdep.c (linux_infcall_mmap, linux_infcall_munmap): Adjust
calls to call_function_by_hand.
* m2-typeprint.c (m2_procedure): Handle functions with unknown
return type.
* objc-lang.c (lookup_objc_class, lookup_child_selector)
(value_nsstring, print_object_command): Adjust calls to
call_function_by_hand.
* p-typeprint.c (pascal_type_print_varspec_prefix): Handle
functions with unknown return type.
(pascal_type_print_func_varspec_suffix): New function.
(pascal_type_print_varspec_suffix) <TYPE_CODE_FUNC,
TYPE_CODE_METHOD>: Use it.
* python/py-value.c (valpy_call): Adjust call to
call_function_by_hand.
* rust-lang.c (rust_evaluate_funcall): Adjust call to
call_function_by_hand.
* valarith.c (value_x_binop, value_x_unop): Adjust calls to
call_function_by_hand.
* valops.c (value_allocate_space_in_inferior): Adjust call to
call_function_by_hand.
* typeprint.c (type_print_unknown_return_type): New function.
* typeprint.h (type_print_unknown_return_type): New declaration.
gdb/testsuite/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* gdb.base/break-main-file-remove-fail.exp (test_remove_bp): Cast
return type of munmap in infcall.
* gdb.base/break-probes.exp: Cast return type of foo in infcall.
* gdb.base/checkpoint.exp: Simplify using for loop. Cast return
type of ftell in infcall.
* gdb.base/dprintf-detach.exp (dprintf_detach_test): Cast return
type of getpid in infcall.
* gdb.base/infcall-exec.exp: Cast return type of execlp in
infcall.
* gdb.base/info-os.exp: Cast return type of getpid in infcall.
Bail on failure to extract the pid.
* gdb.base/nodebug.c: #include <stdint.h>.
(multf, multf_noproto, mult, mult_noproto, add8, add8_noproto):
New functions.
* gdb.base/nodebug.exp (test_call_promotion): New procedure.
Change expected output of print/whatis/ptype with functions with
no debug info. Test all supported languages. Call
test_call_promotion.
* gdb.compile/compile.exp: Adjust expected output to expect
warning.
* gdb.threads/siginfo-threads.exp: Likewise.
Calling a prototyped function via a function pointer with the right
prototype doesn't work correctly, if the called function requires
argument coercion... Like, e.g., with:
float mult (float f1, float f2) { return f1 * f2; }
(gdb) p mult (2, 3.5)
$1 = 7
(gdb) p ((float (*) (float, float)) mult) (2, 3.5)
$2 = 0
both calls should have returned the same, of course. The problem is
that GDB misses marking the type of the function pointer target as
prototyped...
Without the fix, the new test fails like this:
(gdb) p ((int (*) (float, float)) t_float_values2)(3.14159,float_val2)
$30 = 0
(gdb) FAIL: gdb.base/callfuncs.exp: p ((int (*) (float, float)) t_float_values2)(3.14159,float_val2)
gdb/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* gdbtypes.c (lookup_function_type_with_arguments): Mark function
types with more than one parameter as prototyped.
gdb/testsuite/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* gdb.base/callfuncs.exp (do_function_calls): New parameter
"prototypes". Test calling float functions via prototyped and
unprototyped function pointers.
(perform_all_tests): New parameter "prototypes". Pass it down.
(top level): Pass down "prototypes" parameter to
perform_all_tests.
This patch improves the loop_break and loop_continue tests to verify
that they work as expected when multiple loops are nested (they affect
the inner loop).
gdb/testsuite/ChangeLog:
* gdb.base/commands.exp (loop_break_test, loop_continue_test):
Test with nested loops.
For some reason I ended up staring at some of the "int flags" in
btrace-related code, and I got confused because I had no clue what the
flags where supposed to indicate.
Fix that by using enum_flags, so that:
#1 - it's clear from the type what the flags are about, and
#2 - the compiler can catch mismatching mistakes
gdb/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* cli/cli-cmds.c (print_disassembly, disassemble_current_function)
(disassemble_command): Use gdb_disassembly_flags instead of bare
int.
* disasm.c (gdb_pretty_print_disassembler::pretty_print_insn)
(dump_insns, do_mixed_source_and_assembly_deprecated)
(do_mixed_source_and_assembly, do_assembly_only, gdb_disassembly):
Use gdb_disassembly_flags instead of bare int.
* disasm.h (DISASSEMBLY_SOURCE_DEPRECATED, DISASSEMBLY_RAW_INSN)
(DISASSEMBLY_OMIT_FNAME, DISASSEMBLY_FILENAME)
(DISASSEMBLY_OMIT_PC, DISASSEMBLY_SOURCE)
(DISASSEMBLY_SPECULATIVE): No longer macros. Instead they're...
(enum gdb_disassembly_flag): ... values of this new enumeration.
(gdb_disassembly_flags): Define.
(gdb_disassembly)
(gdb_pretty_print_disassembler::pretty_print_insn): Use it.
* mi/mi-cmd-disas.c (mi_cmd_disassemble): Use
gdb_disassembly_flags instead of bare int.
* record-btrace.c (btrace_insn_history)
(record_btrace_insn_history, record_btrace_insn_history_range)
(record_btrace_insn_history_from): Use gdb_disassembly_flags
instead of bare int.
* record.c (get_insn_history_modifiers, cmd_record_insn_history):
Use gdb_disassembly_flags instead of bare int.
* target-debug.h (target_debug_print_gdb_disassembly_flags):
Define.
* target-delegates.c: Regenerate.
* target.c (target_insn_history, target_insn_history_from)
(target_insn_history_range): Use gdb_disassembly_flags instead of
bare int.
* target.h: Include "disasm.h".
(struct target_ops) <to_insn_history, to_insn_history_from,
to_insn_history_range>: Use gdb_disassembly_flags instead of bare
int.
(target_insn_history, target_insn_history_from)
(target_insn_history_range): Use gdb_disassembly_flags instead of
bare int.
I grepped the testsuite for loop_break and loop_continue and didn't find
anything, so I wrote some simple tests for those.
gdb/testsuite/ChangeLog:
* gdb.base/commands.exp: Call the new procedures.
(loop_break_test, loop_continue_test): New procedures.
When using "if" (or while) without args directly on gdb's command line,
you get this:
(gdb) if
if/while commands require arguments
When doing the same when entering a command list, you only get an error
when the command is executed, when parse_exp_in_context_1 fails to
evaluate the expression.
(gdb) define foo
Type commands for definition of "foo".
End with a line saying just "end".
>if
>end
>end
(gdb) foo
Argument required (expression to compute).
I think it would make more sense to error out when inputting the command
list directly:
(gdb) define foo
Type commands for definition of "foo".
End with a line saying just "end".
>if
if/while commands require arguments.
The only required change is to check whether args is an empty string in
build_command_line.
gdb/ChangeLog:
* cli/cli-script.c (build_command_line): For if/while commands,
check whether args is empty.
gdb/testsuite/ChangeLog:
* gdb.base/commands.exp: Call new procedure.
(define_if_without_arg_test): New procedure.
I think it would make more sense if the types and function declarations
related to command lines were in cli-script.h rather than defs.h, since
the related function definitions are in cli-script.c.
I had to add a few includes here and there. I also had to rename the
"lines" parameter of command_lines_deleter::operator(), because ncurses
has a "#define lines ..." that was interfering when cli-script.h is
included by some TUI source files that also include ncurses header files.
gdb/ChangeLog:
* cli/cli-script.h (enum misc_command_type): Move from defs.h.
(enum command_control_type): Likewise.
(struct command_line): Likewise.
(free_command_lines): Likewise.
(struct command_lines_deleter): Likewise.
(command_line_up): Likewise.
(read_command_lines): Likewise.
(read_command_lines_1): Likewise.
* defs.h (enum misc_command_type): Move to cli/cli-script.h.
(enum command_control_type): Likewise.
(struct command_line): Likewise.
(free_command_lines): Likewise.
(struct command_lines_deleter): Likewise.
(command_line_up): Likewise.
(read_command_lines): Likewise.
(read_command_lines_1): Likewise.
* breakpoint.h: Include cli/cli-script.h.
* extension-priv.h: Likewise.
* gdbcmd.h: Likewise.
If you "make" from scratch in gdbserver/, you'll notice that make
deletes the files it considers as intermediary at the end:
$ make clean && make
...
rm i386-mmx-linux-generated.c x32-avx-avx512-linux-generated.c ...
Then, if you type make again, make will rebuild these files and rebuild
gdbserver. To avoid this, we can add the .SECONDARY special target. If
it has no pre-requisites, all intermediary files will be kept.
gdb/gdbserver/ChangeLog:
* Makefile.in (.SECONDARY): Define target.
Instead, make symtab_and_line initialize its members itself. Many
symtab_and_line declarations are moved to where the object is
initialized at the same time both for clarity and to avoid double
initialization. A few functions, like e.g., find_frame_sal are
adjusted to return the sal using normal function return instead of an
output parameter likewise to avoid having to default-construct a sal
and then immediately have the object overwritten.
gdb/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* ada-lang.c (is_known_support_routine): Move sal declaration to
where it is initialized.
* breakpoint.c (create_internal_breakpoint, init_catchpoint)
(parse_breakpoint_sals, decode_static_tracepoint_spec)
(clear_command, update_static_tracepoint): Remove init_sal
references. Move declarations closer to initializations.
* cli/cli-cmds.c (list_command): Move sal declarations closer to
initializations.
* elfread.c (elf_gnu_ifunc_resolver_stop): Remove init_sal
references. Move sal declarations closer to initializations.
* frame.c (find_frame_sal): Return a symtab_and_line via function
return instead of output parameter. Remove init_sal references.
* frame.h (find_frame_sal): Return a symtab_and_line via function
return instead of output parameter.
* guile/scm-frame.c (gdbscm_frame_sal): Adjust.
* guile/scm-symtab.c (stscm_make_sal_smob): Use in-place new
instead of memset.
(gdbscm_find_pc_line): Remove init_sal reference.
* infcall.c (call_function_by_hand_dummy): Remove init_sal
references. Move declarations closer to initializations.
* infcmd.c (set_step_frame): Update. Move declarations closer to
initializations.
(finish_backward): Remove init_sal references. Move declarations
closer to initializations.
* infrun.c (process_event_stop_test, handle_step_into_function)
(insert_hp_step_resume_breakpoint_at_frame)
(insert_step_resume_breakpoint_at_caller): Likewise.
* linespec.c (create_sals_line_offset, decode_digits_ordinary)
(symbol_to_sal): Likewise.
* probe.c (parse_probes_in_pspace): Remove init_sal reference.
* python/py-frame.c (frapy_find_sal): Move sal declaration closer
to its initialization.
* reverse.c (save_bookmark_command): Use new/delete. Remove
init_sal references. Move declarations closer to initializations.
* source.c (get_current_source_symtab_and_line): Remove brace
initialization.
(set_current_source_symtab_and_line): Now takes the sal by const
reference. Remove brace initialization.
(line_info): Remove init_sal reference.
* source.h (set_current_source_symtab_and_line): Now takes a
symtab_and_line via const reference.
* stack.c (set_current_sal_from_frame): Adjust.
(print_frame_info): Adjust.
(get_last_displayed_sal): Return the sal via function return
instead of via output parameter. Simplify.
(frame_info): Adjust.
* stack.h (get_last_displayed_sal): Return the sal via function
return instead of via output parameter.
* symtab.c (init_sal): Delete.
(find_pc_sect_line): Remove init_sal references. Move
declarations closer to initializations.
(find_function_start_sal): Remove init_sal references. Move
declarations closer to initializations.
* symtab.h (struct symtab_and_line): In-class initialize all
fields.
* tracepoint.c (set_traceframe_context)
(print_one_static_tracepoint_marker): Remove init_sal references.
Move declarations closer to initializations.
* tui/tui-disasm.c (tui_show_disassem_and_update_source): Adjust.
* tui/tui-stack.c (tui_show_frame_info): Adjust. Move
declarations closer to initializations.
* tui/tui-winsource.c (tui_update_source_window_as_is): Remove
init_sal references. Adjust.
This replaces "struct symtabs_and_lines" with
std::vector<symtab_and_line> in most cases. This removes a number of
cleanups.
In some cases, the sals objects do not own the sals they point at.
Instead they point at some sal that lives on the stack. Typically
something like this:
struct symtab_and_line sal;
struct symtabs_and_lines sals;
// fill in sal
sals.nelts = 1;
sals.sals = &sal;
// use sals
Instead of switching those cases to std::vector too, such usages are
replaced by gdb::array_view<symtab_and_line> instead. This avoids
introducing heap allocations.
gdb/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* ax-gdb.c (agent_command_1): Use range-for.
* break-catch-throw.c (re_set_exception_catchpoint): Update.
* breakpoint.c: Include "common/array-view.h".
(init_breakpoint_sal, create_breakpoint_sal): Change sals
parameter from struct symtabs_and_lines to
array_view<symtab_and_line>. Adjust. Use range-for. Update.
(breakpoint_sals_to_pc): Change sals parameter from struct
symtabs_and_lines to std::vector reference.
(check_fast_tracepoint_sals): Change sals parameter from struct
symtabs_and_lines to std::array_view. Use range-for.
(decode_static_tracepoint_spec): Return a std::vector instead of
symtabs_and_lines. Update.
(create_breakpoint): Update.
(break_range_command, until_break_command, clear_command): Update.
(base_breakpoint_decode_location, bkpt_decode_location)
(bkpt_probe_create_sals_from_location)
(bkpt_probe_decode_location, tracepoint_decode_location)
(tracepoint_probe_decode_location)
(strace_marker_create_sals_from_location): Return a std::vector
instead of symtabs_and_lines.
(strace_marker_create_breakpoints_sal): Update.
(strace_marker_decode_location): Return a std::vector instead of
symtabs_and_lines. Update.
(update_breakpoint_locations): Change struct symtabs_and_lines
parameters to gdb::array_view. Adjust.
(location_to_sals): Return a std::vector instead of
symtabs_and_lines. Update.
(breakpoint_re_set_default): Use std::vector instead of struct
symtabs_and_lines.
(decode_location_default): Return a std::vector instead of
symtabs_and_lines. Update.
* breakpoint.h: Include "common/array-view.h".
(struct breakpoint_ops) <decode_location>: Now returns a
std::vector instead of returning a symtabs_and_lines via output
parameter.
(update_breakpoint_locations): Change sals parameters to use
gdb::array_view.
* cli/cli-cmds.c (edit_command, list_command): Update to use
std::vector and gdb::array_view.
(ambiguous_line_spec): Adjust to use gdb::array_view and
range-for.
(compare_symtabs): Rename to ...
(cmp_symtabs): ... this. Change parameters to symtab_and_line
const reference and adjust.
(filter_sals): Rewrite using std::vector and standard algorithms.
* elfread.c (elf_gnu_ifunc_resolver_return_stop): Simplify.
(jump_command): Update to use std::vector.
* linespec.c (struct linespec_state) <canonical_names>: Update
comment.
(add_sal_to_sals_basic): Delete.
(add_sal_to_sals, filter_results, convert_results_to_lsals)
(decode_line_2, create_sals_line_offset)
(convert_address_location_to_sals, convert_linespec_to_sals)
(convert_explicit_location_to_sals, parse_linespec)
(event_location_to_sals, decode_line_full, decode_line_1)
(decode_line_with_current_source)
(decode_line_with_last_displayed, decode_objc)
(decode_digits_list_mode, decode_digits_ordinary, minsym_found)
(linespec_result::~linespec_result): Adjust to use std::vector
instead of symtabs_and_lines.
* linespec.h (linespec_sals::sals): Now a std::vector.
(struct linespec_result): Use std::vector, bool, and in-class
initialization.
(decode_line_1, decode_line_with_current_source)
(decode_line_with_last_displayed): Return std::vector.
* macrocmd.c (info_macros_command): Use std::vector.
* mi/mi-main.c (mi_cmd_trace_find): Use std::vector.
* probe.c (parse_probes_in_pspace, parse_probes): Adjust to use
std::vector.
* probe.h (parse_probes): Return a std::vector.
* python/python.c (gdbpy_decode_line): Use std::vector and
gdb::array_view.
* source.c (select_source_symtab, line_info): Use std::vector.
* stack.c (func_command): Use std::vector.
* symtab.h (struct symtabs_and_lines): Delete.
* tracepoint.c (tfind_line_command, scope_info): Use std::vector.
An array_view is an abstraction that provides a non-owning view over a
sequence of contiguous objects.
A way to put it is that array_view is to std::vector (and std::array
and built-in arrays with rank==1) like std::string_view is to
std::string.
The main intent of array_view is to use it as function input parameter
type, making it possible to pass in any sequence of contiguous
objects, irrespective of whether the objects live on the stack or heap
and what actual container owns them. Implicit construction from the
element type is supported too, making it easy to call functions that
expect an array of elements when you only have one element (usually on
the stack). For example:
struct A { .... };
void function (gdb::array_view<A> as);
std::vector<A> std_vec = ...;
std::array<A, N> std_array = ...;
A array[] = {...};
A elem;
function (std_vec);
function (std_array);
function (array);
function (elem);
Views can be either mutable or const. A const view is simply created
by specifying a const T as array_view template parameter, in which
case operator[] of non-const array_view objects ends up returning
const references. (Making the array_view itself const is analogous to
making a pointer itself be const. I.e., disables re-seating the
view/pointer.) Normally functions will pass around array_views by
value.
Uses of gdb::array_view (other than the ones in the unit tests) will
be added in a follow up patch.
gdb/ChangeLog
2017-09-04 Pedro Alves <palves@redhat.com>
* Makefile.in (SUBDIR_UNITTESTS_SRCS): Add
unittests/array-view-selftests.c.
(SUBDIR_UNITTESTS_OBS): Add array-view-selftests.o.
* common/array-view.h: New file.
* unittests/array-view-selftests.c: New file.
Currently, with "list LINESPEC1,LINESPEC2", if one of the linespecs is
ambiguous, i.e., if it expands to multiple locations, you get this
seemingly odd output:
(gdb) list foo,bar
file: "file0.c", line number: 26
file: "file1.c", line number: 29
Since "foo" above expands to multiple locations, the specified range
is indeterminate, and GDB is trying to be helpful by showing you what
was ambiguous. It looks confusing to me, though. I think it'd be
much more user friendly if GDB actually told you that, like this:
(gdb) list foo,bar
Specified first line 'foo' is ambiguous:
file: "file0.c", line number: 26
file: "file1.c", line number: 29
(gdb) list bar,foo
Specified last line 'foo' is ambiguous:
file: "file0.c", line number: 26
file: "file1.c", line number: 29
Note, I'm using "first" and "last" in the output because that's what
the manual uses:
~~~
list first,last
Print lines from first to last. [...]
~~~
Tested on x86-64 GNU/Linux.
gdb/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* cli/cli-cmds.c (edit_command): Pass message to
ambiguous_line_spec.
(list_command): Pass message to ambiguous_line_spec. Say
"first"/"last" instead of "start" and "end" to be consistent with
the manual.
(ambiguous_line_spec): Add 'format' and vararg parameters. Use
them to print formatted message.
gdb/testsuite/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* gdb.base/list-ambiguous.exp: New file.
* gdb.base/list-ambiguous0.c: New file.
* gdb.base/list-ambiguous1.c: New file.
* gdb.base/list.exp (test_list_range): Adjust expected output.
Fix build regression introduced by 0860c437cb ("btrace: Store
btrace_insn in an std::vector"):
src/gdb/btrace.c: In function ‘void ftrace_add_pt(btrace_thread_info*, pt_insn_decoder*, int*, std::vector<unsigned int>&)’:
src/gdb/btrace.c:1329:38: error: invalid initialization of reference of type ‘const btrace_insn&’ from expression of type ‘btrace_insn*’
ftrace_update_insns (bfun, &btinsn);
^
src/gdb/btrace.c:648:1: note: in passing argument 2 of ‘void ftrace_update_insns(btrace_function*, const btrace_insn&)’
ftrace_update_insns (struct btrace_function *bfun, const btrace_insn &insn)
^
gdb/ChangeLog:
2017-09-04 Pedro Alves <palves@redhat.com>
* btrace.c (ftrace_add_pt): Pass btrace_insn to
ftrace_update_insns by reference instead of pointer.
This patch remove the usage of tdesc_i386_mmx in i386-go32-tdep.c, and use
i386_target_description to get it instead.
gdb:
2017-09-04 Yao Qi <yao.qi@linaro.org>
* i386-go32-tdep.c: Include x86-xstate.h.
(i386_go32_init_abi): Call i386_target_description.
* i386-tdep.c (i386_target_description): Return tdesc_i386_mmx
if xcr0 is X86_XSTATE_X87_MASK.
* i386-tdep.h (tdesc_i386): Remove the declaration.
(tdesc_i386_mmx): Likewise.
i386fbsd_core_read_xcr0 reads the value of xcr0 from the corefile. If
it fails, returns 0. This makes its caller {i386,amd64}_target_description
has to handle this special value. IMO, i386fbsd_core_read_xcr0 should
return the default xcr0 in case of error.
gdb:
2017-09-04 Yao Qi <yao.qi@linaro.org>
* i386-fbsd-tdep.c (i386fbsd_core_read_xcr0): Return
X86_XSTATE_SSE_MASK instead of 0.
GDB can call function i386_target_description to get the right target
description rather than tdesc_i386
gdb:
2017-09-04 Yao Qi <yao.qi@linaro.org>
* amd64-fbsd-nat.c (amd64fbsd_read_description): Call
i386_target_description.
* i386-fbsd-nat.c (i386fbsd_read_description): Call
i386_target_description.
* i386-tdep.c (i386_gdbarch_init): Likewise.
This patch changes amd64-*-tdep.c files to use function
amd64_target_description to get the right target description rather than
use the variable tdesd_amd64.
gdb:
2017-09-04 Yao Qi <yao.qi@linaro.org>
* amd64-darwin-tdep.c: Include "x86-xstate.h".
(x86_darwin_init_abi_64): Call amd64_target_description.
* amd64-dicos-tdep.c: Likewise.
* amd64-fbsd-nat.c: Likewise.
* amd64-fbsd-tdep.c: Likewise.
* amd64-nbsd-tdep.c: Likewise.
* amd64-obsd-tdep.c: Likewise.
* amd64-sol2-tdep.c: Likewise.
* amd64-windows-tdep.c: Likewise.
* amd64-tdep.h (tdesc_amd64): Remove the declaration.
Because it contains a non-POD type field (flags), the type btrace_insn
should be new'ed/delete'd. Replace the VEC (btrace_insn_s) in
btrace_function with an std::vector.
gdb/ChangeLog:
* btrace.h (btrace_insn_s, DEF_VEC_O (btrace_insn_s)): Remove.
(btrace_function) <insn>: Change type to use std::vector.
* btrace.c (ftrace_debug, ftrace_call_num_insn,
ftrace_find_call, ftrace_new_gap, ftrace_update_function,
ftrace_update_insns, ftrace_compute_global_level_offset,
btrace_stitch_bts, btrace_clear, btrace_insn_get,
btrace_insn_end, btrace_insn_next, btrace_insn_prev): Adjust to
change to std::vector.
(ftrace_update_insns): Adjust to change to std::vector, change
type of INSN parameter.
(btrace_compute_ftrace_bts): Adjust call to ftrace_update_insns.
* record-btrace.c (btrace_call_history_insn_range,
btrace_compute_src_line_range,
record_btrace_frame_prev_register): Adjust to change to
std::vector.
* python/py-record-btrace.c (recpy_bt_func_instructions): Adjust
to change to std::vector.
This changes reopen_exec_file to use a std::string, removing a
cleanup.
ChangeLog
2017-09-03 Tom Tromey <tom@tromey.com>
* corefile.c (reopen_exec_file): Use std::string.
Change various things in the compile/ code to use std::string or
unique_xmalloc_ptr as appropriate. This allows the removal of some
cleanups.
ChangeLog
2017-09-03 Tom Tromey <tom@tromey.com>
* compile/compile.c (compile_register_name_mangled): Return
std::string.
* compile/compile-loc2c.c (pushf_register_address): Update.
(pushf_register): Update.
* compile/compile-c-types.c (convert_array): Update.
* compile/compile-c-symbols.c (generate_vla_size): Update.
(error_symbol_once): Use a gdb::unique_xmalloc_ptr.
(symbol_substitution_name): Return a gdb::unique_xmalloc_ptr.
(convert_one_symbol): Update.
(generate_c_for_for_one_variable): Update.
* compile/compile-c-support.c (c_get_range_decl_name): Return a
std::string.
(generate_register_struct): Update.
* compile/compile-internal.h (c_get_range_decl_name): Return a
std::string.
(compile_register_name_mangled): Return std::string.
Change perror_string to return a std::string, removing a cleanup in
the process.
ChangeLog
2017-09-03 Tom Tromey <tom@tromey.com>
* utils.c (perror_string): Return a std::string.
(throw_perror_with_name, perror_warning_with_name): Update.
Change demangle_command to use std::string and unique_xmalloc_ptr,
removing some cleanups.
ChangeLog
2017-09-03 Tom Tromey <tom@tromey.com>
* demangle.c (demangle_command): Use std::string,
unique_xmalloc_ptr.
Change do_set_command to use std::string, removing a cleanup and some
manual resizing code.
ChangeLog
2017-09-03 Tom Tromey <tom@tromey.com>
* cli/cli-setshow.c (do_set_command): Use std::string.
Change cd_command to use unique_xmalloc_ptr, removing a cleanup.
ChangeLog
2017-09-03 Tom Tromey <tom@tromey.com>
* cli/cli-cmds.c (cd_command): Use gdb::unique_xmalloc_ptr.
Change mi_cmd_interpreter_exec to use std::string, removing a cleanup.
ChangeLog
2017-09-03 Tom Tromey <tom@tromey.com>
* mi/mi-interp.c (mi_cmd_interpreter_exec): Use std::string.
Change env_execute_cli_command to use unique_xmalloc_ptr, removing a
cleanup.
ChangeLog
2017-09-03 Tom Tromey <tom@tromey.com>
* mi/mi-cmd-env.c (env_execute_cli_command): Use
gdb::unique_xmalloc_ptr.
This changes a few spots in thread.c to use std::string, removing some
cleanups.
ChangeLog
2017-09-03 Tom Tromey <tom@tromey.com>
* thread.c (print_thread_info_1): Use string_printf.
(thread_apply_command, thread_apply_all_command): Use
std::string.
This changes memory_error_message to return a std::string and fixes up
the callers. This removes some cleanups.
ChangeLog
2017-09-03 Tom Tromey <tom@tromey.com>
* valprint.c (val_print_string): Update.
* gdbcore.h (memory_error_message): Return std::string.
* corefile.c (memory_error_message): Return std::string.
(memory_error): Update.
* breakpoint.c (insert_bp_location): Update.
We need to set tlsdesc_plt for x86-64 if GOT_TLS_GDESC_P is true when
allocating dynamic relocations so that _bfd_x86_elf_size_dynamic_sections
will generate TLSDESC_PLT and TLSDESC_GOT in x86-64 output.
bfd/
PR ld/22071
* elfxx-x86.c (elf_x86_allocate_dynrelocs): Set tlsdesc_plt
for x86-64 if GOT_TLS_GDESC_P is true.
ld/
PR ld/22071
* testsuite/ld-x86-64/pr22071.d: New file.
* testsuite/ld-x86-64/pr22071.s: Likewise.
* testsuite/ld-x86-64/x86-64.exp: Run pr22071.
A quite straightforward change. It does "fix" leaks in record-btrace.c,
although since this is only used in debug printing code, it has no real
world impact.
gdb/ChangeLog:
* target/waitstatus.h (target_waitstatus_to_string): Change
return type to std::string.
* target/waitstatus.c (target_waitstatus_to_string): Return
std::string.
* target.h (target_waitstatus_to_string): Remove declaration.
* infrun.c (resume, clear_proceed_status_thread,
print_target_wait_results, do_target_wait, save_waitstatus,
stop_all_threads): Adjust.
* record-btrace.c (record_btrace_wait): Adjust.
* target-debug.h
(target_debug_print_struct_target_waitstatus_p): Adjust.
gdb/gdbserver/ChangeLog:
* linux-low.c (linux_wait_1): Adjust.
* server.c (queue_stop_reply_callback): Adjust.
Since elfxx-x86.lo needs elf64-x86-64.lo with 64-bit BFD now, add
elf64-x86-64.lo together with elfxx-x86.lo to bfd_backends for 64-bit
BFD.
* configure.ac (bfd_backends): Add elf64-x86-64.lo together
with elfxx-x86.lo for 64-bit BFD.
* configure: Regenerated.
elf_i386_size_dynamic_sections and elf_x86_64_size_dynamic_sections are
very similar, except for the followings:
1. elf_i386_size_dynamic_sections checks GOT_TLS_IE and GOT_TLS_IE_BOTH.
elf_x86_64_size_dynamic_sections checks only GOT_TLS_IE. Since
GOT_TLS_IE_BOTH is never true for x86-64, it is OK to check GOT_TLS_IE
for both i386 and x86-64.
2, x86-64 sets tlsdesc_plt, but i386 doesn't. We set tlsdesc_plt only
if target_id == X86_64_ELF_DATA.
3. x86-64 has
if (s != htab->elf.srelplt)
s->reloc_count = 0;
and i386 has
s->reloc_count = 0;
i386 did have
if (s != htab->srelplt)
s->reloc_count = 0;
in the original commit:
commit 67a4f2b710
Author: Alexandre Oliva <aoliva@redhat.com>
Date: Wed Jan 18 21:07:51 2006 +0000
But it was removed by
commit 5ae0bfb60a
Author: Richard Sandiford <rdsandiford@googlemail.com>
Date: Tue Feb 28 07:16:12 2006 +0000
bfd/
* elf32-i386.c (elf_i386_link_hash_table): Add next_tls_desc_index.
(elf_i386_link_hash_table_create): Initialize it.
(elf_i386_compute_jump_table_size): Use it instead of
srelplt->reloc_count.
(allocate_dynrelocs): Likewise.
(elf_i386_size_dynamic_sections): Likewise.
(elf_i386_relocate_section): Likewise.
A later commit:
commit e1f987424b
Author: H.J. Lu <hjl.tools@gmail.com>
Date: Fri Oct 21 15:13:37 2011 +0000
Put IRELATIVE relocations after JUMP_SLOT.
bfd/
2011-10-21 H.J. Lu <hongjiu.lu@intel.com>
PR ld/13302
* elf32-i386.c (elf_i386_link_hash_table): Add next_jump_slot_index
and next_irelative_index.
(elf_i386_link_hash_table_create): Initialize next_jump_slot_index
and next_irelative_index.
(elf_i386_allocate_dynrelocs): Increment reloc_count instead of
next_tls_desc_index.
(elf_i386_size_dynamic_sections): Set next_tls_desc_index and
next_irelative_index from reloc_count.
(elf_i386_finish_dynamic_symbol): Put R_386_IRELATIVE after
R_386_JUMP_SLOT.
changed it back to use reloc_count again. So it is correct to use
if (s != htab->elf.srelplt)
s->reloc_count = 0;
for both i386 and x86-64 now.
4. i386 and x86-64 use different DT_XXXs. They are handled by adding
them to elf_x86_link_hash_table.
With these changes, we can share _bfd_x86_elf_size_dynamic_sections in
elf32-i386.c and elf64-x86-64.c.
* elf32-i386.c (elf_i386_convert_load): Renamed to ...
(_bfd_i386_elf_convert_load): This. Remove static.
(elf_i386_size_dynamic_sections): Removed.
(elf_backend_size_dynamic_sections): Likewise.
* elf64-x86-64.c (elf_x86_64_convert_load): Renamed to ...
(_bfd_x86_64_elf_convert_load): This. Remove static.
(elf_x86_64_size_dynamic_sections): Removed.
(elf_backend_size_dynamic_sections): Likewise.
* elfxx-x86.c (_bfd_x86_elf_allocate_dynrelocs): Renamed to ...
(elf_x86_allocate_dynrelocs): This. Make it static.
(_bfd_x86_elf_allocate_local_dynrelocs): Renamed to ...
(elf_x86_allocate_local_dynreloc): This. Make it static.
(elf_i386_is_reloc_section): New function.
(elf_x86_64_is_reloc_section): Likewise.
(_bfd_x86_elf_link_hash_table_create): Initialize convert_load,
is_reloc_section, dt_reloc, dt_reloc_sz and dt_reloc_ent.
Rearrange got_entry_size initialization.
(_bfd_x86_elf_size_dynamic_sections): New function.
* elfxx-x86.h (elf_x86_link_hash_table): Add convert_load,
is_reloc_section, dt_reloc, dt_reloc_sz and dt_reloc_ent.
(_bfd_i386_elf_convert_load): New.
(_bfd_x86_64_elf_convert_load): Likewise.
(_bfd_x86_elf_size_dynamic_sections): Likewise.
(elf_backend_size_dynamic_sections): Likewise.
(_bfd_x86_elf_allocate_dynrelocs): Removed.
(_bfd_x86_elf_allocate_local_dynrelocs): Likewise.
elf_i386_size_dynamic_sections has
htab->next_tls_desc_index = htab->elf.srelplt->reloc_count;
htab->sgotplt_jump_table_size = htab->next_tls_desc_index * 4;
This patch changes it to
htab->sgotplt_jump_table_size
= elf_x86_compute_jump_table_size (htab)
Since elf_x86_compute_jump_table_size is defined as
((htab)->elf.srelplt->reloc_count * (htab)->got_entry_size)
there is no change in output. It makes elf_i386_size_dynamic_sections
the same as elf_x86_64_size_dynamic_sections.
* elf32-i386.c (elf_i386_size_dynamic_sections): Set
sgotplt_jump_table_size with elf_x86_compute_jump_table_size.
Since PLT_CIE_LENGTH, PLT_FDE_LENGTH, PLT_FDE_START_OFFSET and
PLT_FDE_LEN_OFFSET are identical in elf32-i386.c and elf64-x86-64.c,
they can be defined in elfxx-x86.h.
* elf32-i386.c (PLT_CIE_LENGTH, PLT_FDE_LENGTH,
PLT_FDE_START_OFFSET, PLT_FDE_LEN_OFFSET): Moved to ...
* elfxx-x86.h (PLT_CIE_LENGTH, PLT_FDE_LENGTH,
PLT_FDE_START_OFFSET, PLT_FDE_LEN_OFFSET): Here.
* elf64-x86-64.c (PLT_CIE_LENGTH, PLT_FDE_LENGTH,
PLT_FDE_START_OFFSET, PLT_FDE_LEN_OFFSET): Removed.