Handle partially optimized out values similarly to unavailable values

This fixes PR symtab/14604, PR symtab/14605, and Jan's test at
https://sourceware.org/ml/gdb-patches/2014-07/msg00158.html, in a tree
with bddbbed reverted:

 2014-07-22  Pedro Alves  <palves@redhat.com>

 	* value.c (allocate_optimized_out_value): Don't mark value as
 	non-lazy.

The PRs are about variables described by the DWARF as being split over
multiple registers using DWARF piece information, but some of those
registers being marked as optimised out (not saved) by a later frame.
GDB currently incorrectly mishandles these partially-optimized-out
values.

Even though we can usually tell from the debug info whether a local or
global is optimized out, handling the case of a local living in a
register that was not saved in a frame requires fetching the variable.
GDB also needs to fetch a value to tell whether parts of it are
"<unavailable>".  Given this, it's not worth it to try to avoid
fetching lazy optimized-out values based on debug info alone.

So this patch makes GDB track which chunks of a value's contents are
optimized out like it tracks <unavailable> contents.  That is, it
makes value->optimized_out be a bit range vector instead of a boolean,
and removes the struct lval_funcs check_validity and check_any_valid
hooks.

Unlike Andrew's series which this is based on (at
https://sourceware.org/ml/gdb-patches/2013-08/msg00300.html, note some
pieces have gone in since), this doesn't merge optimized out and
unavailable contents validity/availability behind a single interface,
nor does it merge the bit range vectors themselves (at least yet).
While it may be desirable to have a single entry point that returns
existence of contents irrespective of what may make them
invalid/unavailable, several places want to treat optimized out /
unavailable / etc. differently, so each spot that potentially could
use it will need to be careful considered on case-by-case basis, and
best done as a separate change.

This fixes Jan's test, because value_available_contents_eq wasn't
considering optimized out value contents.  It does now, and because of
that it's been renamed to value_contents_eq.

A new intro comment is added to value.h describing "<optimized out>",
"<not saved>" and "<unavailable>" values.

gdb/
	PR symtab/14604
	PR symtab/14605
	* ada-lang.c (coerce_unspec_val_to_type): Use
	value_contents_copy_raw.
	* ada-valprint.c (val_print_packed_array_elements): Adjust.
	* c-valprint.c (c_val_print): Use value_bits_any_optimized_out.
	* cp-valprint.c (cp_print_value_fields): Let the common printing
	code handle optimized out values.
	(cp_print_value_fields_rtti): Use value_bits_any_optimized_out.
	* d-valprint.c (dynamic_array_type): Use
	value_bits_any_optimized_out.
	* dwarf2loc.c (entry_data_value_funcs): Remove check_validity and
	check_any_valid fields.
	(check_pieced_value_bits): Delete and inline ...
	(check_pieced_synthetic_pointer): ... here.
	(check_pieced_value_validity): Delete.
	(check_pieced_value_invalid): Delete.
	(pieced_value_funcs): Remove check_validity and check_any_valid
	fields.
	(read_pieced_value): Use mark_value_bits_optimized_out.
	(write_pieced_value): Switch to use
	mark_value_bytes_optimized_out.
	(dwarf2_evaluate_loc_desc_full): Copy the value contents instead
	of assuming the whole value is optimized out.
	* findvar.c (read_frame_register_value): Remove special handling
	of optimized out registers.
	(value_from_register): Use mark_value_bytes_optimized_out.
	* frame-unwind.c (frame_unwind_got_optimized): Use
	mark_value_bytes_optimized_out.
	* jv-valprint.c (java_value_print): Adjust.
	(java_print_value_fields): Let the common printing code handle
	optimized out values.
	* mips-tdep.c (mips_print_register): Remove special handling of
	optimized out registers.
	* opencl-lang.c (lval_func_check_validity): Delete.
	(lval_func_check_any_valid): Delete.
	(opencl_value_funcs): Remove check_validity and check_any_valid
	fields.
	* p-valprint.c (pascal_object_print_value_fields): Let the common
	printing code handle optimized out values.
	* stack.c (read_frame_arg): Remove special handling of optimized
	out values.  Fetch both VAL and ENTRYVAL before comparing
	contents.  Adjust to value_available_contents_eq rename.
	* valprint.c (valprint_check_validity)
	(val_print_scalar_formatted): Use value_bits_any_optimized_out.
	(val_print_array_elements): Adjust.
	* value.c (struct value) <optimized_out>: Now a VEC(range_s).
	(value_bits_any_optimized_out): New function.
	(value_entirely_covered_by_range_vector): New function, factored
	out from value_entirely_unavailable.
	(value_entirely_unavailable): Reimplement.
	(value_entirely_optimized_out): New function.
	(insert_into_bit_range_vector): New function, factored out from
	mark_value_bits_unavailable.
	(mark_value_bits_unavailable): Reimplement.
	(struct ranges_and_idx): New struct.
	(find_first_range_overlap_and_match): New function, factored out
	from value_available_contents_bits_eq.
	(value_available_contents_bits_eq): Rename to ...
	(value_contents_bits_eq): ... this.  Check both unavailable
	contents and optimized out contents.
	(value_available_contents_eq): Rename to ...
	(value_contents_eq): ... this.
	(allocate_value_lazy): Remove reference to the old optimized_out
	boolean.
	(allocate_optimized_out_value): Use
	mark_value_bytes_optimized_out.
	(require_not_optimized_out): Adjust to check whether the
	optimized_out vec is empty.
	(ranges_copy_adjusted): New function, factored out from
	value_contents_copy_raw.
	(value_contents_copy_raw): Also copy the optimized out ranges.
	Assert the destination ranges aren't optimized out.
	(value_contents_copy): Update comment, remove call to
	require_not_optimized_out.
	(value_contents_equal): Adjust to check whether the optimized_out
	vec is empty.
	(set_value_optimized_out, value_optimized_out_const): Delete.
	(mark_value_bytes_optimized_out, mark_value_bits_optimized_out):
	New functions.
	(value_entirely_optimized_out, value_bits_valid): Delete.
	(value_copy): Take a VEC copy of the 'optimized_out' field.
	(value_primitive_field): Remove special handling of optimized out.
	(value_fetch_lazy): Assert that lazy values have no unavailable
	regions.  Use value_bits_any_optimized_out.  Remove some special
	handling for optimized out values.
	* value.h: Add intro comment about <optimized out> and
	<unavailable>.
	(struct lval_funcs): Remove check_validity and check_any_valid
	fields.
	(set_value_optimized_out, value_optimized_out_const): Remove.
	(mark_value_bytes_optimized_out, mark_value_bits_optimized_out):
	New declarations.
	(value_bits_any_optimized_out): New declaration.
	(value_bits_valid): Delete declaration.
	(value_available_contents_eq): Rename to ...
	(value_contents_eq): ... this, and extend comments.

gdb/testsuite/
	PR symtab/14604
	PR symtab/14605
	* gdb.dwarf2/dw2-op-out-param.exp: Remove kfail branches and use
	gdb_test.
This commit is contained in:
Pedro Alves
2014-08-20 00:07:40 +01:00
parent 6694c4110a
commit 9a0dc9e369
19 changed files with 552 additions and 448 deletions

View File

@ -33,6 +33,54 @@ struct language_defn;
struct value_print_options;
struct xmethod_worker;
/* Values can be partially 'optimized out' and/or 'unavailable'.
These are distinct states and have different string representations
and related error strings.
'unavailable' has a specific meaning in this context. It means the
value exists in the program (at the machine level), but GDB has no
means to get to it. Such a value is normally printed as
<unavailable>. Examples of how to end up with an unavailable value
would be:
- We're inspecting a traceframe, and the memory or registers the
debug information says the value lives on haven't been collected.
- We're inspecting a core dump, the memory or registers the debug
information says the value lives aren't present in the dump
(that is, we have a partial/trimmed core dump, or we don't fully
understand/handle the core dump's format).
- We're doing live debugging, but the debug API has no means to
get at where the value lives in the machine, like e.g., ptrace
not having access to some register or register set.
- Any other similar scenario.
OTOH, "optimized out" is about what the compiler decided to generate
(or not generate). A chunk of a value that was optimized out does
not actually exist in the program. There's no way to get at it
short of compiling the program differently.
A register that has not been saved in a frame is likewise considered
optimized out, except not-saved registers have a different string
representation and related error strings. E.g., we'll print them as
<not-saved> instead of <optimized out>, as in:
(gdb) p/x $rax
$1 = <not saved>
(gdb) info registers rax
rax <not saved>
If the debug info describes a variable as being in such a register,
we'll still print the variable as <optimized out>. IOW, <not saved>
is reserved for inspecting registers at the machine level.
When comparing value contents, optimized out chunks, unavailable
chunks, and valid contents data are all considered different. See
value_contents_eq for more info.
*/
/* The structure which defines the type of a value. It should never
be possible for a program lval value to survive over a call to the
inferior (i.e. to be put into the history list or an internal
@ -181,14 +229,6 @@ struct lval_funcs
TOVAL is not considered as an lvalue. */
void (*write) (struct value *toval, struct value *fromval);
/* Check the validity of some bits in VALUE. This should return 1
if all the bits starting at OFFSET and extending for LENGTH bits
are valid, or 0 if any bit is invalid. */
int (*check_validity) (const struct value *value, int offset, int length);
/* Return 1 if any bit in VALUE is valid, 0 if they are all invalid. */
int (*check_any_valid) (const struct value *value);
/* If non-NULL, this is used to implement pointer indirection for
this value. This method may return NULL, in which case value_ind
will fall back to ordinary indirection. */
@ -327,16 +367,29 @@ extern int value_fetch_lazy (struct value *val);
exist in the program, at least partially. If the value is lazy,
this may fetch it now. */
extern int value_optimized_out (struct value *value);
extern void set_value_optimized_out (struct value *value, int val);
/* Like value_optimized_out, but don't fetch the value even if it is
lazy. Mainly useful for constructing other values using VALUE as
template. */
extern int value_optimized_out_const (const struct value *value);
/* Given a value, return true if any of the contents bits starting at
OFFSET and extending for LENGTH bits is optimized out, false
otherwise. */
/* Like value_optimized_out, but return false if any bit in the object
is valid. */
extern int value_entirely_optimized_out (const struct value *value);
extern int value_bits_any_optimized_out (const struct value *value,
int bit_offset, int bit_length);
/* Like value_optimized_out, but return true iff the whole value is
optimized out. */
extern int value_entirely_optimized_out (struct value *value);
/* Mark VALUE's content bytes starting at OFFSET and extending for
LENGTH bytes as optimized out. */
extern void mark_value_bytes_optimized_out (struct value *value,
int offset, int length);
/* Mark VALUE's content bits starting at OFFSET and extending for
LENGTH bits as optimized out. */
extern void mark_value_bits_optimized_out (struct value *value,
int offset, int length);
/* Set or return field indicating whether a variable is initialized or
not, based on debugging information supplied by the compiler.
@ -415,13 +468,6 @@ extern struct value *coerce_ref (struct value *value);
extern struct value *coerce_array (struct value *value);
/* Given a value, determine whether the bits starting at OFFSET and
extending for LENGTH bits are valid. This returns nonzero if all
bits in the given range are valid, zero if any bit is invalid. */
extern int value_bits_valid (const struct value *value,
int offset, int length);
/* Given a value, determine whether the bits starting at OFFSET and
extending for LENGTH bits are a synthetic pointer. */
@ -473,35 +519,53 @@ extern void mark_value_bits_unavailable (struct value *value,
its enclosing type chunk, you'd do:
int len = TYPE_LENGTH (check_typedef (value_enclosing_type (val)));
value_available_contents (val, 0, val, 0, len);
value_contents_eq (val, 0, val, 0, len);
Returns true iff the set of available contents match. Unavailable
contents compare equal with unavailable contents, and different
with any available byte. For example, if 'x's represent an
unavailable byte, and 'V' and 'Z' represent different available
bytes, in a value with length 16:
Returns true iff the set of available/valid contents match.
offset: 0 4 8 12 16
contents: xxxxVVVVxxxxVVZZ
Optimized-out contents are equal to optimized-out contents, and are
not equal to non-optimized-out contents.
Unavailable contente are equal to unavailable contents, and are not
equal to non-unavailable contents.
For example, if 'x's represent an unavailable byte, and 'V' and 'Z'
represent different available/valid bytes, in a value with length
16:
offset: 0 4 8 12 16
contents: xxxxVVVVxxxxVVZZ
then:
value_available_contents_eq(val, 0, val, 8, 6) => 1
value_available_contents_eq(val, 0, val, 4, 4) => 1
value_available_contents_eq(val, 0, val, 8, 8) => 0
value_available_contents_eq(val, 4, val, 12, 2) => 1
value_available_contents_eq(val, 4, val, 12, 4) => 0
value_available_contents_eq(val, 3, val, 4, 4) => 0
value_contents_eq(val, 0, val, 8, 6) => 1
value_contents_eq(val, 0, val, 4, 4) => 0
value_contents_eq(val, 0, val, 8, 8) => 0
value_contents_eq(val, 4, val, 12, 2) => 1
value_contents_eq(val, 4, val, 12, 4) => 0
value_contents_eq(val, 3, val, 4, 4) => 0
We only know whether a value chunk is available if we've tried to
read it. As this routine is used by printing routines, which may
be printing values in the value history, long after the inferior is
gone, it works with const values. Therefore, this routine must not
be called with lazy values. */
If 'x's represent an unavailable byte, 'o' represents an optimized
out byte, in a value with length 8:
extern int value_available_contents_eq (const struct value *val1, int offset1,
const struct value *val2, int offset2,
int length);
offset: 0 4 8
contents: xxxxoooo
then:
value_contents_eq(val, 0, val, 2, 2) => 1
value_contents_eq(val, 4, val, 6, 2) => 1
value_contents_eq(val, 0, val, 4, 4) => 0
We only know whether a value chunk is unavailable or optimized out
if we've tried to read it. As this routine is used by printing
routines, which may be printing values in the value history, long
after the inferior is gone, it works with const values. Therefore,
this routine must not be called with lazy values. */
extern int value_contents_eq (const struct value *val1, int offset1,
const struct value *val2, int offset2,
int length);
/* Read LENGTH bytes of memory starting at MEMADDR into BUFFER, which
is (or will be copied to) VAL's contents buffer offset by