From patchwork Thu Jul 24 12:08:13 2014 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Pedro Alves X-Patchwork-Id: 2158 Received: (qmail 9745 invoked by alias); 24 Jul 2014 12:08:27 -0000 Mailing-List: contact gdb-patches-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Unsubscribe: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sourceware.org Delivered-To: mailing list gdb-patches@sourceware.org Received: (qmail 9698 invoked by uid 89); 24 Jul 2014 12:08:23 -0000 Authentication-Results: sourceware.org; auth=none X-Virus-Found: No X-Spam-SWARE-Status: No, score=-1.9 required=5.0 tests=AWL, BAYES_00, RP_MATCHES_RCVD, SPF_HELO_PASS, SPF_PASS autolearn=ham version=3.3.2 X-HELO: mx1.redhat.com Received: from mx1.redhat.com (HELO mx1.redhat.com) (209.132.183.28) by sourceware.org (qpsmtpd/0.93/v0.84-503-g423c35a) with (AES256-GCM-SHA384 encrypted) ESMTPS; Thu, 24 Jul 2014 12:08:17 +0000 Received: from int-mx14.intmail.prod.int.phx2.redhat.com (int-mx14.intmail.prod.int.phx2.redhat.com [10.5.11.27]) by mx1.redhat.com (8.14.4/8.14.4) with ESMTP id s6OC8FBF027730 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK) for ; Thu, 24 Jul 2014 08:08:15 -0400 Received: from [127.0.0.1] (ovpn01.gateway.prod.ext.ams2.redhat.com [10.39.146.11]) by int-mx14.intmail.prod.int.phx2.redhat.com (8.14.4/8.14.4) with ESMTP id s6OC8DgW003684; Thu, 24 Jul 2014 08:08:14 -0400 Message-ID: <53D0F72D.9010701@redhat.com> Date: Thu, 24 Jul 2014 13:08:13 +0100 From: Pedro Alves User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.6.0 MIME-Version: 1.0 To: Jan Kratochvil CC: gdb-patches@sourceware.org Subject: [PATCH v2] Handle partially optimized out values similarly to unavailable values References: <20140709103312.GA27884@host2.jankratochvil.net> <53BD2CE0.1000308@redhat.com> <20140709153121.GA7989@host2.jankratochvil.net> <53C41D5D.9030109@redhat.com> <20140716215838.GA29855@host2.jankratochvil.net> In-Reply-To: <20140716215838.GA29855@host2.jankratochvil.net> On 07/16/2014 10:58 PM, Jan Kratochvil wrote: > On Mon, 14 Jul 2014 20:11:41 +0200, Pedro Alves wrote: > [...] >> --- a/gdb/value.c >> +++ b/gdb/value.c > [...] >> @@ -339,6 +330,15 @@ struct value >> value to be available. This is filled in at value read time. The >> unavailable ranges are tracked in bits. */ >> VEC(range_s) *unavailable; >> + >> + /* Likewise, but for optimized out contents (a chunk of the value of >> + a variable that does not actually exist in the program). If LVAL >> + is lval_register, this is a register ($pc, $sp, etc., never a >> + program variable) that has not been saved in the frame. Not >> + saved registers and optimized-out program variables values are >> + treated pretty much the same, except not-saved registers have a >> + different string representation and related error strings. */ >> + VEC(range_s) *optimized_out; > > I miss here explicit description if a bit of the range can appear on both > 'unavailable' and 'optimized_out'. IMO it should not. Yeah, it should not. Added: @@ -327,8 +327,10 @@ struct value /* Unavailable ranges in CONTENTS. We mark unavailable ranges, rather than available, since the common and default case is for a - value to be available. This is filled in at value read time. The - unavailable ranges are tracked in bits. */ + value to be available. This is filled in at value read time. + The unavailable ranges are tracked in bits. Note that a contents + bit that has been optimized out doesn't really exist in the + program, so it can't be marked unavailable either. */ VEC(range_s) *unavailable; >> @@ -1566,7 +1671,6 @@ value_copy (struct value *arg) >> VALUE_FRAME_ID (val) = VALUE_FRAME_ID (arg); >> VALUE_REGNUM (val) = VALUE_REGNUM (arg); >> val->lazy = arg->lazy; >> - val->optimized_out = arg->optimized_out; > > Why 'optimized_out' no longer needs to be copied? Hmm, I blindly assumed the function was using value_contents_copy. Copying that now. Thanks for spotting this. >> @@ -3662,10 +3753,11 @@ value_fetch_lazy (struct value *val) >> if (value_lazy (parent)) >> value_fetch_lazy (parent); >> >> - if (!value_bits_valid (parent, >> - TARGET_CHAR_BIT * offset + value_bitpos (val), >> - value_bitsize (val))) >> - set_value_optimized_out (val, 1); >> + if (value_bits_any_optimized_out (parent, >> + TARGET_CHAR_BIT * offset + value_bitpos (val), >> + value_bitsize (val))) >> + mark_value_bytes_optimized_out (val, value_embedded_offset (val), >> + TYPE_LENGTH (type)); > > That was required before but it could handle partially optimized out > bitfields; but maybe it is not useful much. Yeah, though it doesn't look trivial. We should be able to copy the contents straight from the parent to VAL, but note we're copying only a few bits off the parent, into a wider VAL. So it seems like we'd need to: - have a value_contents_copy function that knows to copy bits, not bytes. - handle sign/zero extension - and handle big/little endian at that Best leave that out for now. >> >> +/* 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. Examples > > "printed as." - is the sentence finished? Whoops, it wasn't, thanks. I meant to write: "Such a value is normally printed as ." > > >> + 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 > > s///? Nope, we really print them as . More here: https://sourceware.org/ml/gdb-patches/2013-09/msg00176.html I've extended the comment based on that email: @@ -60,11 +60,21 @@ struct xmethod_worker; 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 - instead of . + 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 + instead of , as in: + + (gdb) p/x $rax + $1 = + (gdb) info registers rax + rax + + If the debug info describes a variable as being in such a register, + we'll still print the variable as . IOW, + is reserved for inspecting registers at the machine level. >> + Returns true iff the set of available/valid contents match. >> + >> + Optimized out contents compare equal with optimized out contents, >> + and different with any non-optimized-out byte. > > English: s/different/differ/ The "different" is bound to the "compare" before, so it results in "contents compare ... different". I don't know how I'd use differ in a similar vein in the above. Sounds like I best rewrite this then. Here's what I came up with: @@ -513,11 +523,11 @@ extern void mark_value_bits_unavailable (struct value *value, Returns true iff the set of available/valid contents match. - Optimized out contents compare equal with optimized out contents, - and different with any non-optimized-out contents. + Optimized-out contents are equal to optimized-out contents, and are + not equal to non-optimized-out contents. - Unavailable contents compare equal with unavailable contents, and - different with non-unavailable contents. + Unavailable contents are equal to unavailable contents, and are not + equal to non-unavailable contents. > Maybe it would be all easier to have there single map and each range mapping > into enum { VALID, OPTIMIZED_OUT, UNAVAILABLE }. Yeah, as mentioned in the commit log, that was part of Andrew's original series, but I'd rather take baby steps and do that separately, as it required new code in the range merging code. New patch below. Let me know what you think. Thanks for the review. 8<------------ From d856ed9cb43ad48829c805a88dc6da55bfa149f0 Mon Sep 17 00:00:00 2001 From: Pedro Alves Date: Thu, 24 Jul 2014 12:48:35 +0100 Subject: [PATCH] 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 * 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 "". 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 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 "", "" and "" values. gdb/ 2014-07-24 Andrew Burgess Pedro Alves 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) : 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 and . (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/ 2014-07-24 Andrew Burgess Pedro Alves PR symtab/14604 PR symtab/14605 * gdb.dwarf2/dw2-op-out-param.exp: Remove kfail branches and use gdb_test. --- gdb/ada-lang.c | 4 +- gdb/ada-valprint.c | 6 +- gdb/c-valprint.c | 6 +- gdb/cp-valprint.c | 11 +- gdb/d-valprint.c | 5 +- gdb/dwarf2loc.c | 81 +---- gdb/findvar.c | 8 +- gdb/frame-unwind.c | 2 +- gdb/jv-valprint.c | 13 +- gdb/mips-tdep.c | 6 - gdb/opencl-lang.c | 54 --- gdb/p-valprint.c | 5 - gdb/stack.c | 17 +- gdb/testsuite/gdb.dwarf2/dw2-op-out-param.exp | 30 +- gdb/valprint.c | 22 +- gdb/value.c | 467 +++++++++++++++----------- gdb/value.h | 158 ++++++--- 17 files changed, 445 insertions(+), 450 deletions(-) diff --git a/gdb/ada-lang.c b/gdb/ada-lang.c index 6956909..63f84ea 100644 --- a/gdb/ada-lang.c +++ b/gdb/ada-lang.c @@ -682,14 +682,12 @@ coerce_unspec_val_to_type (struct value *val, struct type *type) else { result = allocate_value (type); - memcpy (value_contents_raw (result), value_contents (val), - TYPE_LENGTH (type)); + value_contents_copy_raw (result, 0, val, 0, TYPE_LENGTH (type)); } set_value_component_location (result, val); set_value_bitsize (result, value_bitsize (val)); set_value_bitpos (result, value_bitpos (val)); set_value_address (result, value_address (val)); - set_value_optimized_out (result, value_optimized_out_const (val)); return result; } } diff --git a/gdb/ada-valprint.c b/gdb/ada-valprint.c index b7b3a9c..b4f191f 100644 --- a/gdb/ada-valprint.c +++ b/gdb/ada-valprint.c @@ -186,9 +186,9 @@ val_print_packed_array_elements (struct type *type, const gdb_byte *valaddr, (i * bitsize) / HOST_CHAR_BIT, (i * bitsize) % HOST_CHAR_BIT, bitsize, elttype); - if (!value_available_contents_eq (v0, value_embedded_offset (v0), - v1, value_embedded_offset (v1), - eltlen)) + if (!value_contents_eq (v0, value_embedded_offset (v0), + v1, value_embedded_offset (v1), + eltlen)) break; } diff --git a/gdb/c-valprint.c b/gdb/c-valprint.c index f4694b0..2f2789c 100644 --- a/gdb/c-valprint.c +++ b/gdb/c-valprint.c @@ -173,9 +173,9 @@ c_val_print (struct type *type, const gdb_byte *valaddr, options->format) && value_bytes_available (original_value, embedded_offset, TYPE_LENGTH (type)) - && value_bits_valid (original_value, - TARGET_CHAR_BIT * embedded_offset, - TARGET_CHAR_BIT * TYPE_LENGTH (type))) + && !value_bits_any_optimized_out (original_value, + TARGET_CHAR_BIT * embedded_offset, + TARGET_CHAR_BIT * TYPE_LENGTH (type))) { int force_ellipses = 0; diff --git a/gdb/cp-valprint.c b/gdb/cp-valprint.c index 3e1d6ed..0d6c3fc 100644 --- a/gdb/cp-valprint.c +++ b/gdb/cp-valprint.c @@ -294,12 +294,6 @@ cp_print_value_fields (struct type *type, struct type *real_type, { fputs_filtered (_(""), stream); } - else if (!value_bits_valid (val, - TYPE_FIELD_BITPOS (type, i), - TYPE_FIELD_BITSIZE (type, i))) - { - val_print_optimized_out (val, stream); - } else { struct value_print_options opts = *options; @@ -434,8 +428,9 @@ cp_print_value_fields_rtti (struct type *type, /* We require all bits to be valid in order to attempt a conversion. */ - if (value_bits_valid (val, TARGET_CHAR_BIT * offset, - TARGET_CHAR_BIT * TYPE_LENGTH (type))) + if (!value_bits_any_optimized_out (val, + TARGET_CHAR_BIT * offset, + TARGET_CHAR_BIT * TYPE_LENGTH (type))) { struct value *value; int full, top, using_enc; diff --git a/gdb/d-valprint.c b/gdb/d-valprint.c index 755f180..788b3d5 100644 --- a/gdb/d-valprint.c +++ b/gdb/d-valprint.c @@ -38,8 +38,9 @@ dynamic_array_type (struct type *type, const gdb_byte *valaddr, && TYPE_CODE (TYPE_FIELD_TYPE (type, 0)) == TYPE_CODE_INT && strcmp (TYPE_FIELD_NAME (type, 0), "length") == 0 && strcmp (TYPE_FIELD_NAME (type, 1), "ptr") == 0 - && value_bits_valid (val, TARGET_CHAR_BIT * embedded_offset, - TARGET_CHAR_BIT * TYPE_LENGTH (type))) + && !value_bits_any_optimized_out (val, + TARGET_CHAR_BIT * embedded_offset, + TARGET_CHAR_BIT * TYPE_LENGTH (type))) { CORE_ADDR addr; struct type *elttype; diff --git a/gdb/dwarf2loc.c b/gdb/dwarf2loc.c index b1c7ee1..588abac 100644 --- a/gdb/dwarf2loc.c +++ b/gdb/dwarf2loc.c @@ -1285,8 +1285,6 @@ static const struct lval_funcs entry_data_value_funcs = { NULL, /* read */ NULL, /* write */ - NULL, /* check_validity */ - NULL, /* check_any_valid */ NULL, /* indirect */ entry_data_value_coerce_ref, NULL, /* check_synthetic_pointer */ @@ -1695,7 +1693,7 @@ read_pieced_value (struct value *v) memset (buffer, 0, this_size); if (optim) - set_value_optimized_out (v, 1); + mark_value_bits_optimized_out (v, offset, this_size_bits); if (unavail) mark_value_bits_unavailable (v, offset, this_size_bits); } @@ -1755,7 +1753,7 @@ read_pieced_value (struct value *v) break; case DWARF_VALUE_OPTIMIZED_OUT: - set_value_optimized_out (v, 1); + mark_value_bits_optimized_out (v, offset, this_size_bits); break; default: @@ -1793,7 +1791,7 @@ write_pieced_value (struct value *to, struct value *from) if (frame == NULL) { - set_value_optimized_out (to, 1); + mark_value_bytes_optimized_out (to, 0, TYPE_LENGTH (value_type (to))); return; } @@ -1925,7 +1923,7 @@ write_pieced_value (struct value *to, struct value *from) source_buffer, this_size); break; default: - set_value_optimized_out (to, 1); + mark_value_bytes_optimized_out (to, 0, TYPE_LENGTH (value_type (to))); break; } offset += this_size_bits; @@ -1934,24 +1932,16 @@ write_pieced_value (struct value *to, struct value *from) do_cleanups (cleanup); } -/* A helper function that checks bit validity in a pieced value. - CHECK_FOR indicates the kind of validity checking. - DWARF_VALUE_MEMORY means to check whether any bit is valid. - DWARF_VALUE_OPTIMIZED_OUT means to check whether any bit is - optimized out. - DWARF_VALUE_IMPLICIT_POINTER means to check whether the bits are an - implicit pointer. */ +/* An implementation of an lval_funcs method to see whether a value is + a synthetic pointer. */ static int -check_pieced_value_bits (const struct value *value, int bit_offset, - int bit_length, - enum dwarf_value_location check_for) +check_pieced_synthetic_pointer (const struct value *value, int bit_offset, + int bit_length) { struct piece_closure *c = (struct piece_closure *) value_computed_closure (value); int i; - int validity = (check_for == DWARF_VALUE_MEMORY - || check_for == DWARF_VALUE_IMPLICIT_POINTER); bit_offset += 8 * value_offset (value); if (value_bitsize (value)) @@ -1976,52 +1966,11 @@ check_pieced_value_bits (const struct value *value, int bit_offset, else bit_length -= this_size_bits; - if (check_for == DWARF_VALUE_IMPLICIT_POINTER) - { - if (p->location != DWARF_VALUE_IMPLICIT_POINTER) - return 0; - } - else if (p->location == DWARF_VALUE_OPTIMIZED_OUT - || p->location == DWARF_VALUE_IMPLICIT_POINTER) - { - if (validity) - return 0; - } - else - { - if (!validity) - return 1; - } + if (p->location != DWARF_VALUE_IMPLICIT_POINTER) + return 0; } - return validity; -} - -static int -check_pieced_value_validity (const struct value *value, int bit_offset, - int bit_length) -{ - return check_pieced_value_bits (value, bit_offset, bit_length, - DWARF_VALUE_MEMORY); -} - -static int -check_pieced_value_invalid (const struct value *value) -{ - return check_pieced_value_bits (value, 0, - 8 * TYPE_LENGTH (value_type (value)), - DWARF_VALUE_OPTIMIZED_OUT); -} - -/* An implementation of an lval_funcs method to see whether a value is - a synthetic pointer. */ - -static int -check_pieced_synthetic_pointer (const struct value *value, int bit_offset, - int bit_length) -{ - return check_pieced_value_bits (value, bit_offset, bit_length, - DWARF_VALUE_IMPLICIT_POINTER); + return 1; } /* A wrapper function for get_frame_address_in_block. */ @@ -2170,8 +2119,6 @@ free_pieced_value_closure (struct value *v) static const struct lval_funcs pieced_value_funcs = { read_pieced_value, write_pieced_value, - check_pieced_value_validity, - check_pieced_value_invalid, indirect_pieced_value, NULL, /* coerce_ref */ check_pieced_synthetic_pointer, @@ -2299,6 +2246,8 @@ dwarf2_evaluate_loc_desc_full (struct type *type, struct frame_info *frame, retval = value_from_register (type, gdb_regnum, frame); if (value_optimized_out (retval)) { + struct value *tmp; + /* This means the register has undefined value / was not saved. As we're computing the location of some variable etc. in the program, not a value for @@ -2306,7 +2255,9 @@ dwarf2_evaluate_loc_desc_full (struct type *type, struct frame_info *frame, generic optimized out value instead, so that we show instead of . */ do_cleanups (value_chain); - retval = allocate_optimized_out_value (type); + tmp = allocate_value (type); + value_contents_copy (tmp, 0, retval, 0, TYPE_LENGTH (type)); + retval = tmp; } } break; diff --git a/gdb/findvar.c b/gdb/findvar.c index 9390c8a..346a499 100644 --- a/gdb/findvar.c +++ b/gdb/findvar.c @@ -681,12 +681,6 @@ read_frame_register_value (struct value *value, struct frame_info *frame) struct value *regval = get_frame_register_value (frame, regnum); int reg_len = TYPE_LENGTH (value_type (regval)) - reg_offset; - if (value_optimized_out (regval)) - { - set_value_optimized_out (value, 1); - break; - } - /* If the register length is larger than the number of bytes remaining to copy, then only copy the appropriate bytes. */ if (reg_len > len) @@ -732,7 +726,7 @@ value_from_register (struct type *type, int regnum, struct frame_info *frame) if (!ok) { if (optim) - set_value_optimized_out (v, 1); + mark_value_bytes_optimized_out (v, 0, TYPE_LENGTH (type)); if (unavail) mark_value_bytes_unavailable (v, 0, TYPE_LENGTH (type)); } diff --git a/gdb/frame-unwind.c b/gdb/frame-unwind.c index fdfea6e..1cc3303 100644 --- a/gdb/frame-unwind.c +++ b/gdb/frame-unwind.c @@ -203,7 +203,7 @@ frame_unwind_got_optimized (struct frame_info *frame, int regnum) "". */ val = allocate_value_lazy (type); set_value_lazy (val, 0); - set_value_optimized_out (val, 1); + mark_value_bytes_optimized_out (val, 0, TYPE_LENGTH (type)); VALUE_LVAL (val) = lval_register; VALUE_REGNUM (val) = regnum; VALUE_FRAME_ID (val) = get_frame_id (frame); diff --git a/gdb/jv-valprint.c b/gdb/jv-valprint.c index a7bb494..93a827e 100644 --- a/gdb/jv-valprint.c +++ b/gdb/jv-valprint.c @@ -183,10 +183,10 @@ java_value_print (struct value *val, struct ui_file *stream, set_value_offset (next_v, value_offset (next_v) + TYPE_LENGTH (el_type)); value_fetch_lazy (next_v); - if (!(value_available_contents_eq - (v, value_embedded_offset (v), - next_v, value_embedded_offset (next_v), - TYPE_LENGTH (el_type)))) + if (!value_contents_eq (v, value_embedded_offset (v), + next_v, + value_embedded_offset (next_v), + TYPE_LENGTH (el_type))) break; } @@ -393,11 +393,6 @@ java_print_value_fields (struct type *type, const gdb_byte *valaddr, { fputs_filtered (_(""), stream); } - else if (!value_bits_valid (val, TYPE_FIELD_BITPOS (type, i), - TYPE_FIELD_BITSIZE (type, i))) - { - val_print_optimized_out (val, stream); - } else { struct value_print_options opts; diff --git a/gdb/mips-tdep.c b/gdb/mips-tdep.c index c3e8e77..34b75f7 100644 --- a/gdb/mips-tdep.c +++ b/gdb/mips-tdep.c @@ -6196,12 +6196,6 @@ mips_print_register (struct ui_file *file, struct frame_info *frame, } val = get_frame_register_value (frame, regnum); - if (value_optimized_out (val)) - { - fprintf_filtered (file, "%s: [Invalid]", - gdbarch_register_name (gdbarch, regnum)); - return; - } fputs_filtered (gdbarch_register_name (gdbarch, regnum), file); diff --git a/gdb/opencl-lang.c b/gdb/opencl-lang.c index 2dd76fa..6a3ba69 100644 --- a/gdb/opencl-lang.c +++ b/gdb/opencl-lang.c @@ -240,58 +240,6 @@ lval_func_write (struct value *v, struct value *fromval) value_free_to_mark (mark); } -/* Return nonzero if all bits in V within OFFSET and LENGTH are valid. */ - -static int -lval_func_check_validity (const struct value *v, int offset, int length) -{ - struct lval_closure *c = (struct lval_closure *) value_computed_closure (v); - /* Size of the target type in bits. */ - int elsize = - TYPE_LENGTH (TYPE_TARGET_TYPE (check_typedef (value_type (c->val)))) * 8; - int startrest = offset % elsize; - int start = offset / elsize; - int endrest = (offset + length) % elsize; - int end = (offset + length) / elsize; - int i; - - if (endrest) - end++; - - if (end > c->n) - return 0; - - for (i = start; i < end; i++) - { - int comp_offset = (i == start) ? startrest : 0; - int comp_length = (i == end) ? endrest : elsize; - - if (!value_bits_valid (c->val, c->indices[i] * elsize + comp_offset, - comp_length)) - return 0; - } - - return 1; -} - -/* Return nonzero if any bit in V is valid. */ - -static int -lval_func_check_any_valid (const struct value *v) -{ - struct lval_closure *c = (struct lval_closure *) value_computed_closure (v); - /* Size of the target type in bits. */ - int elsize = - TYPE_LENGTH (TYPE_TARGET_TYPE (check_typedef (value_type (c->val)))) * 8; - int i; - - for (i = 0; i < c->n; i++) - if (value_bits_valid (c->val, c->indices[i] * elsize, elsize)) - return 1; - - return 0; -} - /* Return nonzero if bits in V from OFFSET and LENGTH represent a synthetic pointer. */ @@ -358,8 +306,6 @@ static const struct lval_funcs opencl_value_funcs = { lval_func_read, lval_func_write, - lval_func_check_validity, - lval_func_check_any_valid, NULL, /* indirect */ NULL, /* coerce_ref */ lval_func_check_synthetic_pointer, diff --git a/gdb/p-valprint.c b/gdb/p-valprint.c index 70eb3a8..c796218 100644 --- a/gdb/p-valprint.c +++ b/gdb/p-valprint.c @@ -627,11 +627,6 @@ pascal_object_print_value_fields (struct type *type, const gdb_byte *valaddr, { fputs_filtered (_(""), stream); } - else if (!value_bits_valid (val, TYPE_FIELD_BITPOS (type, i), - TYPE_FIELD_BITSIZE (type, i))) - { - val_print_optimized_out (val, stream); - } else { struct value_print_options opts = *options; diff --git a/gdb/stack.c b/gdb/stack.c index 4db5df5..346ce29 100644 --- a/gdb/stack.c +++ b/gdb/stack.c @@ -385,9 +385,12 @@ read_frame_arg (struct symbol *sym, struct frame_info *frame, { struct type *type = value_type (val); - if (!value_optimized_out (val) - && value_available_contents_eq (val, 0, entryval, 0, - TYPE_LENGTH (type))) + if (value_lazy (val)) + value_fetch_lazy (val); + if (value_lazy (entryval)) + value_fetch_lazy (entryval); + + if (value_contents_eq (val, 0, entryval, 0, TYPE_LENGTH (type))) { /* Initialize it just to avoid a GCC false warning. */ struct value *val_deref = NULL, *entryval_deref; @@ -413,11 +416,9 @@ read_frame_arg (struct symbol *sym, struct frame_info *frame, /* If the reference addresses match but dereferenced content does not match print them. */ if (val != val_deref - && !value_optimized_out (val_deref) - && !value_optimized_out (entryval_deref) - && value_available_contents_eq (val_deref, 0, - entryval_deref, 0, - TYPE_LENGTH (type_deref))) + && value_contents_eq (val_deref, 0, + entryval_deref, 0, + TYPE_LENGTH (type_deref))) val_equal = 1; } diff --git a/gdb/testsuite/gdb.dwarf2/dw2-op-out-param.exp b/gdb/testsuite/gdb.dwarf2/dw2-op-out-param.exp index 5ab5817..d245951 100644 --- a/gdb/testsuite/gdb.dwarf2/dw2-op-out-param.exp +++ b/gdb/testsuite/gdb.dwarf2/dw2-op-out-param.exp @@ -50,36 +50,12 @@ gdb_test "bt" "#0 ($hex in )?breakpt \\(\\)\r\n#1 ($hex in )?int_param_single_ # (2) struct_param_single_reg_loc gdb_continue_to_breakpoint "Stop in breakpt for struct_param_single_reg_loc" -set test "Backtrace for test struct_param_single_reg_loc" -gdb_test_multiple "bt" "$test" { - -re "#0 ($hex in )?breakpt \\(\\)\r\n#1 ($hex in )?struct_param_single_reg_loc \\(operand0={a = 0xdeadbe00deadbe01, b = }, operand1={a = , b = 0xdeadbe04deadbe05}, operand2=\\)\r\n#2 ($hex in )?main \\(\\)\r\n$gdb_prompt $" { - xpass $test - } - -re "#0 ($hex in )?breakpt \\(\\)\r\n#1 ($hex in )?struct_param_single_reg_loc \\(operand0=, operand1=, operand2=\\)\r\n#2 ($hex in )?main \\(\\)\r\n$gdb_prompt $" { - kfail "symtab/14604" $test - } -} +gdb_test "bt" "#0 ($hex in )?breakpt \\(\\)\r\n#1 ($hex in )?struct_param_single_reg_loc \\(operand0={a = 0xdeadbe00deadbe01, b = }, operand1={a = , b = 0xdeadbe04deadbe05}, operand2=\\)\r\n#2 ($hex in )?main \\(\\)" "Backtrace for test struct_param_single_reg_loc" # (3) struct_param_two_reg_pieces gdb_continue_to_breakpoint "Stop in breakpt for struct_param_two_reg_pieces" -set test "Backtrace for test struct_param_two_reg_pieces" -gdb_test_multiple "bt" "$test" { - -re "#0 ($hex in )?breakpt \\(\\)\r\n#1 ($hex in )?struct_param_two_reg_pieces \\(operand0={a = 0xdeadbe04deadbe05, b = }, operand1={a = , b = 0xdeadbe00deadbe01}, operand2=\\)\r\n#2 ($hex in )?main \\(\\)\r\n$gdb_prompt $" { - xpass $test - } - -re "#0 ($hex in )?breakpt \\(\\)\r\n#1 ($hex in )?struct_param_two_reg_pieces \\(operand0=.*, operand1=.*, operand2=.*\\)\r\n#2 ($hex in )?main \\(\\)\r\n$gdb_prompt $" { - kfail "symtab/14605" $test - } -} +gdb_test "bt" "#0 ($hex in )?breakpt \\(\\)\r\n#1 ($hex in )?struct_param_two_reg_pieces \\(operand0={a = 0xdeadbe04deadbe05, b = }, operand1={a = , b = 0xdeadbe00deadbe01}, operand2=\\)\r\n#2 ($hex in )?main \\(\\)" "Backtrace for test struct_param_two_reg_pieces" # (4) int_param_two_reg_pieces gdb_continue_to_breakpoint "Stop in breakpt for int_param_two_reg_pieces" -set test "Backtrace for test int_param_two_reg_pieces" -gdb_test_multiple "bt" "$test" { - -re "#0 ($hex in )?breakpt \\(\\)\r\n#1 ($hex in )?int_param_two_reg_pieces \\(operand0=, operand1=, operand2=\\)\r\n#2 ($hex in )?main \\(\\)\r\n$gdb_prompt $" { - xpass $test - } - -re "#0 ($hex in )?breakpt \\(\\)\r\n#1 ($hex in )?int_param_two_reg_pieces \\(operand0=.*, operand1=.*, operand2=.*\\)\r\n#2 ($hex in )?main \\(\\)\r\n$gdb_prompt $" { - kfail "symtab/14605" $test - } -} +gdb_test "bt" "#0 ($hex in )?breakpt \\(\\)\r\n#1 ($hex in )?int_param_two_reg_pieces \\(operand0=, operand1=, operand2=\\)\r\n#2 ($hex in )?main \\(\\)" "Backtrace for test int_param_two_reg_pieces" diff --git a/gdb/valprint.c b/gdb/valprint.c index 8600b34..981ef9b 100644 --- a/gdb/valprint.c +++ b/gdb/valprint.c @@ -311,8 +311,9 @@ valprint_check_validity (struct ui_file *stream, && TYPE_CODE (type) != TYPE_CODE_STRUCT && TYPE_CODE (type) != TYPE_CODE_ARRAY) { - if (!value_bits_valid (val, TARGET_CHAR_BIT * embedded_offset, - TARGET_CHAR_BIT * TYPE_LENGTH (type))) + if (value_bits_any_optimized_out (val, + TARGET_CHAR_BIT * embedded_offset, + TARGET_CHAR_BIT * TYPE_LENGTH (type))) { val_print_optimized_out (val, stream); return 0; @@ -983,8 +984,9 @@ val_print_scalar_formatted (struct type *type, /* A scalar object that does not have all bits available can't be printed, because all bits contribute to its representation. */ - if (!value_bits_valid (val, TARGET_CHAR_BIT * embedded_offset, - TARGET_CHAR_BIT * TYPE_LENGTH (type))) + if (value_bits_any_optimized_out (val, + TARGET_CHAR_BIT * embedded_offset, + TARGET_CHAR_BIT * TYPE_LENGTH (type))) val_print_optimized_out (val, stream); else if (!value_bytes_available (val, embedded_offset, TYPE_LENGTH (type))) val_print_unavailable (stream); @@ -1685,12 +1687,12 @@ val_print_array_elements (struct type *type, if (options->repeat_count_threshold < UINT_MAX) { while (rep1 < len - && value_available_contents_eq (val, - embedded_offset + i * eltlen, - val, - (embedded_offset - + rep1 * eltlen), - eltlen)) + && value_contents_eq (val, + embedded_offset + i * eltlen, + val, + (embedded_offset + + rep1 * eltlen), + eltlen)) { ++reps; ++rep1; diff --git a/gdb/value.c b/gdb/value.c index 8b44a6d..01e9409 100644 --- a/gdb/value.c +++ b/gdb/value.c @@ -197,15 +197,6 @@ struct value reset, be sure to consider this use as well! */ unsigned int lazy : 1; - /* If nonzero, this is the value of a variable that does not - actually exist in the program. If nonzero, and LVAL is - lval_register, this is a register ($pc, $sp, etc., never a - program variable) that has not been saved in the frame. All - optimized-out values are treated pretty much the same, except - registers have a different string representation and related - error strings. */ - unsigned int optimized_out : 1; - /* If value is a variable, is it initialized or not. */ unsigned int initialized : 1; @@ -336,9 +327,20 @@ struct value /* Unavailable ranges in CONTENTS. We mark unavailable ranges, rather than available, since the common and default case is for a - value to be available. This is filled in at value read time. The - unavailable ranges are tracked in bits. */ + value to be available. This is filled in at value read time. + The unavailable ranges are tracked in bits. Note that a contents + bit that has been optimized out doesn't really exist in the + program, so it can't be marked unavailable either. */ VEC(range_s) *unavailable; + + /* Likewise, but for optimized out contents (a chunk of the value of + a variable that does not actually exist in the program). If LVAL + is lval_register, this is a register ($pc, $sp, etc., never a + program variable) that has not been saved in the frame. Not + saved registers and optimized-out program variables values are + treated pretty much the same, except not-saved registers have a + different string representation and related error strings. */ + VEC(range_s) *optimized_out; }; int @@ -358,6 +360,14 @@ value_bytes_available (const struct value *value, int offset, int length) } int +value_bits_any_optimized_out (const struct value *value, int bit_offset, int bit_length) +{ + gdb_assert (!value->lazy); + + return ranges_contain (value->optimized_out, bit_offset, bit_length); +} + +int value_entirely_available (struct value *value) { /* We can only tell whether the whole value is available when we try @@ -370,17 +380,22 @@ value_entirely_available (struct value *value) return 0; } -int -value_entirely_unavailable (struct value *value) +/* Returns true if VALUE is entirely covered by RANGES. If the value + is lazy, it'll be read now. Note that RANGE is a pointer to + pointer because reading the value might change *RANGE. */ + +static int +value_entirely_covered_by_range_vector (struct value *value, + VEC(range_s) **ranges) { - /* We can only tell whether the whole value is available when we try - to read it. */ + /* We can only tell whether the whole value is optimized out / + unavailable when we try to read it. */ if (value->lazy) value_fetch_lazy (value); - if (VEC_length (range_s, value->unavailable) == 1) + if (VEC_length (range_s, *ranges) == 1) { - struct range *t = VEC_index (range_s, value->unavailable, 0); + struct range *t = VEC_index (range_s, *ranges, 0); if (t->offset == 0 && t->length == (TARGET_CHAR_BIT @@ -391,8 +406,23 @@ value_entirely_unavailable (struct value *value) return 0; } -void -mark_value_bits_unavailable (struct value *value, int offset, int length) +int +value_entirely_unavailable (struct value *value) +{ + return value_entirely_covered_by_range_vector (value, &value->unavailable); +} + +int +value_entirely_optimized_out (struct value *value) +{ + return value_entirely_covered_by_range_vector (value, &value->optimized_out); +} + +/* Insert into the vector pointed to by VECTORP the bit range starting of + OFFSET bits, and extending for the next LENGTH bits. */ + +static void +insert_into_bit_range_vector (VEC(range_s) **vectorp, int offset, int length) { range_s newr; int i; @@ -483,10 +513,10 @@ mark_value_bits_unavailable (struct value *value, int offset, int length) */ - i = VEC_lower_bound (range_s, value->unavailable, &newr, range_lessthan); + i = VEC_lower_bound (range_s, *vectorp, &newr, range_lessthan); if (i > 0) { - struct range *bef = VEC_index (range_s, value->unavailable, i - 1); + struct range *bef = VEC_index (range_s, *vectorp, i - 1); if (ranges_overlap (bef->offset, bef->length, offset, length)) { @@ -507,18 +537,18 @@ mark_value_bits_unavailable (struct value *value, int offset, int length) else { /* #3 */ - VEC_safe_insert (range_s, value->unavailable, i, &newr); + VEC_safe_insert (range_s, *vectorp, i, &newr); } } else { /* #4 */ - VEC_safe_insert (range_s, value->unavailable, i, &newr); + VEC_safe_insert (range_s, *vectorp, i, &newr); } /* Check whether the ranges following the one we've just added or touched can be folded in (#5 above). */ - if (i + 1 < VEC_length (range_s, value->unavailable)) + if (i + 1 < VEC_length (range_s, *vectorp)) { struct range *t; struct range *r; @@ -526,11 +556,11 @@ mark_value_bits_unavailable (struct value *value, int offset, int length) int next = i + 1; /* Get the range we just touched. */ - t = VEC_index (range_s, value->unavailable, i); + t = VEC_index (range_s, *vectorp, i); removed = 0; i = next; - for (; VEC_iterate (range_s, value->unavailable, i, r); i++) + for (; VEC_iterate (range_s, *vectorp, i, r); i++) if (r->offset <= t->offset + t->length) { ULONGEST l, h; @@ -552,11 +582,17 @@ mark_value_bits_unavailable (struct value *value, int offset, int length) } if (removed != 0) - VEC_block_remove (range_s, value->unavailable, next, removed); + VEC_block_remove (range_s, *vectorp, next, removed); } } void +mark_value_bits_unavailable (struct value *value, int offset, int length) +{ + insert_into_bit_range_vector (&value->unavailable, offset, length); +} + +void mark_value_bytes_unavailable (struct value *value, int offset, int length) { mark_value_bits_unavailable (value, @@ -684,48 +720,53 @@ memcmp_with_bit_offsets (const gdb_byte *ptr1, size_t offset1_bits, return 0; } -/* Helper function for value_available_contents_eq. The only difference is - that this function is bit rather than byte based. +/* Helper struct for find_first_range_overlap_and_match and + value_contents_bits_eq. Keep track of which slot of a given ranges + vector have we last looked at. */ - Compare LENGTH bits of VAL1's contents starting at OFFSET1 bits with - LENGTH bits of VAL2's contents starting at OFFSET2 bits. Return true - if the available bits match. */ +struct ranges_and_idx +{ + /* The ranges. */ + VEC(range_s) *ranges; + + /* The range we've last found in RANGES. Given ranges are sorted, + we can start the next lookup here. */ + int idx; +}; + +/* Helper function for value_contents_bits_eq. Compare LENGTH bits of + RP1's ranges starting at OFFSET1 bits with LENGTH bits of RP2's + ranges starting at OFFSET2 bits. Return true if the ranges match + and fill in *L and *H with the overlapping window relative to + (both) OFFSET1 or OFFSET2. */ static int -value_available_contents_bits_eq (const struct value *val1, int offset1, - const struct value *val2, int offset2, - int length) +find_first_range_overlap_and_match (struct ranges_and_idx *rp1, + struct ranges_and_idx *rp2, + int offset1, int offset2, + int length, ULONGEST *l, ULONGEST *h) { - int idx1 = 0, idx2 = 0; - - /* See function description in value.h. */ - gdb_assert (!val1->lazy && !val2->lazy); + rp1->idx = find_first_range_overlap (rp1->ranges, rp1->idx, + offset1, length); + rp2->idx = find_first_range_overlap (rp2->ranges, rp2->idx, + offset2, length); - while (length > 0) + if (rp1->idx == -1 && rp2->idx == -1) + { + *l = length; + *h = length; + return 1; + } + else if (rp1->idx == -1 || rp2->idx == -1) + return 0; + else { range_s *r1, *r2; ULONGEST l1, h1; ULONGEST l2, h2; - idx1 = find_first_range_overlap (val1->unavailable, idx1, - offset1, length); - idx2 = find_first_range_overlap (val2->unavailable, idx2, - offset2, length); - - /* The usual case is for both values to be completely available. */ - if (idx1 == -1 && idx2 == -1) - return (memcmp_with_bit_offsets (val1->contents, offset1, - val2->contents, offset2, - length) == 0); - /* The contents only match equal if the available set matches as - well. */ - else if (idx1 == -1 || idx2 == -1) - return 0; - - gdb_assert (idx1 != -1 && idx2 != -1); - - r1 = VEC_index (range_s, val1->unavailable, idx1); - r2 = VEC_index (range_s, val2->unavailable, idx2); + r1 = VEC_index (range_s, rp1->ranges, rp1->idx); + r2 = VEC_index (range_s, rp2->ranges, rp2->idx); /* Get the unavailable windows intersected by the incoming ranges. The first and last ranges that overlap the argument @@ -734,7 +775,7 @@ value_available_contents_bits_eq (const struct value *val1, int offset1, h1 = min (offset1 + length, r1->offset + r1->length); l2 = max (offset2, r2->offset); - h2 = min (offset2 + length, r2->offset + r2->length); + h2 = min (offset2 + length, offset2 + r2->length); /* Make them relative to the respective start offsets, so we can compare them for equality. */ @@ -744,31 +785,93 @@ value_available_contents_bits_eq (const struct value *val1, int offset1, l2 -= offset2; h2 -= offset2; - /* Different availability, no match. */ + /* Different ranges, no match. */ if (l1 != l2 || h1 != h2) return 0; - /* Compare the _available_ contents. */ + *h = h1; + *l = l1; + return 1; + } +} + +/* Helper function for value_contents_eq. The only difference is that + this function is bit rather than byte based. + + Compare LENGTH bits of VAL1's contents starting at OFFSET1 bits + with LENGTH bits of VAL2's contents starting at OFFSET2 bits. + Return true if the available bits match. */ + +static int +value_contents_bits_eq (const struct value *val1, int offset1, + const struct value *val2, int offset2, + int length) +{ + /* Each array element corresponds to a ranges source (unavailable, + optimized out). '1' is for VAL1, '2' for VAL2. */ + struct ranges_and_idx rp1[2], rp2[2]; + + /* See function description in value.h. */ + gdb_assert (!val1->lazy && !val2->lazy); + + /* We shouldn't be trying to compare past the end of the values. */ + gdb_assert (offset1 + length + <= TYPE_LENGTH (val1->enclosing_type) * TARGET_CHAR_BIT); + gdb_assert (offset2 + length + <= TYPE_LENGTH (val2->enclosing_type) * TARGET_CHAR_BIT); + + memset (&rp1, 0, sizeof (rp1)); + memset (&rp2, 0, sizeof (rp2)); + rp1[0].ranges = val1->unavailable; + rp2[0].ranges = val2->unavailable; + rp1[1].ranges = val1->optimized_out; + rp2[1].ranges = val2->optimized_out; + + while (length > 0) + { + ULONGEST l, h; + int i; + + for (i = 0; i < 2; i++) + { + ULONGEST l_tmp, h_tmp; + + /* The contents only match equal if the invalid/unavailable + contents ranges match as well. */ + if (!find_first_range_overlap_and_match (&rp1[i], &rp2[i], + offset1, offset2, length, + &l_tmp, &h_tmp)) + return 0; + + /* We're interested in the lowest/first range found. */ + if (i == 0 || l_tmp < l) + { + l = l_tmp; + h = h_tmp; + } + } + + /* Compare the available/valid contents. */ if (memcmp_with_bit_offsets (val1->contents, offset1, - val2->contents, offset2, l1) != 0) + val2->contents, offset2, l) != 0) return 0; - length -= h1; - offset1 += h1; - offset2 += h1; + length -= h; + offset1 += h; + offset2 += h; } return 1; } int -value_available_contents_eq (const struct value *val1, int offset1, - const struct value *val2, int offset2, - int length) +value_contents_eq (const struct value *val1, int offset1, + const struct value *val2, int offset2, + int length) { - return value_available_contents_bits_eq (val1, offset1 * TARGET_CHAR_BIT, - val2, offset2 * TARGET_CHAR_BIT, - length * TARGET_CHAR_BIT); + return value_contents_bits_eq (val1, offset1 * TARGET_CHAR_BIT, + val2, offset2 * TARGET_CHAR_BIT, + length * TARGET_CHAR_BIT); } /* Prototypes for local functions. */ @@ -836,7 +939,6 @@ allocate_value_lazy (struct type *type) val->bitsize = 0; VALUE_REGNUM (val) = -1; val->lazy = 1; - val->optimized_out = 0; val->embedded_offset = 0; val->pointed_to_offset = 0; val->modifiable = 1; @@ -905,11 +1007,8 @@ allocate_optimized_out_value (struct type *type) { struct value *retval = allocate_value_lazy (type); - set_value_optimized_out (retval, 1); - /* FIXME: we should be able to avoid allocating the value's contents - buffer, but value_available_contents_bits_eq can't handle - that. */ - /* set_value_lazy (retval, 0); */ + mark_value_bytes_optimized_out (retval, 0, TYPE_LENGTH (type)); + set_value_lazy (retval, 0); return retval; } @@ -1057,7 +1156,7 @@ error_value_optimized_out (void) static void require_not_optimized_out (const struct value *value) { - if (value->optimized_out) + if (!VEC_empty (range_s, value->optimized_out)) { if (value->lval == lval_register) error (_("register has not been saved in frame")); @@ -1097,6 +1196,31 @@ value_contents_all (struct value *value) return result; } +/* Copy ranges in SRC_RANGE that overlap [SRC_BIT_OFFSET, + SRC_BIT_OFFSET+BIT_LENGTH) ranges into *DST_RANGE, adjusted. */ + +static void +ranges_copy_adjusted (VEC (range_s) **dst_range, int dst_bit_offset, + VEC (range_s) *src_range, int src_bit_offset, + int bit_length) +{ + range_s *r; + int i; + + for (i = 0; VEC_iterate (range_s, src_range, i, r); i++) + { + ULONGEST h, l; + + l = max (r->offset, src_bit_offset); + h = min (r->offset + r->length, src_bit_offset + bit_length); + + if (l < h) + insert_into_bit_range_vector (dst_range, + dst_bit_offset + (l - src_bit_offset), + h - l); + } +} + /* Copy LENGTH bytes of SRC value's (all) contents (value_contents_all) starting at SRC_OFFSET, into DST value's (all) contents, starting at DST_OFFSET. If unavailable contents are @@ -1125,6 +1249,9 @@ value_contents_copy_raw (struct value *dst, int dst_offset, replaced. Make sure to remember to implement replacing if it turns out actually necessary. */ gdb_assert (value_bytes_available (dst, dst_offset, length)); + gdb_assert (!value_bits_any_optimized_out (dst, + TARGET_CHAR_BIT * dst_offset, + TARGET_CHAR_BIT * length)); /* Copy the data. */ memcpy (value_contents_all_raw (dst) + dst_offset, @@ -1135,18 +1262,14 @@ value_contents_copy_raw (struct value *dst, int dst_offset, src_bit_offset = src_offset * TARGET_CHAR_BIT; dst_bit_offset = dst_offset * TARGET_CHAR_BIT; bit_length = length * TARGET_CHAR_BIT; - for (i = 0; VEC_iterate (range_s, src->unavailable, i, r); i++) - { - ULONGEST h, l; - l = max (r->offset, src_bit_offset); - h = min (r->offset + r->length, src_bit_offset + bit_length); + ranges_copy_adjusted (&dst->unavailable, dst_bit_offset, + src->unavailable, src_bit_offset, + bit_length); - if (l < h) - mark_value_bits_unavailable (dst, - dst_bit_offset + (l - src_bit_offset), - h - l); - } + ranges_copy_adjusted (&dst->optimized_out, dst_bit_offset, + src->optimized_out, src_bit_offset, + bit_length); } /* Copy LENGTH bytes of SRC value's (all) contents @@ -1154,8 +1277,7 @@ value_contents_copy_raw (struct value *dst, int dst_offset, (all) contents, starting at DST_OFFSET. If unavailable contents are being copied from SRC, the corresponding DST contents are marked unavailable accordingly. DST must not be lazy. If SRC is - lazy, it will be fetched now. If SRC is not valid (is optimized - out), an error is thrown. + lazy, it will be fetched now. It is assumed the contents of DST in the [DST_OFFSET, DST_OFFSET+LENGTH) range are wholly available. */ @@ -1164,8 +1286,6 @@ void value_contents_copy (struct value *dst, int dst_offset, struct value *src, int src_offset, int length) { - require_not_optimized_out (src); - if (src->lazy) value_fetch_lazy (src); @@ -1218,45 +1338,29 @@ value_optimized_out (struct value *value) { /* We can only know if a value is optimized out once we have tried to fetch it. */ - if (!value->optimized_out && value->lazy) + if (VEC_empty (range_s, value->optimized_out) && value->lazy) value_fetch_lazy (value); - return value->optimized_out; + return !VEC_empty (range_s, value->optimized_out); } -int -value_optimized_out_const (const struct value *value) -{ - return value->optimized_out; -} +/* Mark contents of VALUE as optimized out, starting at OFFSET bytes, and + the following LENGTH bytes. */ void -set_value_optimized_out (struct value *value, int val) +mark_value_bytes_optimized_out (struct value *value, int offset, int length) { - value->optimized_out = val; + mark_value_bits_optimized_out (value, + offset * TARGET_CHAR_BIT, + length * TARGET_CHAR_BIT); } -int -value_entirely_optimized_out (const struct value *value) -{ - if (!value->optimized_out) - return 0; - if (value->lval != lval_computed - || !value->location.computed.funcs->check_any_valid) - return 1; - return !value->location.computed.funcs->check_any_valid (value); -} +/* See value.h. */ -int -value_bits_valid (const struct value *value, int offset, int length) +void +mark_value_bits_optimized_out (struct value *value, int offset, int length) { - if (!value->optimized_out) - return 1; - if (value->lval != lval_computed - || !value->location.computed.funcs->check_validity) - return 0; - return value->location.computed.funcs->check_validity (value, offset, - length); + insert_into_bit_range_vector (&value->optimized_out, offset, length); } int @@ -1569,7 +1673,6 @@ value_copy (struct value *arg) VALUE_FRAME_ID (val) = VALUE_FRAME_ID (arg); VALUE_REGNUM (val) = VALUE_REGNUM (arg); val->lazy = arg->lazy; - val->optimized_out = arg->optimized_out; val->embedded_offset = value_embedded_offset (arg); val->pointed_to_offset = arg->pointed_to_offset; val->modifiable = arg->modifiable; @@ -1580,6 +1683,7 @@ value_copy (struct value *arg) } val->unavailable = VEC_copy (range_s, arg->unavailable); + val->optimized_out = VEC_copy (range_s, arg->optimized_out); set_value_parent (val, arg->parent); if (VALUE_LVAL (val) == lval_computed) { @@ -2854,24 +2958,19 @@ value_primitive_field (struct value *arg1, int offset, int bitpos = TYPE_FIELD_BITPOS (arg_type, fieldno); int container_bitsize = TYPE_LENGTH (type) * 8; - if (arg1->optimized_out) - v = allocate_optimized_out_value (type); + v = allocate_value_lazy (type); + v->bitsize = TYPE_FIELD_BITSIZE (arg_type, fieldno); + if ((bitpos % container_bitsize) + v->bitsize <= container_bitsize + && TYPE_LENGTH (type) <= (int) sizeof (LONGEST)) + v->bitpos = bitpos % container_bitsize; else - { - v = allocate_value_lazy (type); - v->bitsize = TYPE_FIELD_BITSIZE (arg_type, fieldno); - if ((bitpos % container_bitsize) + v->bitsize <= container_bitsize - && TYPE_LENGTH (type) <= (int) sizeof (LONGEST)) - v->bitpos = bitpos % container_bitsize; - else - v->bitpos = bitpos % 8; - v->offset = (value_embedded_offset (arg1) - + offset - + (bitpos - v->bitpos) / 8); - set_value_parent (v, arg1); - if (!value_lazy (arg1)) - value_fetch_lazy (v); - } + v->bitpos = bitpos % 8; + v->offset = (value_embedded_offset (arg1) + + offset + + (bitpos - v->bitpos) / 8); + set_value_parent (v, arg1); + if (!value_lazy (arg1)) + value_fetch_lazy (v); } else if (fieldno < TYPE_N_BASECLASSES (arg_type)) { @@ -2884,37 +2983,29 @@ value_primitive_field (struct value *arg1, int offset, if (VALUE_LVAL (arg1) == lval_register && value_lazy (arg1)) value_fetch_lazy (arg1); - /* The optimized_out flag is only set correctly once a lazy value is - loaded, having just loaded some lazy values we should check the - optimized out case now. */ - if (arg1->optimized_out) - v = allocate_optimized_out_value (type); + /* We special case virtual inheritance here because this + requires access to the contents, which we would rather avoid + for references to ordinary fields of unavailable values. */ + if (BASETYPE_VIA_VIRTUAL (arg_type, fieldno)) + boffset = baseclass_offset (arg_type, fieldno, + value_contents (arg1), + value_embedded_offset (arg1), + value_address (arg1), + arg1); else - { - /* We special case virtual inheritance here because this - requires access to the contents, which we would rather avoid - for references to ordinary fields of unavailable values. */ - if (BASETYPE_VIA_VIRTUAL (arg_type, fieldno)) - boffset = baseclass_offset (arg_type, fieldno, - value_contents (arg1), - value_embedded_offset (arg1), - value_address (arg1), - arg1); - else - boffset = TYPE_FIELD_BITPOS (arg_type, fieldno) / 8; + boffset = TYPE_FIELD_BITPOS (arg_type, fieldno) / 8; - if (value_lazy (arg1)) - v = allocate_value_lazy (value_enclosing_type (arg1)); - else - { - v = allocate_value (value_enclosing_type (arg1)); - value_contents_copy_raw (v, 0, arg1, 0, - TYPE_LENGTH (value_enclosing_type (arg1))); - } - v->type = type; - v->offset = value_offset (arg1); - v->embedded_offset = offset + value_embedded_offset (arg1) + boffset; + if (value_lazy (arg1)) + v = allocate_value_lazy (value_enclosing_type (arg1)); + else + { + v = allocate_value (value_enclosing_type (arg1)); + value_contents_copy_raw (v, 0, arg1, 0, + TYPE_LENGTH (value_enclosing_type (arg1))); } + v->type = type; + v->offset = value_offset (arg1); + v->embedded_offset = offset + value_embedded_offset (arg1) + boffset; } else { @@ -2925,12 +3016,7 @@ value_primitive_field (struct value *arg1, int offset, if (VALUE_LVAL (arg1) == lval_register && value_lazy (arg1)) value_fetch_lazy (arg1); - /* The optimized_out flag is only set correctly once a lazy value is - loaded, having just loaded some lazy values we should check for - the optimized out case now. */ - if (arg1->optimized_out) - v = allocate_optimized_out_value (type); - else if (value_lazy (arg1)) + if (value_lazy (arg1)) v = allocate_value_lazy (type); else { @@ -3649,6 +3735,11 @@ value_fetch_lazy (struct value *val) { gdb_assert (value_lazy (val)); allocate_value_contents (val); + /* A value is either lazy, or fully fetched. The + availability/validity is only established as we try to fetch a + value. */ + gdb_assert (VEC_empty (range_s, val->optimized_out)); + gdb_assert (VEC_empty (range_s, val->unavailable)); if (value_bitsize (val)) { /* To read a lazy bitfield, read the entire enclosing value. This @@ -3665,10 +3756,11 @@ value_fetch_lazy (struct value *val) if (value_lazy (parent)) value_fetch_lazy (parent); - if (!value_bits_valid (parent, - TARGET_CHAR_BIT * offset + value_bitpos (val), - value_bitsize (val))) - set_value_optimized_out (val, 1); + if (value_bits_any_optimized_out (parent, + TARGET_CHAR_BIT * offset + value_bitpos (val), + value_bitsize (val))) + mark_value_bytes_optimized_out (val, value_embedded_offset (val), + TYPE_LENGTH (type)); else if (!unpack_value_bits_as_long (value_type (val), value_contents_for_printing (parent), offset, @@ -3743,16 +3835,12 @@ value_fetch_lazy (struct value *val) if (value_lazy (new_val)) value_fetch_lazy (new_val); - /* If the register was not saved, mark it optimized out. */ - if (value_optimized_out (new_val)) - set_value_optimized_out (val, 1); - else - { - set_value_lazy (val, 0); - value_contents_copy (val, value_embedded_offset (val), - new_val, value_embedded_offset (new_val), - TYPE_LENGTH (type)); - } + /* Copy the contents and the unavailability/optimized-out + meta-data from NEW_VAL to VAL. */ + set_value_lazy (val, 0); + value_contents_copy (val, value_embedded_offset (val), + new_val, value_embedded_offset (new_val), + TYPE_LENGTH (type)); if (frame_debug) { @@ -3805,11 +3893,6 @@ value_fetch_lazy (struct value *val) else if (VALUE_LVAL (val) == lval_computed && value_computed_funcs (val)->read != NULL) value_computed_funcs (val)->read (val); - /* Don't call value_optimized_out on val, doing so would result in a - recursive call back to value_fetch_lazy, instead check the - optimized_out flag directly. */ - else if (val->optimized_out) - /* Keep it optimized out. */; else internal_error (__FILE__, __LINE__, _("Unexpected lazy value type.")); diff --git a/gdb/value.h b/gdb/value.h index 86ebd70..bff5220 100644 --- a/gdb/value.h +++ b/gdb/value.h @@ -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 + . 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 + instead of , as in: + + (gdb) p/x $rax + $1 = + (gdb) info registers rax + rax + + If the debug info describes a variable as being in such a register, + we'll still print the variable as . IOW, + 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. */ + +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. */ -/* 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 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. @@ -416,13 +469,6 @@ 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. */ extern int value_bits_synthetic_pointer (const struct value *value, @@ -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/valid contents match. + + Optimized-out contents are equal to optimized-out contents, and are + not equal to non-optimized-out contents. - 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: + Unavailable contente are equal to unavailable contents, and are not + equal to non-unavailable contents. - offset: 0 4 8 12 16 - contents: xxxxVVVVxxxxVVZZ + 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 - - 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. */ - -extern int value_available_contents_eq (const struct value *val1, int offset1, - const struct value *val2, int offset2, - int length); + 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 + + If 'x's represent an unavailable byte, 'o' represents an optimized + out byte, in a value with length 8: + + 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