Jim Wilson [Sat, 17 Nov 2018 23:31:01 +0000 (23:31 +0000)]
RISC-V: Fix epilogue unwind info with fp and single sp adjust.
gcc/
* config/riscv/riscv.c (epilogue_cfa_sp_offset): New.
(riscv_restore_reg): If restoring HARD_FRAME_POINTER_REGNUM, and
epilogue_cfa_sp_offset set, then add REG_CFA_DEF_CFA regnote.
(riscv_expand_epilogue): Initialize epilogue_cfa_sp_offset. Set it
to step2 if frame_pointer_needed and step1 is 0.
Jakub Jelinek [Sat, 17 Nov 2018 15:10:48 +0000 (16:10 +0100)]
re PR tree-optimization/87546 (Gcc miscompiles at -O3 on valid code)
PR tree-optimization/87546
* tree-vect-patterns.c (vect_look_through_possible_promotion): Add
min_precision variable, initially set it to orig_precision, only does
something if op_type's precision is <= min_precision and update
min_precision whenever calling set_op.
Jakub Jelinek [Sat, 17 Nov 2018 15:09:43 +0000 (16:09 +0100)]
re PR rtl-optimization/88018 (ICE in insert_insn_on_edge at cfgrtl.c:1952 since r255066)
PR rtl-optimization/88018
* cfgrtl.c (fixup_abnormal_edges): Guard moving insns to fallthru edge
on the presence of fallthru edge, rather than if it is a USE or not.
Jan Hubicka [Sat, 17 Nov 2018 11:35:01 +0000 (12:35 +0100)]
re PR ipa/87957 (ICE tree check: expected tree that contains ‘decl minimal’ structure, have ‘identifier_node’ in warn_odr, at ipa-devirt.c:1051 since r265519)
PR ipa/87957
* ipa-devirt.c (warn_odr): Look for main variant to get TYPE_DECL.
re PR target/87927 (ICE: segmentation fault with patchable_function_entry attribute for msp430-elf -mlarge)
2018-11-16 Jozef Lawrynowicz <jozef.l@mittosystems.com>
PR target/87927
* target-def.h: Initialize TARGET_ASM_{,UN}ALIGNED_P{S,D,T}I_OP.
Add them to the TARGET_ASM_{,UN}ALIGNED_INT_OP structs.
* target.def: Enumerate TARGET_ASM_{,UN}ALIGNED_P{S,D,T}I_OP in
the byte_op hook.
* target.h: Add psi, pdi, pti to struct asm_int_op definition.
* targhooks.c (default_print_patchable_function_entry): Assert
asm_int_op does not return a NULL string.
* varasm.c (integer_asm_op): Return the op for a partial int type
when the requested size does not correspond to an integer type.
* config/msp430/msp430.c: Initialize TARGET_ASM_{,UN}ALIGNED_PSI_OP.
* doc/tm.texi: Regenerate.
Jason Merrill [Fri, 16 Nov 2018 21:49:42 +0000 (16:49 -0500)]
Implement P0479R5, [[likely]] and [[unlikely]].
[[likely]] and [[unlikely]] are equivalent to the GNU hot/cold attributes,
except that they can be applied to arbitrary statements as well as labels;
this is most likely to be useful for marking if/else branches as likely or
unlikely. Conveniently, PREDICT_EXPR fits the bill nicely as a
representation.
I also had to fix marking case labels as hot/cold, which didn't work before.
Which then required me to force __attribute ((fallthrough)) to apply to the
statement rather than the label.
Jakub Jelinek [Fri, 16 Nov 2018 16:43:49 +0000 (17:43 +0100)]
re PR rtl-optimization/87475 (ICE in patch_jump_insn, at cfgrtl.c:1275)
PR rtl-optimization/87475
* cfgrtl.c (patch_jump_insn): Allow redirection failure for
CROSSING_JUMP_P insns.
(cfg_layout_redirect_edge_and_branch): Don't ICE if ret is NULL.
Jakub Jelinek [Fri, 16 Nov 2018 16:40:53 +0000 (17:40 +0100)]
re PR middle-end/88032 (ICE in operand_subword_force, at emit-rtl.c:1793)
PR middle-end/88032
* optabs.c (expand_binop): For op0_mode use GET_MODE (op0), unless it
is VOIDmode, in which case use int_mode. Similarly for op1_mode.
When running the testsuite on boards that can't report an error status DejaGNU...
When running the testsuite on boards that can't report an error status
DejaGNU uses a special wrapper to print the exit code on stdout and
parses stdout to find whether an execution failed or passed.
In testcases that use "freopen (..., ..., stdout)" this special line is
printed to the alternate location described in the freopen call and
DejaGNU can't find the error code.
This results in DejaGNU using a default return status of 2 and the test
failing.
This patch skips the two testcases that use freopen on stdout when
testing a board that requires this wrapper.
Testing done by running these two tests on arm-none-eabi cross build and
observing that they are unsupported.
gcc/testsuite/ChangeLog:
2018-11-16 Matthew Malcomson <matthew.malcomson@arm.com>
* gcc.c-torture/execute/printf-2.c: Skip on wrapped boards.
* gcc.c-torture/execute/user-printf.c: Likewise.
* tree-switch-conversion.h (switch_decision_tree::emit_case_nodes): Add
location_t parameter.
(switch_decision_tree::emit_cmp_and_jump_insns): Likewise.
(switch_decision_tree::do_jump_if_equal): Likewise.
* tree-switch-conversion.c (switch_decision_tree::emit): Pass location
of switch statement to emit_case_nodes.
(switch_decision_tree::emit_cmp_and_jump_insns): Add LOC parameter and
set it on the newly built GIMPLE comparison statement.
(switch_decision_tree::do_jump_if_equal): Likewise.
(switch_decision_tree::emit_case_nodes): Add LOC parameter and pass it
in calls to do_jump_if_equal as well as recursive calls.
Andi Kleen [Fri, 16 Nov 2018 15:46:31 +0000 (15:46 +0000)]
Allow memory operands for PTWRITE
The earlier PTWRITE builtin definition was unnecessarily restrictive,
only allowing register input to PTWRITE. The instruction actually
supports memory operands too, so allow that too.
gcc/:
2018-11-16 Andi Kleen <ak@linux.intel.com>
* config/i386/i386.md: Allow memory operands to ptwrite.
Nathan Sidwell [Fri, 16 Nov 2018 15:01:55 +0000 (15:01 +0000)]
[PR c++/87269] Mark string operator overload in template defn.
https://gcc.gnu.org/ml/gcc-patches/2018-11/msg01458.html
PR c++/87269
* parser.c (lookup_literal_operator): Mark overload for keeping
when inside template. Refactor.
Martin Liska [Fri, 16 Nov 2018 12:49:32 +0000 (13:49 +0100)]
Fix ICE in lto_symtab_merge_symbols_1 (PR lto/88004).
2018-11-16 Martin Liska <mliska@suse.cz>
PR lto/88004
* lto-symtab.c (lto_symtab_merge_symbols_1): Do not call
lto_symtab_symbol_p as it does checking of transparent alias.
These needs to be also merged in the function.
Richard Biener [Fri, 16 Nov 2018 12:20:05 +0000 (12:20 +0000)]
re PR tree-optimization/88011 (r266028 causes a bunch of go failures)
2018-11-16 Richard Biener <rguenther@suse.de>
PR tree-optimization/88011
* tree-vrp.c (extract_range_from_binary_expr): Fix error in
replacing set_value_range_to_undefined and
set_value_range_to_varying with method calls.
with an extra subreg, which appears because pseudos, unlike hard
registers, can be accessed only using their natural mode.
This patch adds a special case for that. Also, it performs r*sbg
bit index computations during gcc run, so that expectations do not
depend on which concrete pattern was matched.
gcc/ChangeLog:
2018-11-16 Ilya Leoshkevich <iii@linux.ibm.com>
* config/s390/s390.md
(*r<noxa>sbg_<mode>_srl_bitmask): Do not delegate arithmetic to
assembler.
(*r<noxa>sbg_<mode>_sll): Likewise.
(*r<noxa>sbg_<mode>_srl): Likewise.
(*r<noxa>sbg_sidi_srl): New pattern.
gcc/testsuite/ChangeLog:
2018-11-16 Ilya Leoshkevich <iii@linux.ibm.com>
* gcc.target/s390/md/rXsbg_mode_sXl.c: Do not use arithmetic in
r{o,x}sbg expectations.
* gcc.target/s390/risbg-ll-2.c: Likewise.
Jerome Lambourg [Fri, 16 Nov 2018 09:23:22 +0000 (09:23 +0000)]
Allow target overrides of ctors/dtors support advertisement on VxWorks
2018-11-16 Jerome Lambourg <lambourg@adacore.com>
* config/vxworks.h (TARGET_VXWORKS_HAVE_CTORS_DTORS): New macro.
Default to TARGET_VXWORKS_RTP.
(SUPPORTS_INIT_PRIORITY): Use TARGET_VXWORKS_HAVE_CTORS_DTORS instead
of TARGET_VXWORKS_RTP.
* config/vxworksae.h: Also define TARGET_VXWORKS_HAVE_CTORS_DTORS.
* config/vxworks.c: Use TARGET_VXWORKS_HAVE_CTORS_DTORS instead
of TARGET_VXWORKS_RTP to set targetm.have_ctors_dtors.
Martin Sebor [Thu, 15 Nov 2018 22:53:57 +0000 (22:53 +0000)]
PR c++/87541 - ICE using a constant decl as an attribute alloc_size argument
PR c++/87541 - ICE using a constant decl as an attribute alloc_size argument
PR c++/87542 - bogus error on attribute format with a named constant argument
gcc/ChangeLog:
PR c++/87541
PR c++/87542
* tree.c (type_argument_type): New function.
* tree.h (type_argument_type): Declare it.
* gcc/doc/extend.texi (alloc_align): Update and clarify.
(alloc_size, nonnull, sentinel): Same.
gcc/c-family/ChangeLog:
PR c++/87541
PR c++/87542
* c-attribs.c (positional_argument): New function.
(handle_alloc_size_attribute): Use it and simplify.
(handle_alloc_align_attribute): Same.
(handle_assume_aligned_attribute): Same.
(handle_nonnull_attribute): Same.
* c-common.c (check_function_arguments): Pass fntype to
check_function_format.
* c-common.h (check_function_format): Add an argument.
(PosArgFlags, positional_argument): Declare new type and function.
* c-format.c (decode_format_attr): Add arguments.
(check_format_string, get_constant): Same.
(convert_format_name_to_system_name): Adjust.
Martin Sebor [Thu, 15 Nov 2018 22:18:54 +0000 (22:18 +0000)]
PR c/83656 - missing -Wbuiltin-declaration-mismatch on declaration without prototype
gcc/c/ChangeLog:
PR c/83656
* c-decl.c (header_for_builtin_fn): Declare.
(diagnose_mismatched_decls): Diagnose declarations of built-in
functions without a prototype.
* c-typeck.c (maybe_warn_builtin_no_proto_arg): New function.
(convert_argument): Same.
(convert_arguments): Factor code out into convert_argument.
Detect mismatches between built-in formal arguments in calls
to built-in without prototype.
(build_conditional_expr): Same.
(type_or_builtin_type): New function.
(convert_for_assignment): Add argument. Conditionally issue
warnings instead of errors for mismatches.
Nathan Sidwell [Thu, 15 Nov 2018 19:54:25 +0000 (19:54 +0000)]
[PR c++/86246] ICE tsubst explicit operator call
https://gcc.gnu.org/ml/gcc-patches/2018-11/msg01405.html
PR c++/86246
PR c++/87989
* typeck.c (finish_class_member_access_expr): Conversion operator
to dependent type is dependent.
Andrew Stubbs [Thu, 15 Nov 2018 17:14:15 +0000 (17:14 +0000)]
Handle vectors that don't fit in an integer.
GCN vector sizes range between 64 and 512 bytes, none of which have
correspondingly sized integer modes. This breaks a number of assumptions
throughout the compiler, but I don't really want to create modes just for this
purpose.
Instead, this patch fixes up the cases that I've found, so far, such that the
compiler tries something else, or fails to optimize, rather than just ICE.
2018-11-15 Andrew Stubbs <ams@codesourcery.com>
Kwok Cheung Yeung <kcy@codesourcery.com>
Nathan Sidwell [Thu, 15 Nov 2018 15:51:39 +0000 (15:51 +0000)]
[C++ DR 2336] virtual dtors, exception specs & abstract classes
https://gcc.gnu.org/ml/gcc-patches/2018-11/msg01389.html
DR 2336
* cp-tree.h (enum special_function_kind): Add sfk_virtual_destructor.
* method.c (type_has_trivial_fn): Add it.
(SFK_DTOR_P): Likewise.
(synthesized_method_base_walk): Don't check access of vbases of
abstract classes when sfk_virtual_destructor.
(synthesized_method_walk): Skip vbases of abstract classes except
when sfk_virtual_destructor.
(get_defaulted_eh_spec): Set sfk_virtual_destructor as needed.
David Malcolm [Thu, 15 Nov 2018 15:18:43 +0000 (15:18 +0000)]
doc/ux.texi: auto_diagnostic_group now does something
As of r266186, auto_diagnostic_group is now user-visible
(via -fdiagnostics-format=json), so update the diagnostic
guidelines accordingly.
gcc/ChangeLog:
* doc/ux.texi (Group logically-related diagnostics): Move
discussion of auto_diagnostic_group into this new subsection.
Give an example of where this grouping is used.
This patch implements a -fdiagnostics-format=json option which
converts the diagnostics to be output to stderr in a JSON format;
see the documentation in invoke.texi.
Logically-related diagnostics are nested at the JSON level, using
the auto_diagnostic_group mechanism.
gcc/ChangeLog:
PR other/19165
* Makefile.in (OBJS): Move json.o to...
(OBJS-libcommon): ...here and add diagnostic-format-json.o.
* common.opt (fdiagnostics-format=): New option.
(diagnostics_output_format): New enum.
* diagnostic-format-json.cc: New file.
* diagnostic.c (default_diagnostic_final_cb): New function, taken
from start of diagnostic_finish.
(diagnostic_initialize): Initialize final_cb to
default_diagnostic_final_cb.
(diagnostic_finish): Move "being treated as errors" messages to
default_diagnostic_final_cb. Call any final_cb.
(default_diagnostic_finalizer): Add diagnostic_t param.
(diagnostic_report_diagnostic): Pass "orig_diag_kind" to
diagnostic_finalizer callback.
* diagnostic.h (enum diagnostics_output_format): New enum.
(diagnostic_finalizer_fn): Reimplement, adding diagnostic_t param.
(struct diagnostic_context): Add "final_cb".
(default_diagnostic_finalizer): Add diagnostic_t param.
(diagnostic_output_format_init): New decl.
* doc/invoke.texi (-fdiagnostics-format): New option.
* dwarf2out.c (gen_producer_string): Ignore
OPT_fdiagnostics_format_.
* gcc.c (driver_handle_option): Handle OPT_fdiagnostics_format_.
* lto-wrapper.c (append_diag_options): Ignore it.
* opts.c (common_handle_option): Handle it.
Richard Biener [Thu, 15 Nov 2018 13:42:13 +0000 (13:42 +0000)]
re PR tree-optimization/88031 (ICE in vectorizable_reduction, at tree-vect-loop.c:6953)
2018-11-15 Richard Biener <rguenther@suse.de>
PR tree-optimization/88031
* tree-vect-loop.c (vectorizable_reduction): Move check
for multiple types earlier so we get the expected dump.
Simplify calls to vectorizable_condition.
* tree-vect-stmts.h (vectorizable_condition): Update prototype.
* tree-vect-stmts.c (vectorizable_condition): Instead of
reduc_def and reduc_index take just a flag. Simplify
code-generation now that we can rely on the defs being set up.
(vectorizable_comparison): Remove unused argument.
Nathan Sidwell [Thu, 15 Nov 2018 13:22:15 +0000 (13:22 +0000)]
[C++ DR 2336] Clean up synth walkers first
https://gcc.gnu.org/ml/gcc-patches/2018-11/msg01376.html
* cp-tree.h (enum special_function_kind): Reorder and comment.
* method.c (SFK_CTOR_P, SFK_DTOR_P, SFK_ASSIGN_P, SFK_COPY_P)
(SFK_MOVE_P): New predicates.
(walk_field_subobs, synthesized_method_base_walk): Drop
copy_arg_p, move_p, assign_p args. Use new SFK predicates. Order
parameters consistently.
(synthesized_method_walk): Drop ctor_p, copy_arg_p, move_p,
assign_p calculations. Use new SFK predicates. Adjust calls to
worker functions.
Richard Biener [Thu, 15 Nov 2018 10:42:15 +0000 (10:42 +0000)]
re PR tree-optimization/88030 (ICE in calc_dfs_tree, at dominance.c:458)
2018-11-15 Richard Biener <rguenther@suse.de>
PR tree-optimization/88030
* tree-complex.c (need_eh_cleanup): New global.
(update_complex_assignment): Mark blocks that need EH update.
(expand_complex_comparison): Likewise.
(tree_lower_complex): Allocate and deallocate need_eh_cleanup,
perform EH cleanup and schedule CFG cleanup if that did anything.
Jakub Jelinek [Thu, 15 Nov 2018 08:47:21 +0000 (09:47 +0100)]
re PR rtl-optimization/88018 (ICE in insert_insn_on_edge at cfgrtl.c:1952 since r255066)
PR rtl-optimization/88018
* cfgrtl.c (fixup_abnormal_edges): Guard moving insns to fallthru edge
on the presence of fallthru edge, rather than if it is a USE or not.
gcc/
* doc/invoke.texi (Optimize Options): Remove bad example about
interaction between -flto and -O. Replace it with a note that
you need to compile with -O and not just link. Copy-edit -flto
discussion to reduce verbiage and improve flow.
Jonathan Wakely [Thu, 15 Nov 2018 00:04:19 +0000 (00:04 +0000)]
Optimize pool resource allocation
A recent change caused a performance regression. This restores the
previous performance and adds a performance test.
* scripts/check_performance: Allow tests to choose a -std flag.
* src/c++17/memory_resource.cc (bitset::get_first_unset()): Use local
variables of the right types. Call update_next_word() unconditionally.
* testsuite/20_util/unsynchronized_pool_resource/cons.cc: New test.
* testsuite/performance/20_util/memory_resource/pools.cc: New test.
* testsuite/util/testsuite_performance.h (time_counter): Allow
timer to be restarted.
gcc/
* doc/invoke.texi (Optimize Options): Clarify that -O0 and -Og
also suppress many optimizations. Alphabetize option lists for
-O1, -O2, and -Os. Add list of options disabled with -Og, and
correct documentation for those options to say that.
* opts.c (default_options_table): Sort table by level and option
name, to make it easier to correlate to the manual.
Jakub Jelinek [Wed, 14 Nov 2018 16:46:06 +0000 (17:46 +0100)]
P1236R1 - Signed integers are two's complement
P1236R1 - Signed integers are two's complement
gcc/cp/
* constexpr.c (cxx_eval_check_shift_p): Disable the signed LSHIFT_EXPR
checks for c++2a.
gcc/c-family/
* c-warn.c (maybe_warn_shift_overflow): Don't warn for c++2a.
* c-ubsan.c (ubsan_instrument_shift): Make signed shifts
with in-range second operand well defined for -std=c++2a.
gcc/
* doc/invoke.texi (Wshift-overflow): Adjust documentation for
c++2a.
gcc/testsuite/
* g++.dg/cpp2a/constexpr-shift1.C: New test.
* g++.dg/warn/permissive-1.C (enum A, enum D): Don't expect
diagnostics here for c++2a.
* g++.dg/cpp0x/constexpr-shift1.C (fn3, i3, fn4, i4): Don't expect
diagnostics here for c++2a.
* g++.dg/cpp0x/constexpr-60049.C (f3, x3, y3): Likewise.
* g++.dg/ubsan/cxx11-shift-1.C (main): Add some further tests.
* g++.dg/ubsan/cxx11-shift-2.C (main): Likewise.
* g++.dg/ubsan/cxx2a-shift-1.C: New test.
* g++.dg/ubsan/cxx2a-shift-2.C: New test.
Jonathan Wakely [Wed, 14 Nov 2018 14:25:00 +0000 (14:25 +0000)]
Add missing dir to create_testsuite_files script
* scripts/create_testsuite_files: Add special_functions to the list
of directories to search. Add comment referring to conformance.exp.
* testsuite/libstdc++-dg/conformance.exp: Add comment referring
to create_testsuite_files.
Ilya Leoshkevich [Wed, 14 Nov 2018 13:07:13 +0000 (13:07 +0000)]
[PATCH 1/2] S/390: Disable 3 global-array-* tests for 31-bit mode
These tests rely on larl->movdi merge, which is not implemented for
31-bit mode.
gcc/testsuite/ChangeLog:
2018-11-14 Ilya Leoshkevich <iii@linux.ibm.com>
* gcc.target/s390/global-array-almost-huge-element.c: Run only
in 64-bit mode.
* gcc.target/s390/global-array-almost-negative-huge-element.c:
Likewise.
* gcc.target/s390/global-array-even-element.c: Likewise.
Wilco Dijkstra [Wed, 14 Nov 2018 12:45:29 +0000 (12:45 +0000)]
Simplify floating point comparisons
This patch implements some of the optimizations discussed in PR71026.
Simplify (C / x >= 0.0) into x >= 0.0 with -funsafe-math-optimizations
(since C / x can underflow to zero if x is huge, it's not safe otherwise).
If C is negative the comparison is reversed.
Simplify (x * C1) > C2 into x > (C2 / C1) with -funsafe-math-optimizations.
If C1 is negative the comparison is reversed.
gcc/
PR 71026/tree-optimization
* match.pd: Simplify floating point comparisons.
gcc/testsuite/
PR 71026/tree-optimization
* gcc.dg/div-cmp-1.c: New test.
* gcc.dg/div-cmp-2.c: New test.
Co-Authored-By: Jackson Woodruff <jackson.woodruff@arm.com>
From-SVN: r266142
Jakub Jelinek [Wed, 14 Nov 2018 12:38:20 +0000 (13:38 +0100)]
re PR other/88007 (ICE in gt_ggc_m_S, at ggc-page.c:1474)
PR other/88007
* c-common.c (parse_optimize_options): Allocate option string from
opts_obstack rather than as GC memory. Move the allocation after
warning for invalid option.
This is due to Combine merging a DUP instruction with either a load
or MLA - we can't force it to prefer one over the other. However the
generated vector loop is fast either way since it generates MLA and
merges the DUP either with a load or MLA. So relax the conditions
slightly and check we still generate MLA and there is no DUP or FMOV.
Ed Schonberg [Wed, 14 Nov 2018 11:42:49 +0000 (11:42 +0000)]
[Ada] Problem with boundary case of XOR operation and unnesting
The XOR operation applied to a boolean array whose component type has
the range True .. True raises constraint error. Previous to this patch,
the expansion of the operation could lead to uplevel references that
were not handled properly when unnesting is in effect.
2018-11-14 Ed Schonberg <schonberg@adacore.com>
gcc/ada/
* exp_util.ads, exp_util.adb: Change the profile of
Silly_Boolean_Array_Xor_Test, adding a formal that can be a copy
of the right opersnd. This prevents unnesting anomalies when
that operand contains uplevel references.
* exp_ch4.adb (Expand_Boolean_Operation): Use this new profile.
* exp_pakd.adb (Expand_Packed_Boolean_Operator): Ditto.
Patrick Bernardi [Wed, 14 Nov 2018 11:42:43 +0000 (11:42 +0000)]
[Ada] Update signal constants for GNU/Linux
Add the signal SIGSYS and mark the glibc reserved real-time signals
(32-34) as reserved rather than not maskable.
2018-11-14 Patrick Bernardi <bernardi@adacore.com>
gcc/ada/
* libgnarl/a-intnam__linux.ads: Add SIGSYS.
* libgnarl/s-linux__alpha.ads, libgnarl/s-linux__android.ads,
libgnarl/s-linux__hppa.ads, libgnarl/s-linux__mips.ads,
libgnarl/s-linux__riscv.ads, libgnarl/s-linux__sparc.ads,
libgnarl/s-linux__x32.ads: Rename SIGLTHRRES, SIGLTHRCAN and
SIGLTHRDBG to SIG32, SIG33 and SIG34 as their names are
implementation specific.
* libgnarl/s-osinte__linux.ads, libgnarl/s-linux.ads: Add
SIGSYS. Move SIG32, SIG33 and SIG34 from the unmasked list to
the reserved list.
Eric Botcazou [Wed, 14 Nov 2018 11:42:37 +0000 (11:42 +0000)]
[Ada] Fix small regression with others choice in array aggregate
This change is aimed at fixing a fallout of bumping the default value of
the Max_Others_Replicate parameter of the Convert_To_Positional routine.
This parameter is responsible for taming the duplication of the
expression of an others choice in an array aggregate so that it doesn't
result in a code size explosion.
Unfortunately a fine-grained control based on the analysis of the
expression is not really possible because this analysis has not been
done yet by the time the decision is made in most cases, so the usual
syntactic ambiguities of the language come into play and make the
process a bit cumbersome. For example, it is not possible to
distinguish a simple reference to a static constant declared in another
unit from a call to a parameterless function.
Therefore the change errs on the side of caution and allows the
duplication only if the expression is unambiguously static and
sufficiently simple.
For the following three aggregates, the duplication must be blocked and
the elaboration of the aggregates must be done by means of a loop:
with Q; use Q;
procedure P is
A : Arr := (others => Get_Value);
B : Arr := (others => Get_Other_Value (0));
C : Arr := (others => Q.Get_Other_Value (1));
begin
null;
end;
package Q is
type Arr is array (1 .. 32) of Integer;
function Get_Value return Integer;
function Get_Other_Value (I : integer) return Integer;
end Q;
2018-11-14 Eric Botcazou <ebotcazou@adacore.com>
gcc/ada/
* exp_aggr.adb (Is_Static_Element): New predicate extracted
from...
(Check_Static_Components): ...here. Call Is_Static_Element on
each element of the component association list, if any.
(Flatten): Duplicate the expression of an others choice only if
it is static or is an aggregate which can itself be flattened.
Eric Botcazou [Wed, 14 Nov 2018 11:42:26 +0000 (11:42 +0000)]
[Ada] Remove couple of recently added dead tests
N_Quantified_Expression and N_Iterated_Component_Association are
unrelated nodes that cannot appear in the same context: the former can
appear wherever an expression node is acceptable whereas the latter can
appear only as an element of a component association list. So a test
combining both most likely contains a dead arm and this change removes a
couple of them.
This patch modifies the analysis (which is really expansion) of null
procedures to set the Ghost mode of the spec when the null procedure
acts as a completion. This ensures that all nodes and entities
generated by the expansion are marked as Ghost, and provide a proper
context for references to Ghost entities.
* sem_ch6.adb (Analyze_Null_Procedure): Capture Ghost and
SPARK-related global state at the start of the routine. Set the
Ghost mode of the completed spec if any. Restore the saved
Ghost and SPARK-related global state on exit from the routine.
gcc/testsuite/
* gnat.dg/ghost1.adb, gnat.dg/ghost1.ads: New testcase.
Eric Botcazou [Wed, 14 Nov 2018 11:42:16 +0000 (11:42 +0000)]
[Ada] More complete information level for -gnatR output
This adds a 4th information level for the -gnatR output, where relevant
compiler-generated types are listed in addition to the information
already output by -gnatR3.
For the following package P:
package P is
type Arr0 is array (Positive range <>) of Boolean;
type Rec (D1 : Positive; D2 : Boolean) is record
C1 : Integer;
C2 : Arr0 (1 .. D1);
case D2 is
when False =>
C3 : Character;
when True =>
C4 : String (1 .. 3);
C5 : Float;
end case;
end record;
type Arr1 is array (1 .. 8) of Rec (1, True);
end P;
the output generated by -gnatR4 must be:
Representation information for unit P (spec)
--------------------------------------------
for Arr0'Alignment use 1;
for Arr0'Component_Size use 8;
for Rec'Object_Size use 17179869344;
for Rec'Value_Size use (if (#2 != 0) then ((((#1 + 15) & -4) + 8) * 8)
else ((((#1 + 15) & -4) + 1) * 8) end);
for Rec'Alignment use 4;
for Rec use record
D1 at 0 range 0 .. 31;
D2 at 4 range 0 .. 7;
C1 at 8 range 0 .. 31;
C2 at 12 range 0 .. ((#1 * 8)) - 1;
C3 at ((#1 + 15) & -4) range 0 .. 7;
C4 at ((#1 + 15) & -4) range 0 .. 23;
C5 at (((#1 + 15) & -4) + 4) range 0 .. 31;
end record;
for Arr1'Size use 1536;
for Arr1'Alignment use 4;
for Arr1'Component_Size use 192;
for Tarr1c'Size use 192;
for Tarr1c'Alignment use 4;
for Tarr1c use record
D1 at 0 range 0 .. 31;
D2 at 4 range 0 .. 7;
C1 at 8 range 0 .. 31;
C2 at 12 range 0 .. 7;
C4 at 16 range 0 .. 23;
C5 at 20 range 0 .. 31;
end record;
2018-11-14 Eric Botcazou <ebotcazou@adacore.com>
gcc/ada/
* doc/gnat_ugn/building_executable_programs_with_gnat.rst
(-gnatR): Document new -gnatR4 level.
* gnat_ugn.texi: Regenerate.
* opt.ads (List_Representation_Info): Bump upper bound to 4.
* repinfo.adb: Add with clause for GNAT.HTable.
(Relevant_Entities_Size): New constant.
(Entity_Header_Num): New type.
(Entity_Hash): New function.
(Relevant_Entities): New set implemented with GNAT.HTable.
(List_Entities): Also list compiled-generated entities present
in the Relevant_Entities set. Consider that the Component_Type
of an array type is relevant.
(List_Rep_Info): Reset Relevant_Entities for each unit.
* switch-c.adb (Scan_Front_End_Switches): Add support for -gnatR4.
* switch-m.adb (Normalize_Compiler_Switches): Likewise
* usage.adb (Usage): Likewise.
Javier Miranda [Wed, 14 Nov 2018 11:42:10 +0000 (11:42 +0000)]
[Ada] Crash on interface equality covered by a renaming declaration
The frontend crashes processing a tagged type that implements an
interface which has an equality primitive (that is, "=") and covers such
primitive by means of a renaming declaration.
2018-11-14 Javier Miranda <miranda@adacore.com>
gcc/ada/
* exp_disp.adb (Expand_Interface_Thunk): Extend handling of
renamings of the predefined equality primitive.
(Make_Secondary_DT): When calling Expand_Interface_Thunk() pass
it the primitive, instead of its Ultimate_Alias; required to
allow the called routine to identify renamings of the predefined
equality operation.
gcc/testsuite/
* gnat.dg/equal5.adb, gnat.dg/equal5.ads: New testcase.
[Ada] Use of Suppress_Initialization with pragma Thread_Local_Storage
This patch allows for aspect/pragma Suppress_Initialization to be an
acceptable form of missing initialization with respect to the semantics
of pragma Thread_Local_Storage.
------------
-- Source --
------------
-- gnat.adc
pragma Initialize_Scalars;
-- pack.ads
with System;
package Pack is
Addr : System.Address
with Thread_Local_Storage, Suppress_Initialization;
end Pack;
* freeze.adb (Check_Pragma_Thread_Local_Storage): New routine. A
variable with suppressed initialization has no initialization
for purposes of the pragma.
(Freeze_Object_Declaration): Remove variable
Has_Default_Initialization as it is no longer used. Use routine
Check_Pragma_Thread_Local_Storage to verify the semantics of
pragma Thread_Local_Storage.
Ed Schonberg [Wed, 14 Nov 2018 11:41:58 +0000 (11:41 +0000)]
[Ada] Missing constraint check on if-expression returning a string
If the context of an if-expression is constrained, its dependent
expressions must obey the constraints of the expected type. Prior to
this patch, this check was performed only for scalar types, by means of
an added conversion. This is now enforced on all types by means of a
qualified expression on each dependent expression.
Compiling ce.adb must yield:
ce.adb:33:21: warning: string length wrong for type "T" defined at line 5
ce.adb:33:21: warning: "Constraint_Error" will be raised at run time
ce.adb:37:39: warning: string length wrong for type "T" defined at line 5
ce.adb:37:39: warning: "Constraint_Error" will be raised at run time
ce.adb:38:39: warning: too few elements for type "T" defined at line 5
ce.adb:38:39: warning: "Constraint_Error" will be raised at run time
ce.adb:39:39: warning: too few elements for type "T" defined at line 5
ce.adb:39:39: warning: "Constraint_Error" will be raised at run time
----
with Text_IO;
procedure Ce is
package Aerodrome_Identifier is
subtype T is String (1 .. 4);
end;
package Flight_Identifier is
type T is
record
ADEP : Aerodrome_Identifier.T;
Counter : Positive;
end record;
end;
procedure Assign (X : Flight_Identifier.T) is
begin
Text_IO.Put_Line (X.ADEP); -- outputs the 4 zero bytes
end;
function Env_Aerodrome_Value return String is ("ABCD");
function Void return String is ("What?");
function Void2 return String is
begin
return "who knows";
end;
Here : Aerodrome_Identifier.T;
type Four is range 1 .. 4;
Nothing : String := "";
begin
Assign((ADEP =>
(if (Void'Length = 5)
then "" --!! This value should always raise Constraint_Error !!
else Env_Aerodrome_Value & "!"),
Counter=> 17));
Here := (if (Void'Length = 5) then "" else Env_Aerodrome_Value);
Here := (if (Void'Length = 5) then Nothing else Env_Aerodrome_Value);
Here := (if (Void'Length = 5) then Void2 (1..3) else Void2 & Void);
end;
----
2018-11-14 Ed Schonberg <schonberg@adacore.com>
gcc/ada/
* sem_res.adb (Resolve_If_Expression): Verify that the subtypes
of all dependent expressions obey the constraints of the
expected type for the if-expression.
(Analyze_Expression): Only add qualificiation to the dependent
expressions when the context type is constrained. Small
adjustment to previous patch.
Eric Botcazou [Wed, 14 Nov 2018 11:41:53 +0000 (11:41 +0000)]
[Ada] Fix assertion failure on pragma Compile_Time_Error in generic unit
There is no point in validating 'Alignment or 'Size of an entity
declared in a generic unit after the back-end has been run, since such
an entity is not passed to the back-end, and this can even lead to an
assertion failure.
2018-11-14 Eric Botcazou <ebotcazou@adacore.com>
gcc/ada/
* sem_prag.adb (Process_Compile_Time_Warning_Or_Error): Don't
register a compile-time warning or error for 'Alignment or 'Size
of an entity declared in a generic unit.
gcc/testsuite/
* gnat.dg/compile_time_error1.adb,
gnat.dg/compile_time_error1.ads,
gnat.dg/compile_time_error1_pkg.ads: New testcase.
Justin Squirek [Wed, 14 Nov 2018 11:41:46 +0000 (11:41 +0000)]
[Ada] Crash on use of generic formal package
This patch fixes an issue whereby a complicated set of generic formal
packages in conjunction with use_clauses may cause a crash during
visibility checking due to a homonym being out of scope during the
checking stage.
2018-11-14 Justin Squirek <squirek@adacore.com>
gcc/ada/
* sem_ch8.adb (Use_One_Package): Add test for out-of-scope
homonyms.
Ed Schonberg [Wed, 14 Nov 2018 11:41:41 +0000 (11:41 +0000)]
[Ada] Compiler crash on decl. with limited aggregate and address clause
This patch fixes a compiler abort on an object declaration whose
expression is an aggregate, when the type of the object is limited and
the declaration is followed by an address clause for the declared
object.
2018-11-14 Ed Schonberg <schonberg@adacore.com>
gcc/ada/
* exp_ch3.adb: (Expand_N_Object_Declaration): If the expression
is a limited aggregate its resolution is delayed until the
object declaration is expanded.
* sem_ch3.adb: (Analyze_Object_Declaration): If the expression
is a limited aggregate and the declaration has a following
address clause indicate that resolution of the aggregate (which
must be built in place) must be delayed.
gcc/testsuite/
* gnat.dg/limited_aggr.adb, gnat.dg/limited_aggr.ads: New
testcase.
Bob Duff [Wed, 14 Nov 2018 11:41:36 +0000 (11:41 +0000)]
[Ada] System'To_Address not always static
System'To_Address is supposed to be static when its parameter is static.
This patch fixes a bug in which it is considered nonstatic when used as
the initial value of a variable with the Thread_Local_Storage aspect, so
the compiler incorrectly gives an error when initializing such a
variable with System'To_Address (0).
2018-11-14 Bob Duff <duff@adacore.com>
gcc/ada/
* sem_attr.adb (To_Address): Simplify setting of
Is_Static_Expression. Remove second (unconditional) call to
Set_Is_Static_Expression -- surely it's not static if the
operand is not. Initialize Static on declaration. Do not try
to fold 'To_Address, even though it's static.
* exp_attr.adb (To_Address): Preserve Is_Static_Expression.
* sinfo.ads, sem_eval.ads, sem_eval.adb (Is_Static_Expression,
Is_OK_Static_Expression, Raises_Constraint_Error): Simplify
documentation. There was too much repetition and redundancy.
projects / thirdparty / gcc.git / log
