Jakub Jelinek [Sat, 25 Feb 2017 08:18:24 +0000 (09:18 +0100)]
re PR debug/77589 (fortran: Missing DW_AT_byte_stride for an array record field selection)
PR debug/77589
include/
* dwarf2.def (DW_OP_GNU_variable_value): New opcode.
gcc/
* dwarf2out.c (struct dw_loc_list_struct): Add noted_variable_value
bitfield.
(size_of_loc_descr): Handle DW_OP_GNU_variable_value.
(output_loc_operands): Handle DW_OP_call_ref and
DW_OP_GNU_variable_value.
(struct variable_value_struct): New type.
(struct variable_value_hasher): Likewise.
(variable_value_hash): New variable.
(string_types): Remove.
(copy_loc_descr): New function.
(add_loc_descr_to_each): Clarify comment. Use copy_loc_descr.
(prepend_loc_descr_to_each): New function.
(add_loc_list): Fix comment typo. Use prepend_loc_descr_to_each
instead of add_loc_descr_to_each if the first argument is single
location list and the second has multiple.
(resolve_args_picking_1): Handle DW_OP_GNU_variable_value.
(loc_list_from_tree_1): For early_dwarf, emit DW_OP_GNU_variable_value
when looking for variable value which doesn't have other location info.
(loc_list_from_tree): Formatting fix.
(gen_array_type_die): Simplify DW_AT_string_length handling.
(adjust_string_types): Remove.
(gen_subprogram_die): Don't call adjust_string_types nor test/set
string_types. Call resolve_variable_values.
(prune_unused_types_walk_loc_descr): Handle DW_OP_GNU_variable_value.
(resolve_addr_in_expr): Likewise. Add A argument.
(copy_deref_exprloc): Remove deref argument. Adjust for the
original expression being DW_OP_GNU_variable_value with optionally
DW_OP_stack_value after it instead of DW_OP_call4 with DW_OP_deref
optionally after it.
(optimize_string_length): Rework for DW_OP_GNU_variable_value.
(resolve_addr): Adjust optimize_string_length and resolve_addr_in_expr
callers. Set remove_AT_byte_size if removing DW_AT_string_length.
(variable_value_hasher::hash, variable_value_hasher::equal): New
methods.
(resolve_variable_value_in_expr, resolve_variable_value,
resolve_variable_values, note_variable_value_in_expr,
note_variable_value): New functions.
(dwarf2out_early_finish): Call note_variable_value on all toplevel
DIEs.
gcc/
* config/nios2/nios2.c (nios2_simple_const_p): Returns false if the
argument isn't a CONST_INT.
(nios2_alternate_compare_const): Assert op is a CONST_INT.
(nios2_valid_compare_const_p): Assert op is a CONST_INT.
(nios2_validate_compare): Bypass alternate compare logic if *op2
is not a CONST_INT.
(ldstwm_operation_p): Return false if first_base is not a REG or
if first_offset is not a CONST_INT.
Matthew Fortune [Fri, 24 Feb 2017 22:34:25 +0000 (22:34 +0000)]
Update paired single tests to cope with ABI fixes for o32/o64
gcc/testsuite/
* gcc.target/mips/mips-ps-type-2.c (move): Force generation
of mov.ps.
* gcc.target/mips/mips-ps-type.c (move): Likewise.
(cond_move1): Simplify condition to force generation of
mov[nz].ps.
(cond_move2): Likewise.
* config/arc/arc.c (arc_ccfsm_advance): Only take the PATTERN of
this_insn if it is an INSN or JUMP_INSN.
(force_offsettable): Look at base, not at addr.
* config/arc/predicates.md (brcc_nolimm_operator): Don't call INTVAL
on things that aren' necessarily CONST_INTs.
Uros Bizjak [Fri, 24 Feb 2017 18:18:57 +0000 (19:18 +0100)]
invoke.texi (x86 Options, [...]): Mention that -mfpmath=sse is the default also for x86-32 targets with SSE2...
* doc/invoke.texi (x86 Options, -mfpmath=sse): Mention that
-mfpmath=sse is the default also for x86-32 targets with SSE2
instruction set when @option{-ffast-math} is enabled
Toma Tabacu [Fri, 24 Feb 2017 11:03:12 +0000 (11:03 +0000)]
Use logical_op_short_circuit to skip targets in ssa-thread-14.c.
gcc/testsuite/
* gcc.dg/tree-ssa/ssa-thread-14.c (dg-options): Use
logical_op_short_circuit to skip targets.
(dg-additional-options): Don't pass -mbranch-cost=2 for MIPS, AVR
and s390.
Eric Botcazou [Fri, 24 Feb 2017 11:03:08 +0000 (11:03 +0000)]
decl.c (gnat_to_gnu_field): Do not remove the wrapper around a justified modular type if...
* gcc-interface/decl.c (gnat_to_gnu_field): Do not remove the wrapper
around a justified modular type if it doesn't have the same scalar
storage order as the enclosing record type.
Aldy Hernandez [Fri, 24 Feb 2017 11:02:16 +0000 (11:02 +0000)]
tree-ssa-loop-niter.c (number_of_iterations_exit): Update function comment to reflect reality.
* tree-ssa-loop-niter.c (number_of_iterations_exit): Update
function comment to reflect reality.
(loop_exits_before_overflow): Fix typo in function description.
Eric Botcazou [Fri, 24 Feb 2017 10:48:38 +0000 (10:48 +0000)]
utils.c (fold_bit_position): New function.
* gcc-interface/utils.c (fold_bit_position): New function.
(rest_of_record_type_compilation): Call it instead of bit_position to
compute the field position and remove the call to remove_conversions.
(compute_related_constant): Factor out the multiplication in both
operands, if any, and streamline the final test.
Eric Botcazou [Fri, 24 Feb 2017 10:21:39 +0000 (10:21 +0000)]
c-ada-spec.c (dump_ada_function_declaration): Add comment about the treatment of parameters with pointer-to-tagged type and...
c-family/
* c-ada-spec.c (dump_ada_function_declaration): Add comment about the
treatment of parameters with pointer-to-tagged type and tidy up.
(print_ada_methods): Remove the special treatment of C++ static member
functions.
ada/
* gcc-interface/decl.c: Include demangle.h.
(is_cplusplus_method): Return again true for a primitive operation
only if it is dispatching. For a subprogram with an interface name,
call the demangler to get the number of C++ parameters and compare it
with the number of Ada parameters.
Eric Botcazou [Fri, 24 Feb 2017 10:07:45 +0000 (10:07 +0000)]
trans.c (Handled_Sequence_Of_Statements_to_gnu): If there is no end label...
* gcc-interface/trans.c (Handled_Sequence_Of_Statements_to_gnu): If
there is no end label, put the location of the At_End procedure on
the call to the procedure.
Eric Botcazou [Fri, 24 Feb 2017 10:00:04 +0000 (10:00 +0000)]
misc.c (gnat_type_max_size): Try to return a meaningful value for array types with TYPE_INDEX_TYPE set on...
* gcc-interface/misc.c (gnat_type_max_size): Try to return a meaningful
value for array types with TYPE_INDEX_TYPE set on their domain type.
* gcc-interface/utils.c (max_size): For operations and expressions, do
not build a new node if the operands have not changed or are missing.
Jakub Jelinek [Thu, 23 Feb 2017 22:05:19 +0000 (23:05 +0100)]
re PR tree-optimization/79389 (30% performance regression in SciMark2 MonteCarlo)
PR tree-optimization/79389
* ifcvt.c (struct noce_if_info): Add rev_cond field.
(noce_reversed_cond_code): New function.
(noce_emit_store_flag): Use rev_cond if non-NULL instead of
reversed_comparison_code. Formatting fix.
(noce_try_store_flag): Test rev_cond != NULL in addition to
reversed_comparison_code.
(noce_try_store_flag_constants): Likewise.
(noce_try_store_flag_mask): Likewise.
(noce_try_addcc): Use rev_cond if non-NULL instead of
reversed_comparison_code.
(noce_try_cmove_arith): Likewise. Formatting fixes.
(noce_try_minmax, noce_try_abs): Clear rev_cond.
(noce_find_if_block): Initialize rev_cond.
(find_cond_trap): Call noce_get_condition with then_bb == trap_bb
instead of false as last argument never attempt to reverse it
afterwards.
Jakub Jelinek [Thu, 23 Feb 2017 07:49:06 +0000 (08:49 +0100)]
re PR middle-end/79665 (gcc's signed (x*x)/200 is slower than clang's)
PR middle-end/79665
* internal-fn.c (get_range_pos_neg): Moved to ...
* tree.c (get_range_pos_neg): ... here. No longer static.
* tree.h (get_range_pos_neg): New prototype.
* expr.c (expand_expr_real_2) <case TRUNC_DIV_EXPR>: If both arguments
are known to be in between 0 and signed maximum inclusive, try to
expand both unsigned and signed divmod and use the cheaper one from
those.
The fsel define_insn uses fpr_reg_operand for its predicates. This
won't work because passes can put a hard register in the operands: in
the testcase, combine likes to forward the parameter registers to what
then is still an smin, and then split1 uses "*s<minmax><mode>3_fpr"
(which has gpc_reg_operand). And then we have a GPR in the operand,
which does not match fpr_reg_operand.
It seems to me the predicates should be gpc_reg_operand here as well.
This patch changes that.
PR target/79211
* config/rs6000/rs6000.md (*fsel<SFDF:mode><SFDF2:mode>4): Use
gpc_reg_operand instead of fpr_reg_operand.
Jakub Jelinek [Wed, 22 Feb 2017 22:36:20 +0000 (23:36 +0100)]
re PR c++/79664 (ICE with #pragma omp parallel in constexpr function)
PR c++/79664
* parser.c (cp_parser_omp_teams, cp_parser_omp_target): Use
SET_EXPR_LOCATION on OMP_TARGET/OMP_TEAMS tree.
* constexpr.c (potential_constant_expression_1): Handle
OMP_*, OACC_* and CILK_* trees. Use error_at with
EXPR_LOC_OR_LOC (t, input_location) computed early
instead of error, or error_at with location_of (t).
Jakub Jelinek [Wed, 22 Feb 2017 21:33:47 +0000 (22:33 +0100)]
dwarf2out.c (gen_variable_die): For -gdwarf-5...
* dwarf2out.c (gen_variable_die): For -gdwarf-5, use DW_TAG_variable
instead of DW_TAG_member for static data member declarations and don't
set no_linkage_name for static inline data members.
(gen_member_die): For -gdwarf-5 don't change DW_TAG_variable
to DW_TAG_member.
Marek Polacek [Wed, 22 Feb 2017 19:31:49 +0000 (19:31 +0000)]
re PR c++/79653 (ICE on invalid c++ code in cp_check_const_attributes in cp/decl2.c:1423)
PR c++/79653
* parser.c (cp_parser_std_attribute_spec): Don't build the attribute
if the alignas expression is erroneous.
* pt.c (tsubst_attribute): If tsubst_pack_expansion fails, return
error_mark_node.
* g++.dg/cpp0x/alignas10.C: New test.
* g++.dg/cpp0x/alignas9.C: New test.
Jakub Jelinek [Wed, 22 Feb 2017 17:17:17 +0000 (18:17 +0100)]
re PR target/70465 (Poor code for x87 asm)
PR target/70465
* reg-stack.c (emit_swap_insn): Treat (float_extend:?F (mem:?F))
and (const_double:?F) like (mem:?F) for the purpose of fxch %st(1)
elimination by swapping fld*.
Richard Biener [Wed, 22 Feb 2017 15:00:39 +0000 (15:00 +0000)]
re PR middle-end/79673 (GIMPLE verification fails when compiling code with __seg_gs)
2017-02-22 Richard Biener <rguenther@suse.de>
PR tree-optimization/79673
* tree-ssa-pre.c (compute_avail): Use wide_int_to_tree to
convert the [TARGET_]MEM_REF offset INTEGER_CST, scrapping off
irrelevant address-space qualifiers and avoiding a
ADDR_SPACE_CONVERT_EXPR from fold_convert.
Richard Biener [Wed, 22 Feb 2017 12:11:27 +0000 (12:11 +0000)]
re PR tree-optimization/79666 (wrong code (SIGFPE) at -O2 on x86_64-linux-gnu (in both 32-bit and 64-bit modes))
2017-02-22 Richard Biener <rguenther@suse.de>
PR tree-optimization/79666
* tree-vrp.c (extract_range_from_binary_expr_1): Make sure
to not symbolically negate if that may introduce undefined
overflow.
Martin Liska [Wed, 22 Feb 2017 09:45:42 +0000 (10:45 +0100)]
Remove wrong assert about gcov_type (PR lto/79587).
2017-02-22 Martin Liska <mliska@suse.cz>
PR lto/79587
* data-streamer-in.c (streamer_read_gcov_count): Remove assert.
* data-streamer-out.c (streamer_write_gcov_count_stream):
Likewise.
* value-prof.c (stream_out_histogram_value): Make assert more
precise based on type of counter.
2017-02-22 Martin Liska <mliska@suse.cz>
PR lto/79587
* gcc.dg/tree-prof/pr79587.c: New test.
Jakub Jelinek [Tue, 21 Feb 2017 18:00:35 +0000 (19:00 +0100)]
re PR c++/79654 (ICE on invalid c++ code in register_dtor_fn in cp/decl.c:7877)
PR c++/79654
* decl.c (cp_finish_decomp): Don't set decl's type to error_mark_node
on error.
* pt.c (tsubst_decomp_names): Return error_mark_node if the first
decl after the decomposition artificial decl has error_mark_node.
* decl2.c (prune_vars_needing_no_initialization): Use error_operand_p
instead of just == error_mark_node comparison.
* g++.dg/cpp1z/decomp26.C: New test.
Co-Authored-By: Paolo Carlini <paolo.carlini@oracle.com>
From-SVN: r245639
Jakub Jelinek [Tue, 21 Feb 2017 17:59:07 +0000 (18:59 +0100)]
re PR sanitizer/79589 (ICE in gimplify_compound_expr (gimplify.c:5712) with -fsanitize=undefined)
PR sanitizer/79589
* decl.c: Include gimplify.h.
(cp_finish_decomp): Make sure there is no sharing of trees
in between DECL_VALUE_EXPR of decomposition decls.
Jakub Jelinek [Tue, 21 Feb 2017 17:46:42 +0000 (18:46 +0100)]
re PR target/79633 (ICE in gimple_call_arg, at gimple.h:3163)
PR target/79633
* tree-chkp-opt.c (chkp_optimize_string_function_calls): Use
is_gimple_call instead of comparing gimple_code with GIMPLE_CALL.
Use gimple_call_builtin_p.
Jakub Jelinek [Tue, 21 Feb 2017 14:39:21 +0000 (15:39 +0100)]
re PR target/79494 (ICE in maybe_record_trace_start, at dwarf2cfi.c:2330)
PR target/79494
* config/i386/i386.c (ix86_expand_split_stack_prologue): Call
make_reg_eh_region_note_nothrow_nononlocal on call_insn.
* config/rs6000/rs6000.c: Include except.h.
(rs6000_expand_split_stack_prologue): Call
make_reg_eh_region_note_nothrow_nononlocal on the call insn.
Martin Jambor [Tue, 21 Feb 2017 13:55:32 +0000 (14:55 +0100)]
[PATCH] Properly deprecate -fipa-cp-alignment
2017-02-21 Martin Jambor <mjambor@suse.cz>
* common.opt (-fipa-cp-alignment): Mark as ignored and preserved
for backward compatibility only.
* doc/invoke.texi (Option Summary): Remove all references to
-fipa-cp-alignment.
Eric Botcazou [Tue, 21 Feb 2017 08:42:54 +0000 (08:42 +0000)]
re PR ada/67205 (eliminate No_Implicit_Dynamic_Code restriction violations)
PR ada/67205
* config/arm/arm.c (TARGET_CUSTOM_FUNCTION_DESCRIPTORS): Define.
(arm_function_ok_for_sibcall): Return false for an indirect call by
descriptor if all the argument registers are used.
(arm_relayout_function): Use FUNCTION_ALIGNMENT macro to adjust the
alignment of the function.
Jakub Jelinek [Tue, 21 Feb 2017 08:08:40 +0000 (09:08 +0100)]
re PR tree-optimization/61441 (ARM aarch64 fails to quiet signaling NaN)
PR tree-optimization/61441
* simplify-rtx.c (simplify_const_unary_operation): For
-fsignaling-nans and sNaN operand, return NULL_RTX rather than
the sNaN unmodified.
* Makefile.in (BUILD_SYSTEM_HEADER_DIR): New make variabe.
(LIMITS_H_TEST, if_multiarch, stmp-fixinc): Use BUILD_SYSTEM_HEADER_DIR
instead of SYSTEM_HEADER_DIR.