rth [Tue, 22 Dec 2015 19:42:31 +0000 (19:42 +0000)]
Give EH_ELSE access to __builtin_eh_pointer
The must-not-throw wrapper for protect_cleanup_actions gets in
the way of being able to access __builtin_eh_pointer without
confusion as the identit of the exception to which we are
referring (b_eh_p has no usable argument up to this point).
Since EH_ELSE never comes from user derived code, let's drop
the c++ specific wrapping.
* tree-eh.c (honor_protect_cleanup_actions): Do not wrap
eh_else in a must-not-throw; set ehp_region for it too.
uros [Tue, 22 Dec 2015 15:38:25 +0000 (15:38 +0000)]
[PATCH] Use call-clobbered register for sibcall via GOT
From: H.J. Lu <hongjiu.lu@intel.com>
Since sibcall never returns, we can only use call-clobbered register as
GOT base. Otherwise, callee-saved register used as GOT base won't be
properly restored. sibcall_memory_operand is changed to allow 32-bit
GOT slot only with pseudo register as GOT base for RTL expansion. 2
new patterns, *sibcall_GOT_32 and *sibcall_value_GOT_32, are added to
expose GOT base register to register allocator so that call-clobbered
register will be used for GOT base.
gcc/
PR target/68937
* config/i386/i386.c (ix86_function_ok_for_sibcall): Count
call to global function via GOT slot as indirect call.
* config/i386/i386.md (*sibcall_GOT_32): New pattern.
(*sibcall_value_GOT_32): Likewise.
* config/i386/predicates.md (sibcall_memory_operand): Rewrite.
Allow 32-bit GOT slot only with pseudo register as GOT base.
(GOT32_symbol_operand): New predicate.
ssaraswati [Tue, 22 Dec 2015 14:04:30 +0000 (14:04 +0000)]
This series of patches fix PR61441. This patch avoids various transformations
with signaling NaN operands when flag_signaling_nans is on, to avoid folding
which would lose exceptions.
Bootstrapped & regression-tested on x86_64-linux-gnu.
gcc/
* fold-const.c (const_binop): Convert sNaN to qNaN when
flag_signaling_nans is off.
(const_unop): Avoid the operation, other than NEGATE and
ABS, if flag_signaling_nans is on and the operand is an sNaN.
(fold_convert_const_real_from_real): Avoid the operation if
flag_signaling_nans is on and the operand is an sNaN.
(integer_valued_real_unary_p): Update comment stating it
returns false for sNaN values.
(integer_valued_real_binary_p, integer_valued_real_call_p): Same.
(integer_valued_real_single_p): Same.
(integer_valued_real_invalid_p, integer_valued_real_p): Same.
* fold-const-call.c (fold_const_pow): Avoid the operation
if flag_signaling_nans is on and the operand is an sNaN.
(fold_const_builtin_load_exponent) Same.
(fold_const_call_sss): Same for CASE_CFN_POWI.
* gimple-fold.c (gimple_assign_integer_valued_real_p): Same.
(gimple_call_integer_valued_real_p): Same.
(gimple_phi_integer_valued_real_p): Same.
(gimple_stmt_integer_valued_real_p): Same.
* simplify-rtx.c (simplify_const_unary_operation): Avoid the
operation if flag_signaling_nans is on and the operand is an sNaN.
(simplify_const_binary_operation): Same.
* tree-ssa-math-opts.c (gimple_expand_builtin_pow): Avoid the
operation if flag_signaling_nans is on and the operand is an sNaN.
law [Tue, 22 Dec 2015 03:09:08 +0000 (03:09 +0000)]
* dbgcnt.def (dom_unreachable_edges): New debug counter.
* tree-ssa-dom.c: Include dbgcnt.h
(dom_opt_dom_walker::before_dom_children): Handle the new counter.
ian [Tue, 22 Dec 2015 01:58:17 +0000 (01:58 +0000)]
* go-gcc.cc (class Bvariable): Remove Gcc_tree parent class. Add
t_ and orig_type_ fields. Add new two parameter constructor. Add
get_tree and get_decl methods.
(Gcc_backend::var_expression): Pass location to var get_tree.
(Gcc_backend::global_variable): Don't add VIEW_CONVERT_EXPR. Use
two parameter constructor for Bvariable.
(Gcc_backend::global_variable_set_init): Don't remove
VIEW_CONVERT_EXPR. Use var get_decl, not get_tree.
(Gcc_backend::write_global_definitions): Likewise.
(Gcc_backend::init_statement): Call var get_decl, not get_tree.
(Gcc_backend::block): Likewise.
(Gcc_backend::implicit_variable_set_init): Likewise.
(Gcc_backend::immutable_struct_set_init): Likewise.
(Gcc_backend::function_set_parameters): Likewise.
ian [Tue, 22 Dec 2015 00:10:07 +0000 (00:10 +0000)]
* go-gcc.cc (Gcc_backend::global_variable): If type is zero-sized,
add a VIEW_CONVERT_EXPR to the tree.
(Gcc_backend::global_variable_set_init): Remove any
VIEW_CONVERT_EXPR.
(Gcc_backend::write_global_definitions): Likewise.
dmalcolm [Mon, 21 Dec 2015 17:45:34 +0000 (17:45 +0000)]
C and C++ FE: fix source ranges for binary ops
gcc/c-family/ChangeLog:
* c-common.c (binary_op_error): Convert first param from
location_t to rich_location * and use it when emitting an error.
* c-common.h (binary_op_error): Convert first param from
location_t to rich_location *.
gcc/c/ChangeLog:
* c-typeck.c: Include "gcc-rich-location.h".
(build_binary_op): In the two places that call binary_op_error,
create a gcc_rich_location and populate it with the location of
the binary op and its two operands.
gcc/cp/ChangeLog:
* typeck.c (cp_build_binary_op): Update for change in signature
of build_binary_op. Use error_at to replace an implicit use
of input_location with param "location" in "invalid operands"
error.
(cp_build_binary_op): Replace an error with an error_at, using
"location", rather than implicitly using input_location.
gcc/testsuite/ChangeLog:
* g++.dg/diagnostic/bad-binary-ops.C: New test case.
* gcc.dg/bad-binary-ops.c: New test case.
gcc.dg/plugin/diagnostic_plugin_show_trees.c (get_range_for_expr):
Remove material copied from gcc-rich-location.c
(gcc_rich_location::add_expr): Likewise.
pmderodat [Mon, 21 Dec 2015 15:44:32 +0000 (15:44 +0000)]
DWARF: allow dynamic data member offsets for inheritance info
An unintended effect of the recently introduced machinery to handle
dynamic data member offsets in variable-length records (when
-fgnat-encodings=minimal) prevented GCC from describing correctly
inheritance information for classes in C++, which is a regression.
This change rectifies this machinery in this case.
gcc/ChangeLog:
* dwarf2out.c (add_data_member_location_attribute): Do not
disable dynamic data member offsets descriptions for TREE_BINFO
members.
ssaraswati [Mon, 21 Dec 2015 07:59:30 +0000 (07:59 +0000)]
This series of patches fix PR61441. This patch makes resulting NaN values
to be quiet NaN for real value operations, irrespective of the flag_signaling_nans
flag. The caller has the responsibility to avoid the operation if flag_signaling_nans
is on.
gcc/
* real.c (do_add): Make resulting NaN value to be qNaN.
(do_multiply, do_divide, do_fix_trunc): Same.
(real_arithmetic, real_ldexp, real_convert): Same.
(real_isinteger): Updated comment stating it returns false for sNaN.
jason [Sun, 20 Dec 2015 18:38:37 +0000 (18:38 +0000)]
PR c++/67411
* lambda.c (generic_lambda_fn_p): Split out from...
(maybe_add_lambda_conv_op): ...here.
* semantics.c (process_outer_var_ref): Don't defer maybe-constant
variables in a generic lambda.
* pt.c (instantiate_non_dependent_or_null): New.
* init.c (constant_value_1): Use it.
* cp-tree.h: Declare it and generic_lambda_fn_p.
ebotcazou [Sun, 20 Dec 2015 10:30:27 +0000 (10:30 +0000)]
* gcc-interface/decl.c (gnat_to_gnu_entity) <E_Record_Type>: During
layout in type_annotate_only mode, skip discriminants of derived
tagged types renaming those of the parent type.
In type_annotate_only mode, if the type is tagged, do not override a
size clause but take into account the alignment of the tag.
(annotate_rep): In type_annotate_only mode, deal with discriminants
of derived tagged types renaming those of the parent type.
ebotcazou [Sun, 20 Dec 2015 10:27:37 +0000 (10:27 +0000)]
* gcc-interface/utils.c (maybe_pad_type): In type_annotate_only mode,
retrieve the component type if this is an array and do not issue the
warning for concurrent types.
ssaraswati [Sun, 20 Dec 2015 09:13:51 +0000 (09:13 +0000)]
This series of patches are for fixing PR61441. This patch modifies code to use
REAL_VALUE_ISSIGNALING_NAN instead of REAL_VALUE_ISNAN to avoid the operatins
only for sNaN operands.
Bootstrapped & regression-tested on x86_64-linux-gnu.
gcc/
* fold-const.c (const_binop): Use REAL_VALUE_ISSIGNALING_NAN instead
of REAL_VALUE_ISNAN to avoid the operation for sNaN operands.
* simplify-rtx.c (simplify_const_binary_operation): Same.
hubicka [Sun, 20 Dec 2015 05:50:29 +0000 (05:50 +0000)]
PR middle-end/65337
* tree-ssa-dce.c (bb_postorder): New static var.
(forward_edge_to_pdom): Remove.
(remove_dead_stmt): Instead of redirecting edges only keep an edge
on a path to nearest live BB.
(eliminate_unnecessary_stmts): Free bb_postorder.
* cfganal.c (dfs_find_deadend): Add START_POINTES.
* cfganal.h (inverted_post_order_compute): Update prototype.
amylaar [Sat, 19 Dec 2015 19:29:12 +0000 (19:29 +0000)]
2015-12-19 Andrew Burgess <andrew.burgess@embecosm.com>
* config/arc/arc.md (*storeqi_update): Use 'memory_operand' and
fix RTL pattern to include the plus.
(*storehi_update): Likewise.
(*storesi_update): Likewise.
(*storesf_update): Likewise.
* config/arc/predicates.md (store_update_operand): Delete.
amylaar [Sat, 19 Dec 2015 19:22:51 +0000 (19:22 +0000)]
gcc:
2015-12-19 Andrew Burgess <andrew.burgess@embecosm.com>
* config/arc/arc.md (*loadqi_update): Use new 'any_mem_operand'
and fix RTL pattern to include the plus.
(*load_zeroextendqisi_update): Likewise.
(*load_signextendqisi_update): Likewise.
(*loadhi_update): Likewise.
(*load_zeroextendhisi_update): Likewise.
(*load_signextendhisi_update): Likewise.
(*loadsi_update): Likewise.
(*loadsf_update): Likewise.
* config/arc/predicates.md (load_update_operand): Delete.
(any_mem_operand): New predicate.
gcc/testsuite:
2015-12-19 Andrew Burgess <andrew.burgess@embecosm.com>
danglin [Sat, 19 Dec 2015 01:50:00 +0000 (01:50 +0000)]
PR target/68729
* config/pa/pa.c (pa_emit_move_sequence): Reorganize handling of
floating-point reloads. Only reload operands that are not valid
floating-point memory operands.
jakub [Fri, 18 Dec 2015 22:16:24 +0000 (22:16 +0000)]
PR debug/68860
* ipa-split.c (split_function): Only perform caller side
modifications for decl_debug_args here.
* cgraph.c: Include gimplify.h.
(cgraph_edge::redirect_call_stmt_to_callee): Add caller side
debug stmts for decl_debug_args. Spelling fix in a comment.
* tree-inline.c (tree_function_versioning): Populate decl_debug_args
for args_to_skip arguments and add callee side debug stmts.
Formatting fixes. Avoid shadowing i variable.
* gcc.dg/guality/pr68860-1.c: New test.
* gcc.dg/guality/pr68860-2.c: New test.
ville [Fri, 18 Dec 2015 21:27:53 +0000 (21:27 +0000)]
2015-12-18 Ville Voutilainen <ville.voutilainen@gmail.com>
Fix a regression introduced by the fix of libstdc++/68276.
* src/c++11/ios.cc (_M_grow_words): Catch bad_alloc again so that
bad_array_new_length is handled properly.
paolo [Fri, 18 Dec 2015 18:18:47 +0000 (18:18 +0000)]
/cp
2015-12-18 Paolo Carlini <paolo.carlini@oracle.com>
PR c++/67592
* decl.c (grokdeclarator): Reject constexpr virtual member functions;
in error messages, prefer %<virtual%> and %<constexpr%> to virtual
and constexpr, respectively.
/testsuite
2015-12-18 Paolo Carlini <paolo.carlini@oracle.com>
rguenth [Fri, 18 Dec 2015 13:52:11 +0000 (13:52 +0000)]
2015-12-18 Richard Biener <rguenther@suse.de>
PR testsuite/68776
* tree-vect-patterns.c (struct vect_recog_func): New.
(vect_vect_recog_func_ptrs): Make an array of function name pairs.
(vect_pattern_recog_1): Print the recognized pattern name. Return
true if a pattern was recognized.
(vect_pattern_recog): Adjust. Stop iterating over pattern recognition
functions on a stmt if a pattern was recognized.
pault [Fri, 18 Dec 2015 09:34:13 +0000 (09:34 +0000)]
2015-12-18 Paul Thomas <pault@gcc.gnu.org>
PR fortran/68196
*expr.c (gfc_has_default_initializer): Prevent infinite recursion
through this function for procedure pointer components.
* trans-array.c (structure_alloc_comps): Ditto twice.
2015-12-18 Paul Thomas <pault@gcc.gnu.org>
PR fortran/68196
* gfortran.dg/proc_ptr_48.f90: New test.
krebbel [Fri, 18 Dec 2015 09:20:53 +0000 (09:20 +0000)]
S/390: Deprecate g5 and g6 CPU levels
gcc/ChangeLog:
2015-12-18 Andreas Krebbel <krebbel@linux.vnet.ibm.com>
* config/s390/driver-native.c (s390_host_detect_local_cpu): Pick
z900 -march -mtune even on g5 and g6 machines.
* config/s390/s390.c (s390_option_override_internal): New argument
to distingiush between options and target attributes. Emit
deprecation warning for g5 and g6 arch/tune options.
(s390_option_override): Set new s390_option_override_internal
argument to true.
(s390_valid_target_attribute_tree): Set new
s390_option_override_internal argument to false.
* config/s390/s390.h: Default to z900 in the driver.
* doc/invoke.texi: Document the deprecation of g5 and g6.
gcc/testsuite/ChangeLog:
2015-12-18 Andreas Krebbel <krebbel@linux.vnet.ibm.com>
* gcc.target/s390/hotpatch-8.c: Add -Wno-deprecated to options.
* gcc.target/s390/hotpatch-9.c: Likewise.
ppalka [Fri, 18 Dec 2015 02:25:39 +0000 (02:25 +0000)]
Fix PR c++/68831 (superfluous -Waddress warning for C++ delete)
gcc/cp/ChangeLog:
PR c++/68831
* init.c (build_delete): Use a warning sentinel to disable
-Waddress warnings when building the conditional that tests
if the operand is NULL.
law [Thu, 17 Dec 2015 23:33:44 +0000 (23:33 +0000)]
[PATCH] Limit path splitting to loops we optimize for speed
* doc/invoke.texi (-O2 options): Remove -fsplit-paths.
(-O3 options): Add -fsplit-paths.
* gimple-ssa-split-paths.c: Include predict.h
(split_paths): Only split paths in a loop that should be
optimized for speed.
* opts.c (default_options_table): Move -fsplit-paths from -O2 to
-O3.
* gcc.dg/tree-ssa/split-path-1.c: Explicitly ask for path
splitting optimizations.
dmalcolm [Thu, 17 Dec 2015 21:25:36 +0000 (21:25 +0000)]
C FE: improvements to ranges of bad return values
gcc/c/ChangeLog:
* c-parser.c (c_parser_statement_after_labels): When calling
c_finish_return, Use the return expression's location if it has
one, falling back to the location of the first token within it.
* c-typeck.c (c_finish_return): When issuing warnings about
the incorrect presence/absence of a return value, issue a note
showing the declaration of the function.
gcc/testsuite/ChangeLog:
* gcc.dg/diagnostic-range-bad-return.c: New test case.
denisc [Thu, 17 Dec 2015 16:55:39 +0000 (16:55 +0000)]
* config/avr/avr.h (MOVE_MAX): Set value to 1.
(MOVE_MAX_PIECES): Define.
(MOVE_RATIO): Define.
* config/avr/avr.c (TARGET_USE_BY_PIECES_INFRASTRUCTURE_P):
Provide target hook.
(avr_use_by_pieces_infrastructure_p): New function.
pmderodat [Thu, 17 Dec 2015 14:10:32 +0000 (14:10 +0000)]
DWARF: describe properly Ada packed arrays
gcc/ada/ChangeLog:
* gcc-interface/ada-tree.h
(TYPE_IMPLEMENTS_PACKED_ARRAY_P, TYPE_CAN_HAVE_DEBUG_TYPE_P,
TYPE_ORIGINAL_PACKED_ARRAY, SET_TYPE_ORIGINAL_PACKED_ARRAY): New
macros.
* gcc-interface/decl.c (add_parallel_type_for_packed_array):
Rename to associate_original_type_to_packed_array. When
-fgnat-encodings=minimal, set original packed array type as so
instead of as a parallel type to the implementation type. In
this case, also rename the implementation type to the name of
the original array type.
(gnat_to_gnu_entity): Update invocations to
add_parallel_type_for_packed_array. Tag ARRAY_TYPE nodes for
packed arrays with the TYPE_PACKED flag.
When -fgnat-encodings=minimal:
- strip ___XP suffixes in packed arrays' names;
- set the debug type for padding records around packed arrays
to the packed array;
- do not attach ___XUP types as parallel types of constrained
array types.
* gcc-interface/misc.c (gnat_print_type): Update to handle
orignal packed arrays.
(gnat_get_debug_type): Update to reject packed arrays
implementation types.
(get_array_bit_stride): New.
(gnat_get_array_descr_info): Add packed arrays handling.
* gcc-interface/utils.c (maybe_pad_type): When
-fgnat-encodings=minimal, set the name of the padding type to
the one of the original packed type, if any. Fix TYPE_DECL
peeling around the name of the input type.
pmderodat [Thu, 17 Dec 2015 14:10:24 +0000 (14:10 +0000)]
DWARF: add a language hook for scalar biased types
Front-ends like GNAT for Ada sometimes use biased encodings for integral
types. This change creates a new language hook so that the bias
information can make it into the debugging information back-end and
introduces an experimental DWARF attribute to hold it.
gcc/ada/ChangeLog:
* gcc-interface/misc.c (gnat_get_type_bias): New.
(LANG_HOOKS_GET_TYPE_BIAS): Redefine macro to implement the
get_type_bias language hook.
gcc/ChangeLog:
* langhooks.h (struct lang_hooks_for_types): New get_bias_field.
* langhooks-def.h (LANG_HOOKS_GET_TYPE_BIAS): New.
(LANG_HOOKS_FOR_TYPES_INITIALIZER): Initialize the
get_bias_field.
* dwarf2out.c
(base_type_die): In non-strict DWARF mode, invoke the
get_type_bias language hook for INTEGER_TYPE nodes. If it
returns a bias, emit an attribute for it.
(subrange_type_die): Change signature to handle bias. If
non-strict DWARF mode, emit an attribute for it, if one passed.
(modified_type_die): For subrange types, invoke the
get_type_bias langage hook and pass the bias to
subrange_type_die.
projects / thirdparty / gcc.git / log
