Jakub Jelinek [Thu, 28 Apr 2016 17:10:14 +0000 (19:10 +0200)]
re PR target/70821 (x86_64: __atomic_fetch_add/sub() uses XADD rather than DECL in some cases)
PR target/70821
* config/i386/sync.md (define_peephole2 *atomic_fetch_add_cmp<mode>):
Add new peephole2 where the first insn is *mov<mode>_or instead of
*mov<mode>_internal.
Expanders do not have more elements in the operands array than declared
in the pattern. So, we cannot use operands[5] here. Instead just
declare and use another rtx.
PR target/70668
* config/nds32/nds32.md (casesi): Don't access the operands array
out of bounds.
Bill Seurer [Thu, 28 Apr 2016 16:01:52 +0000 (16:01 +0000)]
This patch adds support for the signed and unsigned int versions of the...
This patch adds support for the signed and unsigned int versions of the
vec_adde altivec builtins from the Power Architecture 64-Bit ELF V2 ABI
OpenPOWER ABI for Linux Supplement (16 July 2015 Version 1.1). There are
many of the builtins that are missing and this is the first of a series
of patches to add them.
There aren't instructions for the int versions of vec_adde so the
output code is built from other built-ins that do have instructions
which in this case is just two vec_adds with a vec_and to ensure the
carry vector is comprised of only the values 0 or 1.
The new test cases are executable tests which verify that the generated
code produces expected values. C macros were used so that the same
test case could be used for both the signed and unsigned versions. An
extra executable test case is also included to ensure that the modified
support for the __int128 versions of vec_adde is not broken. The same
test case could not be used for both int and __int128 because of some
differences in loading and storing the vectors.
Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no
regressions. Is this ok for trunk?
[gcc]
2016-04-28 Bill Seurer <seurer@linux.vnet.ibm.com>
* config/rs6000/rs6000-builtin.def (vec_adde): Change vec_adde to a
special case builtin.
* config/rs6000/rs6000-c.c (altivec_overloaded_builtins): Remove
ALTIVEC_BUILTIN_VEC_ADDE.
* config/rs6000/rs6000-c.c (altivec_resolve_overloaded_builtin): Add
support for ALTIVEC_BUILTIN_VEC_ADDE.
* config/rs6000/rs6000.c (altivec_init_builtins): Add definition
for __builtin_vec_adde.
[gcc/testsuite]
2016-04-28 Bill Seurer <seurer@linux.vnet.ibm.com>
* gcc.target/powerpc/vec-adde.c: New test.
* gcc.target/powerpc/vec-adde-int128.c: New test.
* gcc.target/i386/avx-vround-1.c: New test.
* gcc.target/i386/avx-vround-2.c: New test.
* gcc.target/i386/avx512vl-vround-1.c: New test.
* gcc.target/i386/avx512vl-vround-2.c: New test.
Martin Jambor [Thu, 28 Apr 2016 14:35:04 +0000 (16:35 +0200)]
Verify that context of local DECLs is the current function
2016-04-28 Martin Jambor <mjambor@suse.cz>
* tree-cfg.c (verify_expr): Verify that local declarations belong to
this function. Call verify_expr on MEM_REFs and bases of other
handled_components.
[ARC] Don't use drsub* instructions when selecting fpuda.
The double precision floating point assist instructions are not
implementing the reverse double subtract instruction (drsub) found in
the FPX extension.
* config/arc/arc.md (cpu_facility): Add fpx variant.
(subdf3): Prohibit use reverse sub when assist operations option
is enabled.
* config/arc/fpx.md (subdf3_insn, *dsubh_peep2_insn): Allow drsub
instructions only when FPX is enabled.
* testsuite/gcc.target/arc/trsub.c: New test.
i386.md (*fop_<mode>_1_mixed): Do not check for mult_operator when calculating "type" attribute.
* config/i386/i386.md (*fop_<mode>_1_mixed): Do not check for
mult_operator when calculating "type" attribute.
(*fop_<mode>_1_i387): Ditto.
(*fop_xf_1_i387): Ditto.
(x87 stack loads peephole2): Add "reg = op (mem, reg)" peephole2.
Use std::swap to swap operands. Use RTL expressions to generate
converted pattern.
Eduard Sanou [Thu, 28 Apr 2016 09:12:05 +0000 (09:12 +0000)]
c-common.c (get_source_date_epoch): New function...
gcc/c-family/ChangeLog:
2016-04-28 Eduard Sanou <dhole@openmailbox.org>
Matthias Klose <doko@debian.org>
* c-common.c (get_source_date_epoch): New function, gets the environment
variable SOURCE_DATE_EPOCH and parses it as long long with error
handling.
* c-common.h (get_source_date_epoch): Prototype.
* c-lex.c (c_lex_with_flags): set parse_in->source_date_epoch.
gcc/ChangeLog:
2016-04-28 Eduard Sanou <dhole@openmailbox.org>
Matthias Klose <doko@debian.org>
2016-04-28 Eduard Sanou <dhole@openmailbox.org>
Matthias Klose <doko@debian.org>
* include/cpplib.h (cpp_init_source_date_epoch): Prototype.
* init.c (cpp_init_source_date_epoch): New function.
* internal.h: Added source_date_epoch variable to struct
cpp_reader to store a reproducible date.
* macro.c (_cpp_builtin_macro_text): Set pfile->date timestamp from
pfile->source_date_epoch instead of localtime if source_date_epoch is
set, to be used for __DATE__ and __TIME__ macros to help reproducible
builds.
Patrick Palka [Wed, 27 Apr 2016 21:18:05 +0000 (21:18 +0000)]
Reduce nesting of parentheses in conditionals generated by genattrtab
gcc/ChangeLog:
* genattrtab.c (write_test_expr): New parameter EMIT_PARENS
which defaults to true. Emit an outer pair of parentheses only if
EMIT_PARENS. When continuing a chain of && or || (or & or |),
don't emit parentheses for the right-hand operand.
Ryan Burn [Wed, 27 Apr 2016 20:41:52 +0000 (20:41 +0000)]
re PR c++/69024 ([cilkpus] cilk_spawn is broken for initializations with implicit conversion operators defined)
PR c++/69024
PR c++/68997
* cilk.c (cilk_ignorable_spawn_rhs_op): Change to external linkage.
(cilk_recognize_spawn): Renamed from recognize_spawn and change to
external linkage.
(cilk_detect_and_unwrap): Corresponding changes.
(extract_free_variables): Don't extract free variables from
AGGR_INIT_EXPR slot.
* c-common.h (cilk_ignorable_spawn_rhs_op): Prototype.
(cilk_recognize_spawn): Likewise.
PR c++/69024
PR c++/68997
* cp-gimplify.c (cp_gimplify_expr): Call cilk_cp_detect_spawn_and_unwrap
instead of cilk_detect_spawn_and_unwrap.
* cp-cilkplus.c (is_conversion_operator_function_decl_p): New.
(find_spawn): New.
(cilk_cp_detect_spawn_and_unwrap): New.
* lambda.c: Include cp-cilkplus.h.
* parser.c: Include cp-cilkplus.h.
* cp-tree.h (cpp_validate_cilk_plus_loop): Move prototype into...
* cp-cilkpus.h: New file.
PR c++/69024
PR c++/68997
* g++.dg/cilk-plus/CK/pr68001.cc: Fix to not depend on broken
diagnostic.
* g++.dg/cilk-plus/CK/pr69024.cc: New test.
* g++.dg/cilk-plus/CK/pr68997.cc: New test.
Co-Authored-By: Jeff Law <law@redhat.com>
From-SVN: r235534
gcc/
* config/aarch64/aarch64.md
(*movhf_aarch64): Add "movi %0, #0" to zero up register and
remove the "fp" attributes.
(*movsf_aarch64): Add "movi %0, #0" to zero up register and
add the "simd" attributes.
(*movdf_aarch64): Likewise.
(*movtf_aarch64): Remove the "fp" attributes.
* testsuite/gcc.target/aarch64/fmovf-zero-reg.c: Update accordingly.
* testsuite/gcc.target/aarch64/fmovd-zero-reg.c: Likewise.
David Malcolm [Wed, 27 Apr 2016 18:18:45 +0000 (18:18 +0000)]
df: make df_problem instances "const"
The various struct df_problem instances are constant data; mark them
as such.
gcc/ChangeLog:
* df-core.c (df_add_problem): Make the problem param be const.
(df_remove_problem): Make local "problem" be const.
* df-problems.c (problem_RD): Make const.
(problem_LR): Likewise.
(problem_LIVE): Likewise.
(problem_MIR): Likewise.
(problem_CHAIN): Likewise.
(problem_WORD_LR): Likewise.
(problem_NOTE): Likewise.
(problem_MD): Likewise.
* df-scan.c (problem_SCAN): Likewise.
* df.h (struct df_problem): Make field "dependent_problem" be
const.
(struct dataflow): Likewise for field "problem".
(df_add_problem): Make param const.
i386.c (ix86_spill_class): Enable for TARGET_SSE2 when inter-unit moves to/from vector registers are enabled.
* config/i386/i386.c (ix86_spill_class): Enable for TARGET_SSE2 when
inter-unit moves to/from vector registers are enabled. Do not disable
for TARGET_MMX.
Eric Botcazou [Wed, 27 Apr 2016 18:08:39 +0000 (18:08 +0000)]
sem_aux.adb (Is_By_Reference_Type): Also return true for a tagged incomplete type without full view.
* sem_aux.adb (Is_By_Reference_Type): Also return true for a tagged
incomplete type without full view.
* sem_ch6.adb (Exchange_Limited_Views): Change into a function and
return the list of changes.
(Restore_Limited_Views): New procedure to undo the transformation made
by Exchange_Limited_Views.
(Analyze_Subprogram_Body_Helper): Adjust call to Exchange_Limited_Views
and call Restore_Limited_Views at the end, if need be.
(Possible_Freeze): Do not delay freezing because of incomplete types.
(Process_Formals): Remove kludges for class-wide types.
* types.h (By_Copy_Return): Delete.
* gcc-interface/ada-tree.h (TYPE_MAX_ALIGN): Move around.
(TYPE_DUMMY_IN_PROFILE_P): New macro.
* gcc-interface/gigi.h (update_profiles_with): Declare.
(finish_subprog_decl): Likewise.
(get_minimal_subprog_decl): Delete.
(create_subprog_type): Likewise.
(create_param_decl): Adjust prototype.
(create_subprog_decl): Likewise.
* gcc-interface/decl.c (defer_limited_with): Rename into...
(defer_limited_with_list): ...this.
(gnat_to_gnu_entity): Adjust to above renaming.
(finalize_from_limited_with): Likewise.
(tree_entity_vec_map): New structure.
(gt_pch_nx): New helpers.
(dummy_to_subprog_map): New hash table.
(gnat_to_gnu_param): Set the SLOC here. Remove MECH parameter and
add FIRST parameter. Deal with the mechanism here instead of...
Do not make read-only variant of types. Simplify expressions.
In the by-ref case, test the mechanism before must_pass_by_ref
and also TYPE_IS_BY_REFERENCE_P before building the reference type.
(gnat_to_gnu_subprog_type): New static function extracted from...
Do not special-case the type_annotate_only mode. Call
gnat_to_gnu_profile_type instead of gnat_to_gnu_type on return type.
Deal with dummy return types. Likewise for parameter types. Deal
with by-reference types explicitly and add a kludge for null procedures
with untagged incomplete types. Remove assertion on the types and be
prepared for multiple elaboration of the declarations. Skip the whole
CICO processing if the profile is incomplete. Handle the completion of
a previously incomplete profile.
(gnat_to_gnu_entity) <E_Variable>: Rename local variable.
Adjust couple of calls to create_param_decl.
<E_Access_Subprogram_Type, E_Anonymous_Access_Subprogram_Type>:
Remove specific deferring code.
<E_Access_Type>: Also deal with E_Subprogram_Type designated type.
Simplify handling of dummy types and remove obsolete comment.
Constify a couple of variables. Do not set TYPE_UNIVERSAL_ALIASING_P
on dummy types.
<E_Access_Subtype>: Tweak comment and simplify condition.
<E_Subprogram_Type>: ...here. Call it and clean up handling. Remove
obsolete comment and adjust call to gnat_to_gnu_param. Adjust call to
create_subprog_decl.
<E_Incomplete_Type>: Add a couple of 'const' qualifiers and get rid of
inner break statements. Tidy up condition guarding direct use of the
full view.
(get_minimal_subprog_decl): Delete.
(finalize_from_limited_with): Call update_profiles_with on dummy types
with TYPE_DUMMY_IN_PROFILE_P set.
(is_from_limited_with_of_main): Delete.
(associate_subprog_with_dummy_type): New function.
(update_profile): Likewise.
(update_profiles_with): Likewise.
(gnat_to_gnu_profile_type): Likewise.
(init_gnat_decl): Initialize dummy_to_subprog_map.
(destroy_gnat_decl): Destroy dummy_to_subprog_map.
* gcc-interface/misc.c (gnat_get_alias_set): Add guard for accessing
TYPE_UNIVERSAL_ALIASING_P.
(gnat_get_array_descr_info): Minor tweak.
* gcc-interface/trans.c (gigi): Adjust calls to create_subprog_decl.
(build_raise_check): Likewise.
(Compilation_Unit_to_gnu): Likewise.
(Identifier_to_gnu): Accept mismatches coming from a limited context.
(Attribute_to_gnu): Remove kludge for dispatch table entities.
(process_freeze_entity): Do not retrieve old definition if there is an
address clause on the entity. Call update_profiles_with on dummy types
with TYPE_DUMMY_IN_PROFILE_P set.
* gcc-interface/utils.c (build_dummy_unc_pointer_types): Also set
TYPE_REFERENCE_TO to the fat pointer type.
(create_subprog_type): Delete.
(create_param_decl): Remove READONLY parameter.
(finish_subprog_decl): New function extracted from...
(create_subprog_decl): ...here. Call it. Remove CONST_FLAG and
VOLATILE_FLAG parameters and adjust.
(update_pointer_to): Also clear TYPE_REFERENCE_TO in the unconstrained
case.
David Malcolm [Wed, 27 Apr 2016 17:54:42 +0000 (17:54 +0000)]
Fix comment in rtl.def
Commit r210360 removed the first "i" field from the various instruction
nodes in rtx.def, moving it to an explicit "int insn_uid;" field
of the union "u2" within rtx_def.
Update the comment in rtl.def to reflect this change. Also, fix
a stray apostrophe.
gcc/ChangeLog:
* rtl.def: Update comment for "things in the instruction chain" to
reflect the removal of the leading "i" field for INSN_UID in
r210360. Fix bogus apostrophe.
H.J. Lu [Wed, 27 Apr 2016 17:32:40 +0000 (17:32 +0000)]
Extend STV pass to 64-bit mode
128-bit SSE load and store instructions can be used for load and store
of 128-bit integers if they are the only operations on 128-bit integers.
To convert load and store of 128-bit integers to 128-bit SSE load and
store, the original STV pass, which is designed to convert 64-bit integer
operations to SSE2 operations in 32-bit mode, is extended to 64-bit mode
in the following ways:
1. Class scalar_chain is turned into base class. The 32-bit specific
member functions are moved to the new derived class, dimode_scalar_chain.
The new derived class, timode_scalar_chain, is added to convert oad and
store of 128-bit integers to 128-bit SSE load and store.
2. Add the 64-bit version of scalar_to_vector_candidate_p and
remove_non_convertible_regs. Only TImode load and store are allowed
for conversion. If one instruction on the chain of dependent
instructions aren't TImode load or store, the chain of instructions
won't be converted.
3. In 64-bit, we only convert from TImode to V1TImode, which have the
same size. The difference is only vector registers are allowed in
TImode so that 128-bit SSE load and store instructions will be used
for load and store of 128-bit integers.
4. Put the 64-bit STV pass before the CSE pass so that instructions
changed or generated by the STV pass can be CSEed.
convert_scalars_to_vector calls free_dominance_info in 64-bit mode to
work around ICE in fwprop pass:
DWARF: turn dw_loc_descr_node field into hash map for frame offset check
As discussed on
<https://gcc.gnu.org/ml/gcc-patches/2016-02/msg01708.html>, this change
removes a field in the dw_loc_descr_node structure so we can get rid of
the CHECKING_P macro usage.
This field was used to perform consistency checks for frame offset in
DWARF procedures. As a replacement, this commit turns the "visited
nodes" set in resolve_args_picking_1 into a map that remembers for each
dw_loc_descr_node the frame offset associated to it, so that the
consistency check is still operational.
Boostrapped and regtested on x86_64-linux.
2016-04-27 Pierre-Marie de Rodat <derodat@adacore.com>
* dwarf2out.h (struct dw_loc_descr_node): Remove the
dw_loc_frame_offset field.
* dwarf2out.c (new_loc_descr): Likewise.
(resolve_args_picking_1): Turn the VISITED hash set into a
FRAME_OFFSET hash map. Use it to associate a frame offset to
visited nodes. Remove uses of the CHECKING_P macro.
(resolve_args_picking): Update call to resolve_args_picking_1.
Richard Biener [Wed, 27 Apr 2016 14:10:04 +0000 (14:10 +0000)]
re PR ipa/70760 (wrong generated code for std::make_unique with -fipa-pta)
2016-04-27 Richard Biener <rguenther@suse.de>
PR ipa/70760
* tree-ssa-structalias.c (find_func_aliases_for_call): Use
aggregate_value_p to determine if a function result is
returned by reference.
(ipa_pta_execute): Functions having their address taken are
not automatically nonlocal.
* g++.dg/ipa/ipa-pta-2.C: New testcase.
* gcc.dg/ipa/ipa-pta-1.c: Adjust.
Jakub Jelinek [Wed, 27 Apr 2016 13:29:34 +0000 (15:29 +0200)]
re PR sanitizer/70683 (-fcompare-debug bug with -fsanitize=address)
PR sanitizer/70683
* tree-core.h (enum operand_equal_flag): Add OEP_NO_HASH_CHECK.
* fold-const.c (operand_equal_p): If flag_checking and
OEP_NO_HASH_CHECK is not set in flag, recurse with OEP_NO_HASH_CHECK
and if it returns non-zero, assert iterative_hash_expr on both
args is the same.
constexpr.c (get_fundef_copy): Use the original function for non-recursive evaluations.
cp/
* constexpr.c (get_fundef_copy): Use the original function for
non-recursive evaluations.
(save_fundef_copy): Always expect a slot to be available.
* sem_prag.adb (Analyze_Pragma): An object
renaming declaration resulting from the expansion of an object
declaration is a suitable context for pragma Ghost.
2016-04-27 Doug Rupp <rupp@adacore.com>
* init.c: Refine last checkin so the only requirement is the
signaling compilation unit is compiled with the same mode as
the compilation unit containing the initial landing pad.
2016-04-27 Ed Schonberg <schonberg@adacore.com>
* sem_ch13.adb (Valid_Default_Iterator): Better filter of illegal
specifications for Default_Iterator, including overloaded cases
where no interpretations are legal, and return types that are
not iterator types.
* exp_ch5.adb (Expand_N_Assignment_Statement): Do not install
an accessibility check when the left hand side of the assignment
denotes a container cursor.
* exp_util.ads, exp_util.adb (Find_Primitive_Operations): Removed.
* sem_ch4.adb (Find_Indexing_Operations): New routine.
(Try_Container_Indexing): Code cleanup.
* sem_ch13.adb (Analyze_Aspect_Specifications, case Pre/Post):
Check that the classwide version is illegal when the prefix is
an operation of an untagged synchronized type.
* ghost.adb (In_Subprogram_Body_Profile): New routine.
(Is_OK_Declaration): Treat an unanalyzed expression
function as an OK context. Treat a reference to a Ghost entity
as OK when it appears within the profile of a subprogram body.
2016-04-27 Bob Duff <duff@adacore.com>
* errout.ads: Document the fact that informational messages
don't have to be warnings.
* errout.adb (Error_Msg_Internal): In statistics counts, deal
correctly with informational messages that are not warnings.
(Error_Msg_NEL): Remove useless 'if' aroung Set_Posted, because
Set_Posted already checks for errors and ignores others.
* erroutc.adb (Prescan_Message): Set Is_Serious_Error to False
if Is_Info_Msg; the previous code was assuming that Is_Info_Msg
implies Is_Warning_Msg.
* errutil.adb (Error_Msg): In statistics counts, deal correctly
with informational messages that are not warnings.
* sem_util.ads, sem_util.adb (Is_Null_Record_Type): New predicate
to determine whether a record type is a null record.
* sem_ch3.adb (Analyze_Object_Declaration): If the type is a
null record and there is no expression in the declaration,
no predicate check applies to the object.
* sem_ch7.adb (Analyze_Package_Body_Helper): The body of an
instantiated package should not cause freezing of previous contracts.
2016-04-27 Ed Schonberg <schonberg@adacore.com>
* sem_dim.adb (Analyze_Dimension): Handle subtype declarations
that do not come from source.
(Analyze_Dimension_Subtype_Declaration): Allow confirming
dimensions on subtype entity, either inherited from base type
or provided by aspect specification.
2016-04-27 Ed Schonberg <schonberg@adacore.com>
* s-gearop.ads (Matrix_Vector_Solution, Matrix_Matrix_Solution):
Add scalar formal object Zero, to allow detection and report
when the matrix is singular.
* s-gearop.adb (Matrix_Vector_Solution, Matrix_Matrix_Solution):
Raise Constraint_Error if the Forward_Eliminate pass has
determined that determinant is Zero.o
* s-ngrear.adb (Solve): Add actual for Zero in corresponding
instantiations.
* s-ngcoar.adb (Solve): Ditto.
* sem_dim.adb (Analyze_Dimension, case N_Identifier): Check
that identifier has a usable type before analysis, to handle
properly identifiers introduced after some lexical/syntactic
recovery that created new identifiers.
* a-coinve.adb, a-comutr.adb, a-conhel.adb, a-convec.adb,
exp_util.adb: Remove assertions that can fail in obscure cases when
assertions are turned on but tampering checks are turned off.
2016-04-27 Javier Miranda <miranda@adacore.com>
* exp_ch6.adb (Add_Call_By_Copy_Code,
Add_Simple_Call_By_Copy_Code, Expand_Actuals): Handle formals
whose type comes from the limited view.
2016-04-27 Yannick Moy <moy@adacore.com>
* a-textio.adb: Complete previous patch.
2016-04-27 Yannick Moy <moy@adacore.com>
* inline.adb (Expand_Inlined_Call): Use Cannot_Inline instead of
Error_Msg_N to issue message about impossibility to inline call,
with slight change of message.
* exp_spark.adb (Expand_Potential_Renaming): Removed.
(Expand_SPARK): Update the call to expand a potential renaming.
(Expand_SPARK_Potential_Renaming): New routine.
* exp_spark.ads (Expand_SPARK_Potential_Renaming): New routine.
* sem.adb Add with and use clauses for Exp_SPARK.
(Analyze): Expand a non-overloaded potential renaming for SPARK.
2016-04-27 Ed Schonberg <schonberg@adacore.com>
* sem_ch3.adb (Constrain_Discriminated_Type): In an instance,
check full view for the presence of defaulted discriminants,
even when the partial view of a private type has no visible and
no unknown discriminants.
* sem_prag.adb (Process_Atomic_Independent_Shared_Volatile): Code
cleanup. Check the original node when trying to determine the node kind
of pragma Volatile's argument to account for untagged derivations
where the type is transformed into a constrained subtype.
2016-04-27 Olivier Hainque <hainque@adacore.com>
* mkdir.c (__gnat_mkdir): Rework the vxworks section to use a
consistent posix interface on the caller side.
2016-04-27 Ed Schonberg <schonberg@adacore.com>
* sem_ch10.adb (Build_Limited_View, Decorate_Type): If this
is a limited view of a type, initialize the Limited_Dependents
field to catch misuses of the type in a client unit.
2016-04-27 Thomas Quinot <quinot@adacore.com>
* a-strunb-shared.adb (Finalize): add missing Reference call.
* s-strhas.adb: minor grammar fix and extension of comment
* sem_ch8.adb: minor whitespace fixes
2016-04-27 Ed Schonberg <schonberg@adacore.com>
* lib-xref.adb (Get_Type_Reference): Handle properly the case
of an object declaration whose type definition is a class-wide
subtype and whose expression is a function call that returns a
classwide type.
* gnatcmd.adb: For "gnat ls -V -P", recognize switch
--unchecked-shared-lib-imports and set the flag
Opt.Unchecked_Shared_Lib_Imports accordingly.
2016-04-27 Ed Schonberg <schonberg@adacore.com>
* sem_prag.adb (Analyze_Pre_Post_Condition_In_Decl_Part):
A generic subprogram is never a primitive operation, and thus
a classwide condition for it is not legal.
* sem_res.adb (Flag_Effectively_Volatile_Objects): New routine.
(Resolve_Actuals): Flag effectively volatile objects with enabled
property Async_Writers or Effective_Reads as illegal.
* sem_util.adb (Is_OK_Volatile_Context): Comment reformatting.
2016-04-27 Javier Miranda <miranda@adacore.com>
* exp_ch3.adb (Make_Predefined_Primitive_Specs):
Do not generate the profile of the equality operator if it has
been explicitly defined as abstract in the parent type. Required
to avoid reporting an spurious error.
2016-04-27 Ed Schonberg <schonberg@adacore.com>
* sem_dim.ads, sem_dim.adb (Check_Expression_Dimensions): New
procedure to compute the dimension vector of a scalar expression
and compare it with the dimensions if its expected subtype. Used
for the ultimate components of a multidimensional aggregate,
whose components typically are themselves aggregates that are
expanded separately. Previous to this patch, dimensionality
checking on such aggregates generated spurious errors.
* sem_aggr.adb (Resolve_Array_Aggregate): Use
Check_Expression_Dimensions when needed.
2016-04-27 Javier Miranda <miranda@adacore.com>
* einfo.ads, einfo.adb (Corresponding_Function): New attribute
(applicable to E_Procedure).
(Corresponding_Procedure): New attribute (applicable to E_Function).
* exp_util.adb (Build_Procedure_Form): Link the function with
its internally built proc and viceversa.
* sem_ch6.adb (Build_Subprogram_Declaration): Propagate the
attribute Rewritten_For_C and Corresponding_Procedure to the body.
* exp_ch6.adb (Rewritten_For_C_Func_Id): Removed.
(Rewritten_For_C_Proc_Id): Removed.
* exp_unst.adb (Note_Uplevel_Ref): Use the new attribute to
locate the corresponding procedure.
* sem_ch13.adb (Analyze_Aspect_Export_Import): Signal that there is no
corresponding pragma.
2016-04-27 Bob Duff <duff@adacore.com>
* exp_ch3.adb: Minor comment improvement.
2016-04-27 Ed Schonberg <schonberg@adacore.com>
* exp_ch6.adb (Make_Build_In_Place_Call_In_Object_Declaration): If the
return type is an untagged limited record with only access
discriminants and no controlled components, the return value does not
need to use the secondary stack.
2016-04-27 Javier Miranda <miranda@adacore.com>
* exp_util.adb (Remove_Side_Effects): When
generating C code handle object declarations that have
discriminants and are initialized by means of a call to a
function.
2016-04-27 Ed Schonberg <schonberg@adacore.com>
* a-textio.adb (Get_Line function): Handle properly the case of
a line that has the same length as the buffer (or a multiple
thereof) and there is no line terminator.
* a-tigeli.adb (Get_Line procedure): Do not store an end_of_file
in the string when there is no previous line terminator and we
need at most one additional character.
* s-rident.ads: Make No_Implicit_Loops non partition wide.
2016-04-27 Arnaud Charlet <charlet@adacore.com>
* sem_ch11.adb (Analyze_Handled_Statements): check useless
assignments also in entries and task bodies, not only in
procedures and declaration blocks.
* sem_ch5.adb (Analyze_Block_Statement): check useless
assignements in declaration blocks as part of processing their
handled statement sequence, just like it was done for procedures
and now is also done for entries and task bodies.
* sem_warn.adb (Warn_On_Useless_Assignment): detect boundries
of entries and task bodies just like of procedures.
* sem_util.adb (Is_Volatile_Function): Recognize
a function declared within a protected type as well as the
protected/unprotected version of a function.
Bob Duff [Wed, 27 Apr 2016 12:20:54 +0000 (12:20 +0000)]
exp_ch3.adb (Expand_N_Object_Declaration): Rewrite an object declaration of the form "X ...
2016-04-27 Bob Duff <duff@adacore.com>
* exp_ch3.adb (Expand_N_Object_Declaration): Rewrite an object
declaration of the form "X : T := Func (...);", where T is
controlled, as a renaming.
* a-strunb-shared.adb (Finalize): Set the Unbounded_String Object
to be an empty string, instead of null-ing out the Reference.
* exp_util.adb (Needs_Finalization): Remove redundant code.
Nick Clifton [Wed, 27 Apr 2016 11:29:20 +0000 (11:29 +0000)]
re PR go/49889 (Calling a function whose name is obscured by a local variable does not produce an error)
PR middle-end/49889
gcc * varasm.c (merge_weak): Generate an error if an attempt is made
to convert a non-weak static function into a weak, public function.
* aspects.ads Aspects Export and Import do not require delay. They
were classified as delayed aspects, but treated as non-delayed
by the analysis of aspects.
* freeze.adb (Copy_Import_Pragma): New routine.
(Wrap_Imported_Subprogram): Copy the import pragma by first
resetting all semantic fields to avoid an infinite loop when
performing the copy.
* sem_ch13.adb (Analyze_Aspects_At_Freeze_Point): Add
comment on the processing of aspects Export and Import
at the freeze point.
(Analyze_Aspect_Convention: New routine.
(Analyze_Aspect_Export_Import): New routine.
(Analyze_Aspect_External_Link_Name): New routine.
(Analyze_Aspect_External_Or_Link_Name): Removed.
(Analyze_Aspect_Specifications): Factor out the analysis of
aspects Convention, Export, External_Name, Import, and Link_Name
in their respective routines. Aspects Export and Import should
not generate a Boolean pragma because their corresponding pragmas
have a very different syntax.
(Build_Export_Import_Pragma): New routine.
(Get_Interfacing_Aspects): New routine.
2016-04-27 Eric Botcazou <ebotcazou@adacore.com>
* inline.adb (Add_Inlined_Body): Overhaul implementation,
robustify handling of -gnatn1, add special treatment for
expression functions.
2016-04-27 Doug Rupp <rupp@adacore.com>
* g-traceb.ads: Update comment.
* exp_ch2.adb: minor style fix in object declaration
* sem_elab.adb (Check_Internal_Call): Do not
consider a call when it appears within pragma Initial_Condition
since the pragma is part of the elaboration statements of a
package body and may only call external subprograms or subprograms
whose body is already available.
(Within_Initial_Condition): New routine.
2016-04-27 Ed Schonberg <schonberg@adacore.com>
* exp_util.adb (Build_Procedure_Form): Prevent double generation
of the procedure form when dealing with an expression function
whose return type is an array.
* sem_ch3.adb: Fix out-of order Has_Predicates setting.
* exp_ch6.adb: Proper conversion for inherited operation in C.
* sem_ch6.adb: Code cleanup.
* lib-xref.ads, sem_ch10.adb: minor style fix in comment
* g-socket.adb: Minor reformatting.
* sinfo.ads: Minor comment correction.
* sem_warn.ads: minor grammar fix in comment
* params.def (MAX_PARTITION_SIZE): New param.
* invoke.texi: Document lto-max-partition.
lto/
* lto-partition.h (lto_balanced_map): New parameter.
* lto-partition.c (lto_balanced_map): New parameter
max_partition_size.
Check if partition size is greater than max_partition_size.
* lto.c (do_whole_program_analysis): Adjust calls to
lto_balanced_map() to pass 2nd argument.
Richard Biener [Wed, 27 Apr 2016 10:42:08 +0000 (10:42 +0000)]
re PR ipa/70785 (LTO bootstrap with IPA PTA is broken)
2016-04-27 Richard Biener <rguenther@suse.de>
PR ipa/70785
* tree-ssa-structalias.c (refered_from_nonlocal_fn): New
function cummulating used_from_other_partition, externally_visible
and force_output from aliases.
(refered_from_nonlocal_var): Likewise.
(ipa_pta_execute): Use call_for_symbol_and_aliases to cummulate
node flags properly.
Jakub Jelinek [Wed, 27 Apr 2016 07:45:57 +0000 (09:45 +0200)]
re PR sanitizer/70683 (-fcompare-debug bug with -fsanitize=address)
PR sanitizer/70683
* tree.h (inchash::add_expr): Add FLAGS argument.
* tree.c (inchash::add_expr): Likewise. If not OEP_ADDRESS_OF,
use STRIP_NOPS first. For INTEGER_CST assert not OEP_ADDRESS_OF.
For REAL_CST and !HONOR_SIGNED_ZEROS (t) hash +/- 0 the same.
Formatting fix. Adjust recursive calls. For tcc_comparison,
if swap_tree_comparison (code) is smaller than code, hash that
and arguments in the other order. Hash CONVERT_EXPR the same
as NOP_EXPR. For OEP_ADDRESS_OF hash MEM_REF with 0 offset
of ADDR_EXPR of decl as the decl itself. Add or remove
OEP_ADDRESS_OF from recursive flags as needed. For
FMA_EXPR, WIDEN_MULT_{PLUS,MINUS}_EXPR hash the first two
operands commutatively and only the third one normally.
For internal CALL_EXPR hash in CALL_EXPR_IFN.
Patrick Palka [Tue, 26 Apr 2016 22:24:43 +0000 (22:24 +0000)]
Fix PR c++/70241 (inconsistent access with in-class enumeration)
gcc/cp/ChangeLog:
PR c++/70241
* decl.c (build_enumerator): Set current_access_specifier when
declaring an enumerator belonging to an in-class enumeration.
* parser.c (cp_parser_check_access_in_redecleration): Also
consider in-class enumerations.
gcc/testsite/ChangeLog:
PR c++/70241
* g++.dg/cpp0x/enum32.C: New test.
* g++.dg/cpp0x/enum33.C: New test.
Ian Lance Taylor [Tue, 26 Apr 2016 20:37:58 +0000 (20:37 +0000)]
compiler: Add Enclosed_var_expression.
Introduces an abstraction for a variable referenced in a closure.
This maintains the underlying expression which accesses a field within
a closure variable and gives easy access to the underlying
Named_object.
Marek Polacek [Tue, 26 Apr 2016 14:59:40 +0000 (14:59 +0000)]
re PR c/67784 (Incorrect parsing when using declarations in for loops and typedefs)
PR c/67784
* c-parser.c (c_parser_maybe_reclassify_token): New function factored
out of ...
(c_parser_for_statement): ... here.
(c_parser_if_statement): Use it.
(c_parser_switch_statement): Use it.
(c_parser_while_statement): Use it.
* gcc.dg/pr67784-3.c: New test.
* gcc.dg/pr67784-4.c: New test.
* gcc.dg/pr67784-5.c: New test.
projects / thirdparty / gcc.git / log
