2014-01-25 Walter Lee <walt@tilera.com>
* config/tilegx/tilegx.c (tilegx_gen_bundles): Delete barrier
insns after bundles have been formed.
* config/tilegx/tilegx.md (tile_network_barrier): Update
comment.
From-SVN: r207081
2014-01-25 Walter Lee <walt@tilera.com>
* config/tilegx/tilegx.c (tilegx_expand_builtin): Set
PREFETCH_SCHEDULE_BARRIER_P to true for prefetches.
* config/tilepro/tilepro.c (tilepro_expand_builtin): Ditto.
From-SVN: r207080
gcc/
* config/mips/constraints.md (kl): Delete.
* config/mips/mips.md (divmod<mode>4, udivmod<mode>4): Turn into
define expands, using...
(divmod<mode>4_mips16, udivmod<mode>4_mips16): ...these new
instructions for MIPS16.
(*divmod<mode>4, *udivmod<mode>4): New patterns, taken from the
non-MIPS16 version of the old divmod<mode>4 and udivmod<mode>4.
From-SVN: r207079
* gcc-interface/utils.c (convert_to_fat_pointer): Un-obfuscate the
conversion from a thin pointer with a shifted value.
* gcc-interface/utils2.c (gnat_build_constructor): Propagate the
read-only flag from the values onto the result.
(gnat_invariant_expr): Accept read-only CONSTRUCTORs.
From-SVN: r207073
* print-rtl.c (in_call_function_usage): New var.
(print_rtx): When in CALL_INSN_FUNCTION_USAGE, always print
EXPR_LIST mode as mode and not as reg note name.
From-SVN: r207066
PR middle-end/59561
* cfgloopmanip.c (copy_loop_info): If
loop->warned_aggressive_loop_optimizations, make sure
the flag is set in target loop too.
From-SVN: r207065
2014-01-24 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
* config/rs6000/rs6000.c (rs6000_expand_vec_perm_const_1): Remove
correction for little endian...
* config/rs6000/vsx.md (vsx_xxpermdi2_<mode>_1): ...and move it to
here.
From-SVN: r207062
PR tree-optimization/59919
* tree-vrp.c (find_assert_locations_1): Do not register asserts
for non-returning calls.
PR tree-optimization/59919
* gcc.c-torture/compile/pr59919.c: New test.
From-SVN: r207061
2014-01-24 Ed Smith-Rowland <3dw4rd@verizon.net>
PR libstdc++/59531
* testsuite/experimental/string_view/operations/copy/char/1.cc: New.
* testsuite/experimental/string_view/operations/copy/wchar_t/1.cc: New.
2014-01-24 Ed Smith-Rowland <3dw4rd@verizon.net>
Peter A. Bigot <pab@pabigot.com>
PR libstdc++/59531
* include/experimental/string_view
(copy(_CharT*, size_type, size_type) const): Correct throw string.
Correct copy start location.
2014-01-24 Ed Smith-Rowland <3dw4rd@verizon.net>
Peter A. Bigot <pab@pabigot.com>
PR libstdc++/59530
* include/experimental/string_view (operator[](size_type) const):
Fix one-off index error in debug check.
* testsuite/experimental/string_view/element_access/char/1.cc: Don't
test basic_string_view at size().
* testsuite/experimental/string_view/element_access/wchar_t/1.cc: Ditto.
2014-01-24 Ed Smith-Rowland <3dw4rd@verizon.net>
Peter A. Bigot <pab@pabigot.com>
PR libstdc++/59529
* include/experimental/string_view
(basic_string_view(const _CharT*, size_type)): Don't care if len == 0.
* testsuite/experimental/string_view/operations/substr/char/1.cc:
Comment out catch of out_of_range; No terminating null
in basic_string_view. Check begin == end.
* testsuite/experimental/string_view/operations/substr/wchar_t/1.cc:
Ditto.
Co-Authored-By: Peter A. Bigot <pab@pabigot.com>
From-SVN: r207060
PR libstdc++/59548
* include/debug/safe_base.h (_Safe_sequence_base): Define copy
constructor to prevent it being implicitly defined as deleted, but
do not copy anything.
* include/debug/safe_unordered_base.h (_Safe_unordered_container_base):
Define copy and move constructors similar to _Safe_sequence_base's.
* testsuite/23_containers/unordered_map/59548.cc: New.
From-SVN: r207059
2014-01-24 Eric Botcazou <ebotcazou@adacore.com>
* set_targ.adb: Set Short_Enums.
* gcc-interface/lang.opt (fshort-enums): New option.
* gcc-interface/misc.c (gnat_handle_option): Handle it.
(gnat_post_options): Do not modify the global settings.
2014-01-24 Robert Dewar <dewar@adacore.com>
* g-rannum.ads, g-rannum.adb (Random_Ordinary_Fixed): New generic
function.
(Random_Decimal_Fixed): New generic function.
* s-rannum.ads: Minor comment clarifications.
From-SVN: r207049
2014-01-24 Robert Dewar <dewar@adacore.com>
* back_end.adb: Remove Short_Enums handling (handled in
Ttypes/Get_Targ now) Minor added comments.
* freeze.adb: Change name Short_Enums_On_Target to
Target_Short_Enums.
* get_targ.ads, get_targ.adb (Get_Short_Enums): New function.
* opt.ads: Minor comment updates.
* sem_ch13.adb: Change name Short_Enums_On_Target to
Target_Short_Enums.
* set_targ.adb: Set Short_Enums from gcc back end.
* set_targ.ads (Short_Enums): New variable.
* targparm.ads, targparm.adb: Remove Short_Enums entries (handled in
Ttypes/Get_Targ now).
* ttypes.ads (Target_Short_Enums): New constant boolean switch
From-SVN: r207045
2014-01-24 Gary Dismukes <dismukes@adacore.com>
* exp_disp.adb (Expand_Dispatching_Call): Call Unqualify on Param when
comparing it with Ctrl_Arg, since Ctrl_Arg may have had qualification
stripped off.
From-SVN: r207043
2014-01-24 Robert Dewar <dewar@adacore.com>
* sinfo.ads, make.adb, prj-env.adb: Minor reformatting.
2014-01-24 Vincent Celier <celier@adacore.com>
* prj.adb (Add_Aggregated_Project): Do not add a project in
the list if it is already there.
2014-01-24 Yannick Moy <moy@adacore.com>
* lib-xref-spark_specific.adb (Enclosing_Subprogram_Or_Package):
Correct the search for a subrogram declaration to which a pragma is
attached.
2014-01-24 Bob Duff <duff@adacore.com>
* gnat_ugn.texi: Document --decimal-grouping and
--based-grouping switches in gnatpp.
From-SVN: r207042
2014-01-24 Ed Schonberg <schonberg@adacore.com>
* sinfo.ads: Documentation update.
2014-01-24 Ed Schonberg <schonberg@adacore.com>
* sem_ch3.adb (Constant_Redeclaration): New declaration is
illegal if previous one has an initial expression that is an
aggregate expanded into assignments.
2014-01-24 Ed Schonberg <schonberg@adacore.com>
* sem_ch5.adb (Analyze_Loop_Parameter_Specification): Small
code reorganization to remove spurious warning on a loop with
an array element iterator that has a null range.
From-SVN: r207041
2014-01-24 Ed Schonberg <schonberg@adacore.com>
* sem_ch3.adb (Analyze_Declarations): At the end of an
appropriate declarative part, call Freeze_All from the first
declaration in the scope, not from the first unfrozen one. This
is necessary to apply visibility checks to entities with delayed
aspects. Otherwise, in the presence of instantiations and cleanups
that they may generate, the delayed aspects may be analyzed too
late and produce spurious visibility errors.
* sem_attr.adb: Place etype on range.
* sem_ch6.adb: Documentation expression functions.
From-SVN: r207039
2014-01-24 Robert Dewar <dewar@adacore.com>
* exp_ch7.adb: Minor change of Indices to Indexes (preferred
terminology in compiler).
2014-01-24 Robert Dewar <dewar@adacore.com>
* scans.ads: Remove Tok_Raise from Sterm, Eterm, After_SM
categories, now that Ada 95 supports raise expressions.
2014-01-24 Robert Dewar <dewar@adacore.com>
* freeze.adb (Freeze_Enumeration_Type): Use new target parameter
Short_Enums_On_Target.
* sem_ch13.adb (Set_Enum_Esize): Take Short_Enums_On_Target
into account.
* targparm.ads, targparm.adb: Add new target parameter Short_Enums.
2014-01-24 Ed Schonberg <schonberg@adacore.com>
* sem_ch5.adb (Analyze_Iterator_Specification): If subtype
indication is given explicity, check that it matches the array
component type or the container element type of the domain
of iteration.
2014-01-24 Tristan Gingold <gingold@adacore.com>
* back_end.adb (Scan_Compiler_Arguments): Set Short_Enums_On_Target.
2014-01-24 Vincent Celier <celier@adacore.com>
* prj-env.adb (Ada_Objects_Path): Use Ada_Objects_Path_No_Libs
to cache the result when Including_Libraries is False.
* prj-env.ads (Ada_Objects_Path): Update documentation
* prj.adb (Free (Project_Id)): Also free Ada_Objects_Path_No_Libs
(Get_Object_Directory): Return the Library_Ali_Dir only when
when Including_Libraries is True.
* prj.ads (Get_Object_Directory): Fix and complete documentation
(Project_Data): New component Ada_Objects_Path_No_Libs
From-SVN: r207036
2014-01-24 Robert Dewar <dewar@adacore.com>
* checks.adb (Expr_Known_Valid): Result of fpt operator never
considered valid.
2014-01-24 Eric Botcazou <ebotcazou@adacore.com>
* back_end.adb: Minor fix in comment.
2014-01-24 Javier Miranda <miranda@adacore.com>
* sem_ch3.adb (Check_Abstract_Overriding): Code reestructuration
required to report the error in case of task types.
2014-01-24 Ed Schonberg <schonberg@adacore.com>
* sem_attr.adb: Additional index checking.
From-SVN: r207035
2014-01-24 Ed Schonberg <schonberg@adacore.com>
* sem_attr.adb (Analyze_Attribute, case 'Update): Analyze
expressions in each component association, and for records note
the entity in each association choice, for subsequent resolution.
(Resolve_Attribute, case 'Update): Complete resolution of
expressions in each component association.
2014-01-24 Robert Dewar <dewar@adacore.com>
* sem.adb (Sem): Avoid premature reference to Current_Sem_Unit
(this was causing Is_Main_Unit_Or_Main_Unit_Spec to be set wrong,
leading to wrong handling of SPARK_Mode for library units).
2014-01-24 Robert Dewar <dewar@adacore.com>
* sem_ch6.adb (Analyze_Subprogram_Body_Helper): Set SPARK_Mode
on generic instances (do not consider them to be internally
generated)
2014-01-24 Doug Rupp <rupp@adacore.com>
* s-osinte-android.ads (pthread_sigmask): Import sigprocmask
vice pthread_sigmask.
2014-01-24 Vincent Celier <celier@adacore.com>
* prj.adb (Debug_Output (Str, Str2)): Output if verbosity is
not default.
2014-01-24 Vincent Celier <celier@adacore.com>
* prj-ext.adb (Add): Do not output anything when Silent is True,
whatever the verbosity. When Source is From_External_Attribute,
set the corresponding environment variable if it is not already set.
* prj-ext.ads (Add): New Boolean parameter Silent, defaulted
to False
* prj-proc.adb (Process_Expression_For_Associative_Array):
For attribute External, call Prj.Ext.Add with Silent set to
True for the child environment, to avoid useless output in non
default verbosity.
2014-01-24 Ed Schonberg <schonberg@adacore.com>
* sem_res.adb (Set_Slice_Subtype): Handle properly a discrete
range given by a subtype indication, and force evaluation of
the bounds, as for a simple range.
* exp_util.adb (Evaluate_Slice_Bounds): Utility to force evaluation
of bounds of slice for various kinds of discrete ranges.
(Evaluate_Name, Evaluate_Subtype_From_Expr): use
Evaluate_Slice_Bounds.
2014-01-24 Bob Duff <duff@adacore.com>
* s-taskin.ads (Activator): Make this Atomic, because
Activation_Is_Complete reads it, and that can be called
from any task. Previously, this component was only
modified by the activator before activation, and by
Self after activation.
* a-taside.ads, a-taside.adb (Environment_Task,
Activation_Is_Complete): Implement these missing functions.
From-SVN: r207034
2014-01-24 Doug Rupp <rupp@adacore.com>
* init.c: Add a handler section for Android.
2014-01-24 Arnaud Charlet <charlet@adacore.com>
* i-cexten.ads (Unsigned_33..64, Unsigned_33..64): New types.
2014-01-24 Ed Schonberg <schonberg@adacore.com>
* sem_ch4.adb (Operator_Check): If one operand is a
Raise_Expression, set its type to that of the other operand.
* sem_res.adb (Resolve_Raise_Expression): new procedure.
(Resolve_Actuals): For an actual that is a Raise_Expression,
set the type to that of the formal.
* sem_type.adb (Find_Unique_Type): If one of the operands is a
Raise_Expression, return type of the other operand.
2014-01-24 Ed Schonberg <schonberg@adacore.com>
* sem_aggr.adb (Resolve_Record_Aggregate): If a scalar
component of the record has a type with a default aspect, and
the corresponding aggregate component is initiaized with a box,
use the default value in the rewritten aggregate.
2014-01-24 Tristan Gingold <gingold@adacore.com>
* s-interr.ads, s-interr.adb, s-interr-hwint.adb, s-interr-vms.adb,
s-interr-sigaction.adb,
s-interr-dummy.adb (Install_Restricted_Handlers): Add Prio parameter.
* exp_ch9.adb (Make_Initialize_Protection): Add Prio parameter
to the call to Install_Restricted_Handlers.
2014-01-24 Emmanuel Briot <briot@adacore.com>
* prj-nmsc.adb (Check_File): Add protection when the source is
not fully initialized.
From-SVN: r207033
2014-01-24 Ed Schonberg <schonberg@adacore.com>
* sem_util.adb (Is_Post_State): In a postcondition, a selected
component that denotes an implicit dereference is a reference
to the post state of the subprogram.
2014-01-24 Robert Dewar <dewar@adacore.com>
* sem_ch6.adb (Analyze_Subprogram_Body_Helper): SPARK_Mode OFF
for generated subprograms.
(Analyze_Subprogram_Specification): Ditto.
2014-01-24 Vincent Celier <celier@adacore.com>
* prj-dect.adb (Check_Attribute_Allowed): Detect more forbidden
attributes in package Builder of aggregate and aggregate library
projects.
* prj-nmsc.adb (Process_Naming_Scheme.Check.Check_Aggregate):
Remove procedure (Process_Naming_Scheme.Check.Check_Aggregated):
Remove parameters. Change error message from "... externally
build library ..." to "... externally built project ...".
(Process_Naming_Scheme.Check): Do not do any check in aggregate
project, as attribute Library_Dir and Library_Name have already
been detected as forbidden.
2014-01-24 Vincent Celier <celier@adacore.com>
* prj-env.adb (Find_Project): If cached project path is not in
project directory, look in current directory first and use cached
project path only if project is not found in project directory.
From-SVN: r207032
2014-01-24 Vincent Celier <celier@adacore.com>
* projects.texi: Document that aggregate projects are only
supported by gprbuild, but not by gnatmake. Document that the
only attribute Switches in package Binder of aggregate projects
will be ignored if its index is not others. Document that
attribute Global_Config_File is allowed in package Binder of
aggregate projects.
2014-01-24 Robert Dewar <dewar@adacore.com>
* sem_prag.adb: Minor code reorganization.
* sem_util.adb: Minor fix of potential latent bug in Is_LHS.
From-SVN: r207029
2014-01-24 Pascal Obry <obry@adacore.com>
* prj-attr.adb, projects.texi, snames.ads-tmpl: Add Excluded_Patterns
attribute definition.
2014-01-24 Vincent Celier <celier@adacore.com>
* makeutl.adb (Queue.Insert_No_Roots): In gprbuild, do not put
in the Queue the same source (same path, same multi-source index)
from the same project file, to avoid compiling several times
the same source.
2014-01-24 Eric Botcazou <ebotcazou@adacore.com>
* einfo.ads (First_Rep_Item): Remove obsolete stuff.
(Has_Gigi_Rep_Item): Likewise.
* sem_prag.adb (Analyze_Pragma) <Pragma_Linker_Section>: Do not set
Has_Gigi_Rep_Item for objects.
* gcc-interface/decl.c (prepend_one_attribute_to): Rename into...
(prepend_one_attribute): ...this.
(prepend_one_attribute_pragma): New function extracted from...
(prepend_attributes): ...here. Swap the parameters for consistency.
(gnat_to_gnu_entity): Adjust calls to prepend_one_attribute_to and to
prepend_attributes.
<object>: Deal with a pragma Linker_Section on a constant
or variable. <E_Function>: Deal with a pragma Linker_Section
on a subprogram.
(get_minimal_subprog_decl): Adjust calls to prepend_one_attribute_to.
2014-01-24 Vincent Celier <celier@adacore.com>
* opt.ads: Minor comment update.
From-SVN: r207028
pushsf for -m64/-mx32 looks like
(set (mem:SF (pre_modify:SI (reg/f:SI 7 sp)
(plus:SI (reg/f:SI 7 sp)
(const_int -8))))
(reg:SF 22 xmm1 [orig:84 D.1790 ] [84]))
Stack adjustment is in push operand and it isn't stack register mode size
which may be 4 bytes for -mx32. This patch gets stack adjustment from
push operand if code of push isn't PRE_DEC.
gcc/
PR target/59929
* config/i386/i386.md (pushsf splitter): Get stack adjustment
from push operand if code of push isn't PRE_DEC.
gcc/testsuite/
PR target/59929
* gcc.target/i386/pr59929.c: New test.
From-SVN: r207023
[gcc]
2014-01-23 Michael Meissner <meissner@linux.vnet.ibm.com>
PR target/59909
* doc/invoke.texi (RS/6000 and PowerPC Options): Document
-mquad-memory-atomic. Update -mquad-memory documentation to say
it is only used for non-atomic loads/stores.
* config/rs6000/predicates.md (quad_int_reg_operand): Allow either
-mquad-memory or -mquad-memory-atomic switches.
* config/rs6000/rs6000-cpus.def (ISA_2_7_MASKS_SERVER): Add
-mquad-memory-atomic to ISA 2.07 support.
* config/rs6000/rs6000.opt (-mquad-memory-atomic): Add new switch
to separate support of normal quad word memory operations (ldq,
stq) from the atomic quad word memory operations.
* config/rs6000/rs6000.c (rs6000_option_override_internal): Add
support to separate non-atomic quad word operations from atomic
quad word operations. Disable non-atomic quad word operations in
little endian mode so that we don't have to swap words after the
load and before the store.
(quad_load_store_p): Add comment about atomic quad word support.
(rs6000_opt_masks): Add -mquad-memory-atomic to the list of
options printed with -mdebug=reg.
* config/rs6000/rs6000.h (TARGET_SYNC_TI): Use
-mquad-memory-atomic as the test for whether we have quad word
atomic instructions.
(TARGET_SYNC_HI_QI): If either -mquad-memory-atomic,
-mquad-memory, or -mp8-vector are used, allow byte/half-word
atomic operations.
* config/rs6000/sync.md (load_lockedti): Insure that the address
is a proper indexed or indirect address for the lqarx instruction.
On little endian systems, swap the hi/lo registers after the lqarx
instruction.
(load_lockedpti): Use indexed_or_indirect_operand predicate to
insure the address is valid for the lqarx instruction.
(store_conditionalti): Insure that the address is a proper indexed
or indirect address for the stqcrx. instruction. On little endian
systems, swap the hi/lo registers before doing the stqcrx.
instruction.
(store_conditionalpti): Use indexed_or_indirect_operand predicate to
insure the address is valid for the stqcrx. instruction.
* gcc/config/rs6000/rs6000-c.c (rs6000_target_modify_macros):
Define __QUAD_MEMORY__ and __QUAD_MEMORY_ATOMIC__ based on what
type of quad memory support is available.
[gcc/testsuite]
2014-01-23 Michael Meissner <meissner@linux.vnet.ibm.com>
PR target/59909
* gcc.target/powerpc/quad-atomic.c: New file to test power8 quad
word atomic functions at runtime.
From-SVN: r207020
