aboutsummaryrefslogtreecommitdiff
path: root/gcc
AgeCommit message (Collapse)AuthorFilesLines
2016-04-21Regenerate .pot files.Joseph Myers2-2953/+2962
gcc/po: * gcc.pot: Regenerate. libcpp/po: * cpplib.pot: Regenerate. From-SVN: r235342
2016-04-21re PR tree-optimization/70725 (Internal compiler error (ICE) on valid code)Richard Biener2-10/+28
2016-04-21 Richard Biener <rguenther@suse.de> PR tree-optimization/70725 * tree-if-conv.c (if_convertible_phi_p): Adjust guard for phi_convertible_by_degenerating_args. (predicate_all_scalar_phis): Handle single-argument PHIs. From-SVN: r235341
2016-04-21re PR middle-end/70747 (ICE on valid code on x86_64-linux-gnu: verify_gimple ↵Richard Biener4-2/+23
failed) 2016-04-21 Richard Biener <rguenther@suse.de> PR middle-end/70747 * fold-const.c (fold_comparison): Return properly typed constant boolean. * gcc.dg/pr70747.c: New testcase. From-SVN: r235335
2016-04-21re PR tree-optimization/70715 (SCEV failed to prove no-overflow-ness ↵Bin Cheng4-1/+31
information unsigned loop IV) PR tree-optimization/70715 * tree-ssa-loop-niter.c (loop_exits_before_overflow): Check equality after expanding BASE using expand_simple_operations. gcc/testsuite/ChangeLog PR tree-optimization/70715 * gcc.dg/tree-ssa/scev-13.c: New test. From-SVN: r235333
2016-04-21min(-x, -y), min(~x, ~y)Marc Glisse4-0/+35
2016-04-21 Marc Glisse <marc.glisse@inria.fr> gcc/ * match.pd (min(-x, -y), max(-x, -y), min(~x, ~y), max(~x, ~y)): New transformations. gcc/testsuite/ * gcc.dg/tree-ssa/minmax-2.c: New testcase. From-SVN: r235332
2016-04-21[multiple changes]Arnaud Charlet5-8/+30
2016-04-21 Hristian Kirtchev <kirtchev@adacore.com> * sem_prag.adb (Match_Constituent): Treat a constant as a legal constituent even if it is not to prevent spurious errors. 2016-04-21 Gary Dismukes <dismukes@adacore.com> * sem_ch4.adb: Minor typo fixes and reformatting. 2016-04-21 Dmitriy Anisimkov <anisimko@adacore.com> * g-calend.ads (No_Time): The same value in any timezone. * g-socket.adb (Raise_Host_Error): Remove ending dot from original error message before append colon delimited host name. From-SVN: r235331
2016-04-21[multiple changes]Arnaud Charlet9-89/+261
2016-04-21 Hristian Kirtchev <kirtchev@adacore.com> * sem_ch3.adb: Code cleanup. * sem_ch6.adb: Code cleanup. (Is_Matching_Limited_View): New routine. (Matches_Limited_With_View): Reimplemented. * sem_ch10.adb (Decorate_Type): Code cleanup. 2016-04-21 Doug Rupp <rupp@adacore.com> * tracebak.c (PPC ELF): Add macro defs for lynxos178e. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * sem_ch4.adb (Try_Container_Indexing): If there are overloaded indexing functions, collect all overloadings of the call firts, and then transfer them to indexing node, to prevent interleaving of the set of interpretations of the nodes involved. * sem_res.adb (Resolve): Suppress cascaded errors that report ambiguities when one of the actuals in an overloaded generatlized indexing operation is illegal and has type Any_Type, as is done for similar cascaded errors in subprogram calls. (Valid_Tagged_Conversion): Cleanup conversion checks when one of the types involved is a class-wide subtype. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * sem_ch12.adb (Load_Parent_Of_Generic): When looking for the subprogram declaration within a wrapper package, skip pragmas that may have been generated by aspect specifications on the generic instance. 2016-04-21 Javier Miranda <miranda@adacore.com> * exp_aggr.adb (Component_Not_OK_For_Backend): Generating C code return True for array identifiers since the backend needs to initialize such component by means of memcpy(). From-SVN: r235330
2016-04-21a-tasatt.adb, [...] (Fast_Path): Rewritten to avoid reading potentially ↵Arnaud Charlet4-8/+23
uninitialized memory. 2016-04-21 Arnaud Charlet <charlet@adacore.com> * a-tasatt.adb, a-tasatt.ads (Fast_Path): Rewritten to avoid reading potentially uninitialized memory. * sem_ch3.adb: Minor style fix in comment. From-SVN: r235329
2016-04-21gnat_rm.texi, [...]: Update doc.Arnaud Charlet14-919/+1130
* gnat_rm.texi, gnat_ugn.texi, doc/gnat_ugn/gnat_project_manager.rst, doc/gnat_ugn/building_executable_programs_with_gnat.rst, doc/gnat_ugn/gnat_and_program_execution.rst, doc/gnat_ugn/gnat_utility_programs.rst, doc/gnat_ugn/the_gnat_compilation_model.rst, doc/gnat_rm/implementation_defined_attributes.rst, doc/gnat_rm/standard_and_implementation_defined_restrictions.rst, doc/gnat_rm/implementation_defined_pragmas.rst, doc/gnat_rm/the_gnat_library.rst, doc/gnat_rm/implementation_defined_aspects.rst: Update doc. * doc/Makefile: Cleanups. From-SVN: r235328
2016-04-21[multiple changes]Arnaud Charlet5-45/+98
2016-04-21 Hristian Kirtchev <kirtchev@adacore.com> * exp_pakd.adb, sem_ch13.adb: Minor reformatting. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * exp_dbug.adb, exp_dbug.ads (Qualify_Entity_Name): Add suffixes to disambiguate local variables that may be hidden from inner visibility by nested block declarations or loop variables. From-SVN: r235327
2016-04-21[multiple changes]Arnaud Charlet16-323/+629
2016-04-21 Jerome Lambourg <lambourg@adacore.com> * s-soflin.adb: Initialize the Stack_Limit global variable. 2016-04-21 Hristian Kirtchev <kirtchev@adacore.com> * lib-writ.adb: Minor reformatting. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * exp_pakd.adb (Compute_Number_Components): New function to build an expression that computes the number of a components of an array that may be multidimensional. (Expan_Packed_Eq): Use it. 2016-04-21 Arnaud Charlet <charlet@adacore.com> * g-traceb.ads: Update list of supported platforms. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * sem_ch13.adb (Add_Predicates): if the type is declared in an inner package it may be frozen outside of the package, and the generated pragma has not been analyzed yet, the expression for the predicate must be captured and added to the predicate function at this point. 2016-04-21 Hristian Kirtchev <kirtchev@adacore.com> * contracts.adb (Analyze_Package_Body_Contract): Do not check for a missing package refinement because 1) packages do not have "refinement" and 2) the check for proper state refinement is performed in a different place. * einfo.adb (Has_Non_Null_Visible_Refinement): Reimplemented. (Has_Null_Visible_Refinement): Reimplemented. * sem_ch3.adb (Analyze_Declarations): Determine whether all abstract states have received a refinement and if not, emit errors. * sem_ch7.adb (Analyze_Package_Declaration): Code cleanup. Determine whether all abstract states of the package and any nested packages have received a refinement and if not, emit errors. (Requires_Completion_In_Body): Add new formal parameter Do_Abstract_States. Update the comment on usage. Propagate the Do_Abstract_States flag to all Unit_Requires_Body calls. (Unit_Requires_Body): Remove formal parameter Ignore_Abstract_States. Add new formal paramter Do_Abstract_States. Propagate the Do_Abstract_States flag to all Requires_Completion_In calls. * sem_ch7.ads (Unit_Requires_Body): Remove formal parameter Ignore_Abstract_States. Add new formal paramter Do_Abstract_States. Update the comment on usage. * sem_ch9.adb (Analyze_Single_Protected_Declaration): Do not initialize the constituent list as this is now done on a need-to-add-element basis. (Analyze_Single_Task_Declaration): Do not initialize the constituent list as this is now done on a need-to-add-element basis. * sem_ch10.adb (Decorate_State): Do not initialize the constituent lists as this is now done on a need-to-add-element basis. * sem_prag.adb (Analyze_Constituent): Set the refinement constituents when adding a new element. (Analyze_Part_Of_In_Decl_Part): Set the Part_Of constituents when adding a new element. (Analyze_Part_Of_Option): Set the Part_Of constituents when adding a new element. (Analyze_Pragma): Set the Part_Of constituents when adding a new element. (Check_Constituent_Usage (all versions)): Reimplemented. (Collect_Constituent): Set the refinement constituents when adding a new element. (Create_Abstract_State): Do not initialize the constituent lists as this is now done on a need-to-add-element basis. (Propagate_Part_Of): Set the Part_Of constituents when adding a new element. * sem_util.adb (Check_State_Refinements): New routine. (Has_Non_Null_Refinement): Reimplemented. (Has_Null_Refinement): Reimplemented. (Requires_State_Refinement): Removed. * sem_util.ads (Check_State_Refinements): New routine. (Requires_State_Refinement): Removed. From-SVN: r235326
2016-04-21lib-writ.adb, [...]: Minor reformatting and code cleanup.Hristian Kirtchev4-13/+24
2016-04-21 Hristian Kirtchev <kirtchev@adacore.com> * lib-writ.adb, sem_ch6.adb: Minor reformatting and code cleanup. * sem.adb: Fix comment. From-SVN: r235325
2016-04-21[multiple changes]Arnaud Charlet6-47/+115
2016-04-21 Ed Schonberg <schonberg@adacore.com> * sem_ch3.adb (Analyze_Subtype_Declaration): A subtype declaration with no aspects, whose subtype_mark is a subtype with predicates, inherits the list of subprograms for the type. 2016-04-21 Arnaud Charlet <charlet@adacore.com> * exp_aggr.adb (Has_Per_Object_Constraint): Refine previous change. 2016-04-21 Thomas Quinot <quinot@adacore.com> * g-socket.adb (Raise_Host_Error): Include additional Name parameter. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * lib-writ.adb (Write_ALI): Do not record in ali file units that are present in the files table but not analyzed. These units are present because they appear in the context of units named in limited_with clauses, and the unit being compiled does not depend semantically on them. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * sem_ch6.adb (Analyze_Subprogram_Body_Helper): Simplify code to create the procedure body for an function returning an array type, when generating C code. Reuse the subprogram body rather than creating a new one, both as an efficiency measure and because in an instance the body may contain global references that must be preserved. From-SVN: r235324
2016-04-21max(INT_MIN, x) -> xMarc Glisse4-8/+35
2016-04-21 Marc Glisse <marc.glisse@inria.fr> gcc/ * match.pd (min(int_max, x), max(int_min, x)): New transformations. gcc/testsuite/ * gcc.dg/tree-ssa/minmax-1.c: New testcase. From-SVN: r235323
2016-04-21sem_ch3.adb, [...]: Minor reformatting.Hristian Kirtchev5-6/+14
2016-04-21 Hristian Kirtchev <kirtchev@adacore.com> * sem_ch3.adb, exp_attr.adb, exp_ch6.adb, exp_aggr.adb: Minor reformatting. From-SVN: r235322
2016-04-21exp_aggr.adb (Component_Check): Extend the check that verifies that the ↵Javier Miranda2-1/+35
aggregate has no function calls to... 2016-04-21 Javier Miranda <miranda@adacore.com> * exp_aggr.adb (Component_Check): Extend the check that verifies that the aggregate has no function calls to handle transformations performed by the frontend. (Ultimate_Original_Expression): New subprogram. From-SVN: r235321
2016-04-21[multiple changes]Arnaud Charlet3-2/+17
2016-04-21 Philippe Gil <gil@adacore.com> * krunch.adb (Krunch): Fix krunching of i-java. 2016-04-21 Arnaud Charlet <charlet@adacore.com> * exp_ch8.adb (Evaluation_Required): Always return True when Modify_Tree_For_C. From-SVN: r235320
2016-04-21ipa-inline.c (can_inline_edge_p): Pass caller info to ultiimate_alias_target.Jan Hubicka2-4/+13
* ipa-inline.c (can_inline_edge_p): Pass caller info to ultiimate_alias_target. (update_callee_keys): Likewise. (lookup_recursive_calls): Likewise. (speculation_useful_p): Likewise. From-SVN: r235319
2016-04-21re PR c++/70018 (Possible issue around IPO and C++ comdats discovered as ↵Jan Hubicka6-19/+110
pure/const) PR ipa/70018 * cgraph.c (cgraph_set_nothrow_flag_1): Rename to ... (set_nothrow_flag_1): ... this; handle interposition correctly; recurse on aliases and thunks. (cgraph_node::set_nothrow_flag): New. * ipa-pure-const.c (ignore_edge_for_nothrow): Ignore calls to functions compiled with non-call exceptions that binds to current def. (propagate_nothrow): Be safe WRT interposition. * cgraph.h (set_nothrow_flag): Update prototype. * g++.dg/ipa/nothrow-1.C: New testcase. From-SVN: r235318
2016-04-21[multiple changes]Arnaud Charlet8-18/+102
2016-04-21 Eric Botcazou <ebotcazou@adacore.com> * gnatlink.adb (Gnatlink): Robustify detection of Windows target. * alloc.ads: Minor comment fixes. * einfo.ads: Fix typo. 2016-04-21 Arnaud Charlet <charlet@adacore.com> * exp_aggr.adb (Component_Not_OK_For_Backend): Redo previous changes to handle all cases of components depending on the discriminant, not just string literals. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * sem_ch3.adb (Analyze_Subtype_Declaration): If the subtype declaration is the generated declaration for a generic actual, inherit predicates from the actual if it is a predicated subtype. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * exp_ch6.adb (Rewrite_Function_Call_For_C): If the function is inherited and its result is controlling, introduce a conversion on the actual for the corresponding procedure call, to avoid spurious type errors. 2016-04-21 Jerome Lambourg <lambourg@adacore.com> * krunch.adb (Krunch): Fix krunching of i-vxworks. From-SVN: r235317
2016-04-21[multiple changes]Arnaud Charlet3-53/+67
2016-04-21 Gary Dismukes <dismukes@adacore.com> * exp_aggr.adb: Minor reformatting and code cleanup. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * sem_ch13.adb (Resolve_Name): Omit quantified expressions from resolution, because they introduce local names. Full resolution will take place when predicate function is constructed. From-SVN: r235316
2016-04-21tree-ssa-loop-unswitch.c (tree_unswitch_single_loop): Use also ↵Jan Hubicka2-0/+10
max_loop_iterations_int. * tree-ssa-loop-unswitch.c (tree_unswitch_single_loop): Use also max_loop_iterations_int. (tree_unswitch_outer_loop): Likewise. From-SVN: r235315
2016-04-21exp_aggr.adb (Component_Not_OK_For_Backend): Refine previous change to take ↵Arnaud Charlet3-7/+15
into account Per_Object_Constraint field rather... 2016-04-21 Arnaud Charlet <charlet@adacore.com> * exp_aggr.adb (Component_Not_OK_For_Backend): Refine previous change to take into account Per_Object_Constraint field rather than special casing strings. * exp_ch6.adb: Fix typo in Replace_Returns. From-SVN: r235314
2016-04-21[multiple changes]Arnaud Charlet7-31/+135
2016-04-21 Hristian Kirtchev <kirtchev@adacore.com> * sem_ch5.adb: Minor reformatting. 2016-04-21 Eric Botcazou <ebotcazou@adacore.com> * exp_ch2.adb (Expand_Current_Value): Make an appropriate character literal if the entity is of a character type. 2016-04-21 Arnaud Charlet <charlet@adacore.com> * exp_aggr.adb (Backend_Processing_Possible): Return False when generating C and aggregate contains function calls. 2016-04-21 Tristan Gingold <gingold@adacore.com> * krunch.adb (Krunch): Only partially krunch children of Interfaces that aren't known. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * sem_prag.adb (Make_Inline): Handle properly the instantiation of a generic subpprogram that carries an Inline aspect. Place inline info on the anonymous subprogram that is constructed in the wrapper package. (Analyze_Pragma, case Pure): Do not check placement if pragma appears within an instantiation, which can be nested at any level. * sem_ch12.adb (Analyze_Instance_And_Renamings): Do not copy Freeze node from anonymous subprogram to its visible renaming. The freeze node will be constructed if the subprogram carries delayed aspects. (Set_Global): Preserve dimension information if present (from code reading). 2016-04-21 Vasiliy Fofanov <fofanov@adacore.com> * gnatlink.adb: Change wording of the warning message on problematic filenames to be more neutral. Add a new substring "patch" introduced on Windows 10. From-SVN: r235313
2016-04-21[multiple changes]Arnaud Charlet12-103/+196
2016-04-21 Philippe Gil <gil@adacore.com> * tracebak.c (__gnat_backtrace): handle bad RIP values (win64 only) 2016-04-21 Javier Miranda <miranda@adacore.com> * exp_aggr.adb (Component_Not_OK_For_Backend): Return true for string literals. 2016-04-21 Hristian Kirtchev <kirtchev@adacore.com> * einfo.adb (Has_Non_Null_Abstract_State): New routine. * einfo.ads New synthesized attribute Has_Non_Null_Abstract_State along with occurrences in entities. (Has_Non_Null_Abstract_State): New routine. * sem_ch7.adb (Unit_Requires_Body): Add local variable Requires_Body. A package declaring an abstract state requires a body only when the state is non-null and the package contains at least one other construct that requires completion in a body. * sem_util.adb (Mode_Is_Off): Removed. (Requires_State_Refinement): Remove an obsolete check. Code cleanup. 2016-04-21 Bob Duff <duff@adacore.com> * sem_attr.adb (Analyze_Attribute): In processing the 'Old attribute, a warning is given for infinite recursion. Fix the code to not crash when the prefix of 'Old denotes a protected function. * sem_ch5.adb (Analyze_Iterator_Specification): Avoid calling Is_Dependent_Component_Of_Mutable_Object in cases where the parameter would not be an object. 2016-04-21 Eric Botcazou <ebotcazou@adacore.com> * sem_eval.adb (Compile_Time_Compare): Be prepared for an empty Etype or Underlying_Type of the operands. 2016-04-21 Eric Botcazou <ebotcazou@adacore.com> * atree.adb (Print_Statistics): Protect against overflows and print the memory consumption in bytes. * table.adb (Reallocate): Do the intermediate calculation of the new size using the Memory.size_t type. From-SVN: r235312
2016-04-21[multiple changes]Arnaud Charlet4-16/+55
2016-04-21 Gary Dismukes <dismukes@adacore.com> * exp_attr.adb (Is_Inline_Floating_Point_Attribute): Suppress expansion of Attribute_Machine and Attribute_Model for AAMP. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * freeze.adb: Disable previous change for now. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * sem_ch8.adb (Find_Selected_Component): If prefix has an access type and designated type is a limited view, introduce an explicit dereference before continuing the analysis, and set its type to the non-limited view of the designated type, if we are in context where it is available. From-SVN: r235311
2016-04-21Minor reformatting.Arnaud Charlet2-14/+13
From-SVN: r235310
2016-04-21freeze.adb: Freeze profile in ASIS mode.Ed Schonberg2-15/+24
2016-04-21 Ed Schonberg <schonberg@adacore.com> * freeze.adb: Freeze profile in ASIS mode. From-SVN: r235309
2016-04-21[multiple changes]Arnaud Charlet9-61/+223
2016-04-21 Hristian Kirtchev <kirtchev@adacore.com> * sem_aux.ads, sem_aux.adb (Has_Rep_Item): New variant. * sem_util.adb (Inherit_Rep_Item_Chain): Reimplemented. 2016-04-21 Hristian Kirtchev <kirtchev@adacore.com> * exp_ch6.adb: Minor reformatting. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * sem_ch13.adb: Minor comment update. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * freeze.ads, freeze.adb (Freeze_Entity, Freeze_Before): Add boolean parameter to determine whether freezing an overloadable entity freezes its profile as well. This is required by AI05-019. The call to Freeze_Profile within Freeze_Entity is conditioned by the value of this flag, whose default is True. * sem_attr.adb (Resolve_Attribute, case 'Access): The attribute reference freezes the prefix, but it the prefix is a subprogram it does not freeze its profile. From-SVN: r235308
2016-04-21[multiple changes]Arnaud Charlet4-11/+38
2016-04-21 Javier Miranda <miranda@adacore.com> * exp_util.adb (Build_Procedure_Form): No action needed for subprogram renamings since the backend can generate the call using the renamed subprogram. This leaves the tree more clean to the backend. * exp_ch6.adb (Expand_Call): Extend previous patch for rewritten-for-c entities to handle subprogram renamings. (Rewrite_Function_Call_For_C): Handle subprogram renamings. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * sem_ch13.adb: Code cleanup. From-SVN: r235307
2016-04-21sem_ch6.adb (Analyze_Subprogram_Body_Helper): If the body is created for ↵Ed Schonberg4-4/+61
SPARK_To_C... 2016-04-21 Ed Schonberg <schonberg@adacore.com> * sem_ch6.adb (Analyze_Subprogram_Body_Helper): If the body is created for SPARK_To_C, the entity must remain invisible so it does not overload subsequent references to the original function. * exp_ch6.adb (Build_Procedure_Body_Form, Replace_Returns): Handle Extended_Return_Statements by replacing it with a block with assignments and a simple return statement. * exp_util.adb (Build_Procedure_Form): Make procedure entity invisible after analyzing declaration, to prevent improper overloading. From-SVN: r235306
2016-04-21sem_ch6.adb (Build_Subprogram_Declaration): Propagate the attribute ↵Javier Miranda4-3/+121
Rewritten_For_C to the body since since the expander... 2016-04-21 Javier Miranda <miranda@adacore.com> * sem_ch6.adb (Build_Subprogram_Declaration): Propagate the attribute Rewritten_For_C to the body since since the expander may generate calls using that entity. * exp_ch6.adb (Expand_Call): For internally generated calls ensure that they reference the entity of the spec of the called function. (Rewritten_For_C_Func_Id): New subprogram. (Rewritten_For_C_Proc_Id): New subprogram. (Rewrite_Function_Call_For_C): Invoke the new subprogram to ensure that we skip freezing entities. * exp_util.adb (Build_Procedure_Form): No action needed if the procedure was already built. From-SVN: r235305
2016-04-21[multiple changes]Arnaud Charlet13-80/+148
2016-04-21 Hristian Kirtchev <kirtchev@adacore.com> * sem_ch3.adb, exp_util.adb, sem_ch13.adb, exp_unst.adb: Minor reformatting. 2016-04-21 Ed Schonberg <schonberg@adacore.com> * sem_util.adb (Denotes_Iterator): Use root type to determine whether the ultimate ancestor is the predefined iterator interface pakage. * exp_ch5.adb (Expand_Iterator_Over_Container): simplify code and avoid reuse of Pack local variable. 2016-04-21 Olivier Hainque <hainque@adacore.com> * system-vxworks-arm.ads, system-vxworks-sparcv9.ads, system-vxworks-ppc.ads, system-vxworks-m68k.ads, system-vxworks-mips.ads, system-vxworks-x86.ads: Define Executable_Extension to ".out". From-SVN: r235304
2016-04-21frontend.adb: Update call to Unnest_Subprograms.Javier Miranda7-73/+97
2016-04-21 Javier Miranda <miranda@adacore.com> * frontend.adb: Update call to Unnest_Subprograms. * exp_ch6.ads, exp_ch6.adb, exp_unst.ads, exp_unst.adb (Unnest_Subprograms): Moved to package exp_unst. * exp_unst.ads (Unnest_Subprogram): Moved to the body of the package. * exp_dbug.adb (Qualify_Entity_Name): Enable qualification of enumeration literals when generating C code. From-SVN: r235303
2016-04-21[multiple changes]Arnaud Charlet7-126/+86
2016-04-21 Javier Miranda <miranda@adacore.com> * frontend.adb: Remove call to initialize Exp_Ch6. * exp_ch6.ads, exp_ch6.adb (Initialize): removed. (Unest_Entry/Unest_Bodies): Removed. (Unnest_Subprograms): Code cleanup. 2016-04-21 Arnaud Charlet <charlet@adacore.com> * set_targ.adb (Read_Target_Dependent_Values): close target description file once its contents is read. * s-os_lib.adb (Non_Blocking_Spawn, version with Stdout_File and Stderr_File): Close local file descriptors before spawning child process. * exp_util.adb (Containing_Package_With_Ext_Axioms): Limit scope of local variables to make the code easier to understand and avoid duplicated calls to Parent and Generic_Parent. From-SVN: r235302
2016-04-21Daily bump.GCC Administrator1-1/+1
From-SVN: r235301
2016-04-20re PR c++/55635 (Deallocation function ("operator delete") not called when ↵Nathan Sidwell5-15/+113
destructor throws exception) cp/ PR c++/55635 * init.c (build_vec_delete_1): Protect operator delete call in try finally. (build_delete): Likewise. * optimize.c (build_delete_destructor_body): Likewise. testsuite/ PR c++/55635 * g++.dg/eh/delete1.C: New. From-SVN: r235297
2016-04-20Add -mtune-ctrl=sse_typeless_stores to avx256-unaligned-store-2.cH.J. Lu2-1/+6
Since avx256-unaligned-store-2.c scans typeless SSE stores, add -mtune-ctrl=sse_typeless_stores to enable typeless SSE stores. * gcc.target/i386/avx256-unaligned-store-2.c: Add -mtune-ctrl=sse_typeless_stores. From-SVN: r235296
2016-04-20Update load scan avx256-unaligned-load-1.cH.J. Lu2-2/+6
Since avx_loadups256 and sse_loadups have been replaced by movv8sf_internal and movv4sf_internal, respectively, we need to scan movv8sf_internal and movv4sf_internal for load. * gcc.target/i386/avx256-unaligned-load-1.c: Update load scan. From-SVN: r235295
2016-04-20re PR tree-optimization/69489 (missed vectorization for boolean loop, missed ↵Bin Cheng4-6/+93
if-conversion) PR tree-optimization/69489 * tree-if-conv.c (phi_convertible_by_degenerating_args): New. (if_convertible_phi_p): Call phi_convertible_by_degenerating_args. Revise dump message. (if_convertible_bb_p): Remove check on edge count of basic block's predecessors. gcc/testsuite/ChangeLog PR tree-optimization/69489 * gcc.dg/tree-ssa/ifc-pr69489-2.c: New test. From-SVN: r235292
2016-04-20re PR c++/69363 (ICE when doing a pragma simd reduction with max)Ilya Verbin15-75/+140
Fix PR c++/69363 gcc/c-family/ PR c++/69363 * c-cilkplus.c (c_finish_cilk_clauses): Remove function. * c-common.h (c_finish_cilk_clauses): Remove declaration. gcc/c/ PR c++/69363 * c-parser.c (c_parser_cilk_all_clauses): Use c_finish_omp_clauses instead of c_finish_cilk_clauses. * c-tree.h (c_finish_omp_clauses): Add new default argument. * c-typeck.c (c_finish_omp_clauses): Add new argument. Allow floating-point variables in the linear clause for Cilk Plus. gcc/cp/ PR c++/69363 * cp-tree.h (finish_omp_clauses): Add new default argument. * parser.c (cp_parser_cilk_simd_all_clauses): Use finish_omp_clauses instead of c_finish_cilk_clauses. * semantics.c (finish_omp_clauses): Add new argument. Allow floating-point variables in the linear clause for Cilk Plus. gcc/testsuite/ PR c++/69363 * c-c++-common/cilk-plus/PS/clauses3.c: Adjust dg-error string. * c-c++-common/cilk-plus/PS/clauses4.c: New test. * c-c++-common/cilk-plus/PS/pr69363.c: New test. From-SVN: r235290
2016-04-20re PR tree-optimization/56625 (After if-conversion vectorizer doesn't ↵Bin Cheng6-21/+150
recognize similar loads) PR tree-optimization/56625 PR tree-optimization/69489 * tree-data-ref.h (DR_INNERMOST): New macro. * tree-if-conv.c (innermost_loop_behavior_hash): New class for hashing struct innermost_loop_behavior. (ref_DR_map): Remove. (innermost_DR_map): New map. (baseref_DR_map): Revise comment. (hash_memrefs_baserefs_and_store_DRs_read_written_info): Store DR to innermost_DR_map accroding to its innermost loop behavior. (ifcvt_memrefs_wont_trap): Get DR from innermost_DR_map according to its innermost loop behavior. (if_convertible_loop_p_1): Remove intialization for ref_DR_map. Add initialization for innermost_DR_map. Record memory reference in DR_BASE_ADDRESS if the reference is compound one or it doesn't have innermost loop behavior. (if_convertible_loop_p): Remove release for ref_DR_map. Release innermost_DR_map. gcc/testsuite/ChangeLog PR tree-optimization/56625 PR tree-optimization/69489 * gcc.dg/vect/pr56625.c: New test. * gcc.dg/tree-ssa/ifc-pr69489-1.c: New test. From-SVN: r235289
2016-04-20semantics.c (finish_compound_lteral): Don't wrap VECTOR_TYPEs in a TARGET_EXPR.Nathan Sidwell2-4/+14
* semantics.c (finish_compound_lteral): Don't wrap VECTOR_TYPEs in a TARGET_EXPR. From-SVN: r235288
2016-04-20i386.md (*lea<mode>_general_1): Rename from *lea_general_1.Uros Bizjak2-109/+97
* config/i386/i386.md (*lea<mode>_general_1): Rename from *lea_general_1. Use explicit SWI12 mode interator. (*lea<mode>_general_2): Rename from *lea_general_2. Use explicit SWI12 mode interator. (*lea<mode>_general_3): Rename from *lea_general_3. Use explicit SWI12 mode interator. (*lea<SWI12:mode>_general_4): Split from *lea_general_4. Use explicit SWI12 mode interator. (*lea<SWI48:mode>_general_4): Split from *lea_general_4. Use explicit SWI48 mode interator. From-SVN: r235286
2016-04-20Simplify ix86_expand_vector_move_misalignH.J. Lu2-171/+90
Since mov<mode>_internal patterns handle both aligned/unaligned load and store, we can simplify ix86_avx256_split_vector_move_misalign and ix86_expand_vector_move_misalign. * config/i386/i386.c (ix86_avx256_split_vector_move_misalign): Short-cut unaligned load and store cases. Handle all integer vector modes. (ix86_expand_vector_move_misalign): Short-cut unaligned load and store cases. Call ix86_avx256_split_vector_move_misalign directly without checking mode class. From-SVN: r235283
2016-04-20[AArch64] Work around PR target/64971Andrew Pinski4-0/+39
2016-04-20 Andrew Pinski <apinski@cavium.com> Kyrylo Tkachov <kyrylo.tkachov@arm.com> PR target/64971 * config/aarch64/aarch64.md (sibcall): Force call address to be DImode for ILP32. (sibcall_value): Likewise. * gcc.c-torture/compile/pr37433-1.c: New testcase. Co-Authored-By: Kyrylo Tkachov <kyrylo.tkachov@arm.com> From-SVN: r235281
2016-04-20Replace -skip-rax-setup with -mskip-rax-setupH.J. Lu2-1/+5
* doc/invoke.texi: Replace -skip-rax-setup with -mskip-rax-setup. From-SVN: r235275
2016-04-20gimple-match.h (maybe_build_generic_op): Adjust prototype.Richard Biener4-8/+15
2016-04-20 Richard Biener <rguenther@suse.de> * gimple-match.h (maybe_build_generic_op): Adjust prototype. * gimple-match-head.c (maybe_build_generic_op): Pass all ops by reference, clear op1 and op2 when GENERICizing BIT_FIELD_REF. (maybe_push_res_to_seq): Adjust. * gimple-fold.c (maybe_build_generic_op): Likewise. From-SVN: r235274
2016-04-20tree-if-conv.c (is_false_predicate): For NULL_TREE return false rather than ↵Marek Polacek2-3/+8
true. * tree-if-conv.c (is_false_predicate): For NULL_TREE return false rather than true. From-SVN: r235272
2016-04-20* gnat.dg/self1.adb: Remove now incorrect test.Arnaud Charlet2-21/+4
From-SVN: r235271