Arnaud Charlet [Thu, 1 Sep 2011 14:02:35 +0000 (16:02 +0200)]
Add missing prototypes.
From-SVN: r178419
Arnaud Charlet [Thu, 1 Sep 2011 13:55:43 +0000 (15:55 +0200)]
[multiple changes]
2011-09-01 Robert Dewar <dewar@adacore.com>
* sem_ch3.adb, s-taprop-linux.adb, gnatls.adb: Minor reformatting.
2011-09-01 Jose Ruiz <ruiz@adacore.com>
* adaint.h (__gnat_cpu_free): Fix the name of this profile.
* adaint.c (__gnat_cpu_alloc, __gnat_cpu_alloc_size, __gnat_cpu_free,
__gnat_cpu_zero, __gnat_cpu_set): Create version of these subprograms
specific for systems where their glibc version does not define the
routines to handle dynamically allocated CPU sets.
2011-09-01 Vincent Celier <celier@adacore.com>
* prj-proc.adb, prj.ads, prj-nmsc.adb, prj-util.adb, prj-util.ads,
prj-env.adb: Implement inheritance of naming exceptions in extending
projects.
From-SVN: r178418
Arnaud Charlet [Thu, 1 Sep 2011 13:52:39 +0000 (15:52 +0200)]
[multiple changes]
2011-09-01 Romain Berrendonner <berrendo@adacore.com>
* gnatls.adb: Display simple message instead of content of
gnatlic.adl.
2011-09-01 Hristian Kirtchev <kirtchev@adacore.com>
* sem_ch3.adb: (Build_Derived_Record_Type) Remove the kludgy update of
access discriminant and anonymous access component scopes.
(Inherit_Component): Reuse the itype of an access discriminant
or anonymous access component by copying it in order to set the proper
scope. This is done only when the parent and the derived type
are in different scopes.
(Set_Anonymous_Etype): New routine.
2011-09-01 Robert Dewar <dewar@adacore.com>
* a-convec.adb: Minor reformatting throughout.
From-SVN: r178417
Jose Ruiz [Thu, 1 Sep 2011 13:40:48 +0000 (13:40 +0000)]
adaint.c, adaint.h (__gnat_cpu_alloc, [...]): Create these wrappers around the CPU_ALLOC...
2011-09-01 Jose Ruiz <ruiz@adacore.com>
* adaint.c, adaint.h (__gnat_cpu_alloc, __gnat_cpu_alloc_size,
__gnat_cpu_set_free): Create these wrappers around the CPU_ALLOC,
CPU_ALLOC_SIZE and CPU_FREE linux macros.
(__gnat_cpu_zero, __gnat_cpu_set): Use the CPU_ZERO_S and
CPU_SET_S respectively because we are now using dynamically allocated
CPU sets which are more portable across different glibc versions.
* s-osinte-linux.ads (cpu_set_t_ptr, CPU_ALLOC, CPU_ALLOC_SIZE,
CPU_FREE): Add this type and subprograms to be able to create cpu_set_t
masks dynamically according to the number of processors in the target
platform.
(CPU_ZERO, CPU_SET): They are now mapped to the CPU_ZERO_S and CPU_SET_S
respectively, so we need to pass the size of the masks as
parameters.
* s-taprop-linux.adb (Create_Task, Set_Task_Affinity): Use dynamically
created cpu_set_t masks
with the number of processors available in the target platform,
instead of static bit arrays. It enhances portability because
it uses the information from the target platform.
* sem_ch8.adb: (Attribute_Renaming): When checking whether we
are using a restricted run-time library, use the flag
Configurable_Run_Time_Mode instead of Restricted_Profile.
From-SVN: r178416
Arnaud Charlet [Thu, 1 Sep 2011 13:37:13 +0000 (15:37 +0200)]
[multiple changes]
2011-09-01 Vincent Celier <celier@adacore.com>
* ug_words: Add /MULTI_UNIT_INDEX= -> -gnateI
* vms_data.ads: Add new VMS qualifier equivalent for -gnateInnn
2011-09-01 Nicolas Roche <roche@adacore.com>
* adaint.c (__gnat_tmp_name): Don't use tmpnam function from the system
on VxWorks in kernel mode.
From-SVN: r178415
Arnaud Charlet [Thu, 1 Sep 2011 13:29:39 +0000 (15:29 +0200)]
[multiple changes]
2011-09-01 Jose Ruiz <ruiz@adacore.com>
* s-taprop-linux.adb (Create_Task, Set_Task_Affinity): Use the linux
macros for handling CPU sets (CPU_ZERO, CPU_SET) instead of modifying
directly the bit array.
* s-osinte-linux.ads (CPU_ZERO, CPU_SET): Import these wrappers around
the linux macros with the same name.
* adaint.h, adaint.c (__gnat_cpu_zero, __gnat_cpu_set): Create these
wrappers around the CPU_ZERO and CPU_SET linux macros.
2011-09-01 Hristian Kirtchev <kirtchev@adacore.com>
* exp_ch7.adb (Find_Insertion_List): Removed.
(Process_Transient_Objects): Insert the declarations of the hook
access type and the hook object before the associated transient object.
2011-09-01 Jose Ruiz <ruiz@adacore.com>
* sem_ch8.adb (Attribute_Renaming): Add missing check to avoid loading
package System.Aux_Dec when using restricted run-time libraries which
do not have this package.
2011-09-01 Tristan Gingold <gingold@adacore.com>
* s-vaflop-vms-alpha.adb: Remove pragma optimize, useless.
2011-09-01 Bob Duff <duff@adacore.com>
* sem_attr.adb (Analyze_Access_Attribute): Do not call
Kill_Current_Values for P'Unrestricted_Access, where P is library level
2011-09-01 Thomas Quinot <quinot@adacore.com>
* exp_ch5.adb: Minor reformatting
* gnat_ugn.texi: Fix minor typos.
* gcc-interface/Make-lang.in: Update dependencies.
From-SVN: r178414
Arnaud Charlet [Thu, 1 Sep 2011 13:22:54 +0000 (15:22 +0200)]
[multiple changes]
2011-09-01 Robert Dewar <dewar@adacore.com>
* inline.adb, sem_aggr.adb: Minor reformatting.
2011-09-01 Ed Schonberg <schonberg@adacore.com>
* a-convec.adb: Proper handling of cursors for
Ada2012 iterators.
From-SVN: r178413
Arnaud Charlet [Thu, 1 Sep 2011 13:21:23 +0000 (15:21 +0200)]
[multiple changes]
2011-09-01 Robert Dewar <dewar@adacore.com>
* prj-proc.adb, exp_ch6.adb, prj-env.adb: Minor reformatting.
2011-09-01 Bob Duff <duff@adacore.com>
* sem_aggr.adb (Resolve_Aggregate): Need to treat "in instance
body" the same as "in inlined body", because visibility shouldn't
apply there.
2011-09-01 Eric Botcazou <ebotcazou@adacore.com>
* inline.adb (Add_Inlined_Body): Refine previous change.
From-SVN: r178412
Arnaud Charlet [Thu, 1 Sep 2011 13:16:58 +0000 (15:16 +0200)]
[multiple changes]
2011-09-01 Gary Dismukes <dismukes@adacore.com>
* exp_ch6.ads (Needs_BIP_Alloc_Form): New utility function.
* exp_ch6.adb (Make_Build_In_Place_Call_In_Object_Declaration):
Test for case where call
initializes an object of a return statement before testing for
a constrained call, to ensure that all such cases get handled
by simply passing on the caller's parameters. Also, in that
case call Needs_BIP_Alloc_Form to determine whether to pass on
the BIP_Alloc_Form parameter of the enclosing function rather
than testing Is_Constrained. Add similar tests for the return
of a BIP call to later processing to ensure consistent handling.
(Needs_BIP_Alloc_Form): New utility function.
* sem_ch6.adb: (Create_Extra_Formals): Replace test for adding
a BIP_Alloc_Form formal with call to new utility function
Needs_BIP_Alloc_Form.
2011-09-01 Pascal Obry <obry@adacore.com>
* prj-part.adb: Minor reformatting.
2011-09-01 Vincent Celier <celier@adacore.com>
* prj-env.adb (Create_Mapping_File.Process): Encode the upper
half character in the unit name.
From-SVN: r178411
Arnaud Charlet [Thu, 1 Sep 2011 13:14:44 +0000 (15:14 +0200)]
[multiple changes]
2011-09-01 Hristian Kirtchev <kirtchev@adacore.com>
* exp_ch4.adb: Minor code and comment reformatting.
2011-09-01 Thomas Quinot <quinot@adacore.com>
* exp_ch6.adb (Expand_Inlined_Call): Remove redundant tests
for Is_Limited_Type and Is_Tagged_Type in condition checking
for by-reference type.
* inline.adb (Add_Inlined_Body): Only exclude init_procs. Other
subprograms may have a completion because of a previous
Inline_Always clause, but the enclosing package must be marked
inlined for the subprogram body to become visible to the backend.
From-SVN: r178410
Thomas Quinot [Thu, 1 Sep 2011 13:10:40 +0000 (13:10 +0000)]
sem_aux.adb, [...]: Minor reformatting
2011-09-01 Thomas Quinot <quinot@adacore.com>
* sem_aux.adb, exp_ch4.adb: Minor reformatting
From-SVN: r178409
Richard Guenther [Thu, 1 Sep 2011 11:46:08 +0000 (11:46 +0000)]
expr.c (expand_expr_real_2): Move COND_EXPR and VEC_COND_EXPR handling here, from ...
2011-08-31 Richard Guenther <rguenther@suse.de>
* expr.c (expand_expr_real_2): Move COND_EXPR and VEC_COND_EXPR
handling here, from ...
(expand_expr_real_1): ... here.
* gimple-pretty-print.c (dump_ternary_rhs): Handle COND_EXPR
and VEC_COND_EXPR.
* gimple.c (gimple_rhs_class_table): Make COND_EXPR and VEC_COND_EXPR
a GIMPLE_TERNARY_RHS.
* tree-cfg.c (verify_gimple_assign_ternary): Handle COND_EXPR
and VEC_COND_EXPR here ...
(verify_gimple_assign_single): ... not here.
* gimple-fold.c (fold_gimple_assign): Move COND_EXPR folding.
* tree-object-size.c (cond_expr_object_size): Adjust.
(collect_object_sizes_for): Likewise.
* tree-scalar-evolution.c (interpret_expr): Don't handle
ternary RHSs.
* tree-ssa-forwprop.c (forward_propagate_into_cond): Fix and
simplify.
(ssa_forward_propagate_and_combine): Adjust.
* tree-ssa-loop-im.c (move_computations_stmt): Build the COND_EXPR
as ternary.
* tree-ssa-threadedge.c (fold_assignment_stmt): Adjust.
* tree-vect-loop.c (vect_is_simple_reduction_1): Likewise.
* tree-vect-stmt.c (vectorizable_condition): Likewise.
* tree-vrp.c (extract_range_from_cond_expr): Likewise.
(extract_range_from_assignment): Likewise.
From-SVN: r178408
Arnaud Charlet [Thu, 1 Sep 2011 11:06:39 +0000 (13:06 +0200)]
[multiple changes]
2011-09-01 Pascal Obry <obry@adacore.com>
* prj-proc.adb, prj.ads, sinput-p.adb: Minor reformatting.
2011-09-01 Ed Schonberg <schonberg@adacore.com>
* sem_ch6.adb (Analyze_Subprogram_Body_Helper): Generate references to
the formals of a subprogram stub that acts as a spec.
2011-09-01 Ed Schonberg <schonberg@adacore.com>
* exp_ch6.adb (Expand_Inlined_Call): If an actual is a by_reference
type, declare a renaming for it, not an object declaration.
From-SVN: r178407
Arnaud Charlet [Thu, 1 Sep 2011 10:59:58 +0000 (12:59 +0200)]
[multiple changes]
2011-09-01 Yannick Moy <moy@adacore.com>
* ali-util.adb, ali-util.ads (Read_Withed_ALIs): Add parameter
Ignore_Errors to ignore failures to read ALI files when True.
2011-09-01 Javier Miranda <miranda@adacore.com>
* exp_ch3.adb (Expand_N_Object_Declaration): Handle non-default
constructor calls associated with non-tagged record types.
2011-09-01 Robert Dewar <dewar@adacore.com>
* g-socthi-vms.adb: Minor reformatting.
2011-09-01 Nicolas Roche <roche@adacore.com>
* sysdep.c: Don't use macro functions for stdio functions on VxWorks
in order to avoid impact of imcompatible changes.
* cstreams.c: Likewise
* cio.c: Likewise
* aux-io.c: Likewise
2011-09-01 Yannick Moy <moy@adacore.com>
* lib-writ.adb (Write_With_Lines): Always output complete information
on "with" line in Alfa mode, as this is required by formal verification
back-end.
From-SVN: r178406
Arnaud Charlet [Thu, 1 Sep 2011 10:52:10 +0000 (12:52 +0200)]
[multiple changes]
2011-09-01 Tristan Gingold <gingold@adacore.com>
* g-socthi-vms.adb: Add comments.
2011-09-01 Pascal Obry <obry@adacore.com>
* prj.ads: Minor reformatting.
From-SVN: r178405
Arnaud Charlet [Thu, 1 Sep 2011 10:50:31 +0000 (12:50 +0200)]
[multiple changes]
2011-09-01 Vincent Celier <celier@adacore.com>
* prj-env.adb (Create_Config_Pragmas_File.Check): Put all naming
exceptions in the config pragmas file.
2011-09-01 Tristan Gingold <gingold@adacore.com>
* gnat_ugn.texi: Document GNAT_STACK_SIZE on IVMS.
From-SVN: r178404
Arnaud Charlet [Thu, 1 Sep 2011 10:48:23 +0000 (12:48 +0200)]
[multiple changes]
2011-09-01 Hristian Kirtchev <kirtchev@adacore.com>
* exp_ch7.adb (Find_Insertion_List): New routine.
(Process_Transient_Objects): Add code to handle the abnormal
finalization of a controlled transient associated with a subprogram
call. Since transients are cleaned up right after the associated
context, an exception raised during a subprogram call may bypass the
finalization code.
2011-09-01 Robert Dewar <dewar@adacore.com>
* exp_ch6.adb (Expand_Call): Check actual for aliased parameter is
aliased.
From-SVN: r178403
Robert Dewar [Thu, 1 Sep 2011 10:45:53 +0000 (10:45 +0000)]
exp_ch4.adb, [...]: Minor reformatting.
2011-09-01 Robert Dewar <dewar@adacore.com>
* exp_ch4.adb, a-exexda.adb: Minor reformatting.
From-SVN: r178402
Arnaud Charlet [Thu, 1 Sep 2011 10:44:14 +0000 (12:44 +0200)]
[multiple changes]
2011-09-01 Robert Dewar <dewar@adacore.com>
* s-taskin.ads, s-tassta.adb, sem_ch13.adb: Minor reformatting.
2011-09-01 Thomas Quinot <quinot@adacore.com>
* Makefile.rtl: Move s-oscons.o from GNATRTL_TASKING_OBJS to
GNATRTL_NONTASKING_OBJS.
2011-09-01 Robert Dewar <dewar@adacore.com>
* einfo.ads (Is_Aliased): Fix existing documentation and add note on
possibility of this flag being set for formals in Ada 2012 mode.
* par-ch6.adb (P_Formal_Part): Handle aliased for parameters for Ada
2012.
* sem_ch6.adb (Process_Formals): Handle aliased parameters in Ada 2012
mode.
* sinfo.adb (Aliased_Present): Allowed in N_Parameter_Specification for
Ada 2012.
* sinfo.ads (Aliased_Present): Allowed in N_Parameter_Specification for
Ada 2012.
2011-09-01 Hristian Kirtchev <kirtchev@adacore.com>
* exp_ch4.adb (Find_Insertion_Node): New routine. Determines the proper
insertion node in a tree of nested Expression_With_Actions nodes.
(Process_Transient_Object): In the case where a complex if statement
has been converted into nested Expression_With_Actions nodes, the
"hook" object and the associated access type must be inserted before
the top most Expression_With_Actions.
From-SVN: r178401
Arnaud Charlet [Thu, 1 Sep 2011 10:36:43 +0000 (12:36 +0200)]
[multiple changes]
2011-09-01 Robert Dewar <dewar@adacore.com>
* a-cbprqu.adb, a-cbprqu.ads, a-cuprqu.adb, a-cuprqu.ads,
a-cbsyqu.adb, a-cbsyqu.ads: Minor reformatting.
2011-09-01 Ed Schonberg <schonberg@adacore.com>
* sem_attr.adb: Conditionalize aliasing predicates to
Ada2012.
2011-09-01 Jose Ruiz <ruiz@adacore.com>
* aspects.ads (Aspect_Id, Aspect_Argument, Aspect_Names): Add the CPU
aspect.
* aspects.adb (Canonical_Aspect): Add entry for the CPU aspect.
* sem_ch13.adb (Analyze_Aspect_Specifications): Analyze the CPU aspect
in a similar way as we do for the Priority or Dispatching_Domain aspect.
* s-mudido-affinity.adb (Dispatching_Domain_Tasks,
Dispatching_Domains_Frozen): Move this local data to package
System.Tasking because with the CPU aspect we need to have access
to this data when creating the task in System.Tasking.Stages.Create_Task
* s-taskin.ads (Dispatching_Domain_Tasks, Dispatching_Domains_Frozen):
Move these variables from the body of
System.Multiprocessors.Dispatching_Domains because with the CPU aspect
we need to have access to this data when creating the task in
System.Tasking.Stages.Create_Task.
* s-taskin.adb (Initialize): Signal the allocation of the environment
task to a CPU, if any, so that we know whether the CPU can be
transferred to a different dispatching domain.
* s-tassta.adb (Create_Task): Check whether the CPU to which this task
is being allocated belongs to the dispatching domain where the task
lives. Signal the allocation of the task to a CPU, if any, so that we
know whether the CPU can be transferred to a different dispatching
domain.
From-SVN: r178400
Ed Schonberg [Thu, 1 Sep 2011 10:33:43 +0000 (10:33 +0000)]
exp_attr.adb, [...]: Implementation of attributes Same_Storage and Overlaps_Storage.
2011-09-01 Ed Schonberg <schonberg@adacore.com>
* exp_attr.adb, sem_attr.adb, snames.ads-tmpl: Implementation of
attributes Same_Storage and Overlaps_Storage.
From-SVN: r178399
Arnaud Charlet [Thu, 1 Sep 2011 10:32:07 +0000 (12:32 +0200)]
[multiple changes]
2011-09-01 Hristian Kirtchev <kirtchev@adacore.com>
* exp_strm.adb: Remove with and use clause for Opt.
(Build_Array_Input_Function): Remove the version-dependent generation
of the return statement. The Ada 2005 tree is now the default.
2011-09-01 Yannick Moy <moy@adacore.com>
* put_alfa.adb: Unconditionnally write files in Alfa section, so that
it is never empty when compiling in Alfa mode.
2011-09-01 Robert Dewar <dewar@adacore.com>
* sem_aggr.adb, sem_ch3.adb, a-direct.adb, s-taprop-vxworks.adb,
comperr.adb, exp_ch9.adb, exp_pakd.adb, sem_ch12.adb, freeze.adb,
s-taprop-mingw.adb, s-taprop-linux.adb, s-taprop-solaris.adb,
gnat1drv.adb, a-rbtgbo.adb, exp_dist.adb: Minor reformatting
2011-09-01 Matthew Heaney <heaney@adacore.com>
* Makefile.rtl, impunit.adb: Add a-csquin.ads, a-cusyqu.ad[sb],
a-cuprqu.ad[sb], a-cbsyqu.ad[sb], a-cbprqu.ad[sb]
* a-csquin.ads: New Ada 2012 unit that specifies the queue interface
* a-cusyqu.ads, a-cusyqu.adb: New Ada 2012 unit that specifies the
unbounded queue container.
* a-cbsyqu.ads, a-cbsyqu.adb: New Ada 2012 unit that specifies the
bounded queue container.
* a-cuprqu.ads, a-cuprqu.adb: New Ada 2012 unit that specifies the
unbounded priority queue container.
* a-cbprqu.ads, a-cbprqu.adb: New Ada 2012 unit that specifies the
bounded priority queue container.
From-SVN: r178398
Ira Rosen [Thu, 1 Sep 2011 08:46:59 +0000 (08:46 +0000)]
re PR tree-optimization/50178 (ICE with gfortran -O3, not with gfortran -02)
PR tree-optimization/50178
* gfortran.dg/vect/pr50178.f90: New test.
From-SVN: r178397
GCC Administrator [Thu, 1 Sep 2011 00:17:55 +0000 (00:17 +0000)]
Daily bump.
From-SVN: r178393
Richard Sandiford [Wed, 31 Aug 2011 21:08:51 +0000 (21:08 +0000)]
i386.md: Use (match_test ...) for attribute tests.
gcc/
* config/i386/i386.md: Use (match_test ...) for attribute tests.
* config/i386/mmx.md: Likewise.
* config/i386/sse.md: Likewise.
* config/i386/predicates.md (call_insn_operand): Use
(not (match_test "...")) instead of (match_test "!...")
* config/i386/constraints.md (w): Likewise.
From-SVN: r178389
Richard Sandiford [Wed, 31 Aug 2011 21:06:59 +0000 (21:06 +0000)]
md.texi: Describe the use of match_tests in attribute tests.
gcc/
* doc/md.texi: Describe the use of match_tests in attribute tests.
* rtl.def (MATCH_TEST): Update commentary.
* genattrtab.c (attr_copy_rtx, check_attr_test, clear_struct_flag)
(write_test_expr, walk_attr_value): Handle MATCH_TEST.
From-SVN: r178388
Richard Sandiford [Wed, 31 Aug 2011 21:06:29 +0000 (21:06 +0000)]
genattrtab.c (attr_rtx_1): Hash SYMBOL_REFs.
gcc/
* genattrtab.c (attr_rtx_1): Hash SYMBOL_REFs.
(attr_string): Use copy_md_ptr_loc.
From-SVN: r178387
Martin Jambor [Wed, 31 Aug 2011 17:17:19 +0000 (19:17 +0200)]
re PR middle-end/49886 (pass_split_functions cannot deal with function type attributes)
2011-08-31 Martin Jambor <mjambor@suse.cz>
PR middle-end/49886
* ipa-inline-analysis.c (compute_inline_parameters): Set
can_change_signature of noes with typde attributes.
* ipa-split.c (split_function): Do not skip any arguments if
can_change_signature is set.
* testsuite/gcc.c-torture/execute/pr49886.c: New testcase.
From-SVN: r178386
Martin Jambor [Wed, 31 Aug 2011 16:32:47 +0000 (18:32 +0200)]
cgraphunit.c (cgraph_redirect_edge_call_stmt_to_callee): Alias check removed.
2011-08-31 Martin Jambor <mjambor@suse.cz>
* cgraphunit.c (cgraph_redirect_edge_call_stmt_to_callee): Alias
check removed.
From-SVN: r178384
Richard Guenther [Wed, 31 Aug 2011 11:39:20 +0000 (11:39 +0000)]
fold-const.c (extract_muldiv_1): Remove bogus TYPE_IS_SIZETYPE special-casing.
2011-08-31 Richard Guenther <rguenther@suse.de>
* fold-const.c (extract_muldiv_1): Remove bogus TYPE_IS_SIZETYPE
special-casing.
From-SVN: r178377
Paolo Carlini [Wed, 31 Aug 2011 10:29:31 +0000 (10:29 +0000)]
chrono (operator*(const _Rep1&, const duration<>&)): Fix order of template parameters per LWG 2004.
2011-08-31 Paolo Carlini <paolo.carlini@oracle.com>
* include/std/chrono (operator*(const _Rep1&, const duration<>&)):
Fix order of template parameters per LWG 2004.
From-SVN: r178374
Arnaud Charlet [Wed, 31 Aug 2011 09:42:50 +0000 (11:42 +0200)]
[multiple changes]
2011-08-31 Pascal Obry <obry@adacore.com>
* a-direct.adb: Do not try to create an UNC path on Windows.
(Create_Path): Skip leading computer name in UNC path if any.
2011-08-31 Hristian Kirtchev <kirtchev@adacore.com>
* exp_strm.adb (Build_Record_Or_Elementary_Input_Function): Remove the
version-dependent generation of the return statement. The Ada 2005 tree
is now the default.
2011-08-31 Thomas Quinot <quinot@adacore.com>
* rtsfind.ads, exp_dist.adb (Find_Numeric_Representation): Predefined
types Stream_Element_Offset and Storage_Offset have a different native
type depending on whether the platform is 32 or 64 bits. When
exchanging them, always convert to 64 bits.
2011-08-31 Yannick Moy <moy@adacore.com>
* debug.adb: Reserve debug option -gnatd.E for passing gnatprove option
--force-alfa to gnat2why.
2011-08-31 Ed Schonberg <schonberg@adacore.com>
* sem_ch12.adb (Save_References): If the node has aspects, save
references within the corresponding expressions in a separate step,
because the aspects are not directly in the tree for the declaration
to which they belong.
2011-08-31 Arnaud Charlet <charlet@adacore.com>
* freeze.adb (Freeze_Record_Type): Omit test on variable size component
in CodePeer mode, since representation clauses are partially ignored.
* gcc-interface/Make-lang.in: Update dependencies.
From-SVN: r178373
Arnaud Charlet [Wed, 31 Aug 2011 09:37:54 +0000 (11:37 +0200)]
[multiple changes]
2011-08-31 Jose Ruiz <ruiz@adacore.com>
* s-taprop-vxworks.adb, s-taprop-mingw.adb, s-taprop-linux.adb,
s-taprop-solaris.adb (Create_Task): Not_A_Specific_CPU can be assigned
to any dispatching domain.
2011-08-31 Thomas Quinot <quinot@adacore.com>
* exp_ch4.adb: Minor reformatting.
2011-08-31 Bob Duff <duff@adacore.com>
* sem_ch6.adb (Get_Generic_Parent_Type): Don't query Subtype_Indication
on nodes for which it is not defined.
(Is_Non_Overriding_Operation): Exit the loop when we find a generic
parent type.
2011-08-31 Bob Duff <duff@adacore.com>
* sem_ch3.adb (Process_Full_View): Disable legality check if
In_Instance, to avoid spurious errors.
* sem_ch12.adb (Validate_Derived_Type_Instance): Disable legality check
if In_Instance, to avoid spurious errors.
2011-08-31 Pascal Obry <obry@adacore.com>
* a-direct.adb: Use Dir_Seps everywhere to properly handle all
directory speparators.
(Compose): Use Dir_Seps to handle both forms.
(Create_Path): Use Dir_Seps instead of explicit check, no semantic
changes.
(Extension): Use Dir_Seps to handle both forms.
2011-08-31 Pascal Obry <obry@adacore.com>
* prj-conf.adb: Minor reformatting.
From-SVN: r178372
Arnaud Charlet [Wed, 31 Aug 2011 09:33:35 +0000 (11:33 +0200)]
[multiple changes]
2011-08-31 Jose Ruiz <ruiz@adacore.com>
* aspects.ads (Aspect_Id, Aspect_Argument, Aspect_Names): Add the
dispatching domain aspect.
* aspects.adb (Canonical_Aspect): Add entry for the dispatching domain
aspect.
* sem_ch13.adb (Analyze_Aspect_Specifications): Analyze the
Dispatching_Domain aspect in a similar way as we do for the Priority
aspect.
* exp_ch9.adb (Expand_N_Task_Type_Declaration): Add the
Dispatching_Domain component if a Dispatching_Domain pragma or aspect
is present.
(Make_Task_Create_Call): Add the Dispatching_Domain when creating a task
* par-prag.adb (Prag): Add Pragma_Dispatching_Domain as a known pragma.
* sem_prag.adb (Analyze_Pragma): Check the correctness of a pragma
Dispatching_Domain and add it to the task definition.
(Sig_Flags): Add Pragma_Dispatching_Domain.
* rtsfind.ads, rtsfind.adb (RTU_Id, RE_Id, Get_Unit_Name): Add the
support to find the types Dispatching_Domain and
Dispatching_Domain_Access.
* sinfo.ads, sinfo.adb (Has_Pragma_Dispatching_Domain,
Set_Has_Pragma_Dispatching_Domain): Add these subprograms to set and
query the availability of a pragma Dispatching_Domain.
* snames.ads-tmpl (Name_uDispatching_Domain): Add this name required by
the expander to pass the Dispatching_Domain when creating a task.
(Name_Dispatching_Domain): Add this new name for a pragma.
(Pragma_Id): Add the new Pragma_Dispatching_Domain.
* s-tassta.ads, s-tassta.adb (Create_Task): Set the domain to which the
task has been allocated at creation time.
* s-tarest.adb (Create_Restricted_Task): The dispatching domain using
Ravenscar is always null.
* s-taskin.ads, s-taskin.adb (Initialize_ATCB): Set the domain to which
the task has been allocated at creation time.
* s-tporft.adb (Register_Foreign_Thread): A foreign task will not have
a specific dispatching domain.
* s-taprop-linux.adb, s-taprop-solaris.adb, s-taprop-vxworks.adb,
s-taprop-mingw.adb (Create_Task): Check whether both Dispatching_Domain
and CPU are specified for the task, and the CPU value is not contained
within the range of processors for the domain.
2011-08-31 Vincent Celier <celier@adacore.com>
* make.adb (Original_Gcc) : New constant String_Access.
(Gnatmake): For VM targets, do not use VM version of the compiler if
--GCC= has been specified.
2011-08-31 Thomas Quinot <quinot@adacore.com>
* sem_ch5.adb: Minor reformatting.
2011-08-31 Ed Schonberg <schonberg@adacore.com>
* exp_pakd.adb (Convert_To_PAT_Type): If prefix is a function call, do
not reanalyze it.
2011-08-31 Bob Duff <duff@adacore.com>
* exp_ch4.adb (Expand_N_Selected_Component): Use the full type, in case
the access type is private; we don't care about privacy in expansion.
2011-08-31 Ed Schonberg <schonberg@adacore.com>
* sem_aggr.adb (Resolve_Aggregate): In an instance, ignore aggregate
subcomponents tnat may be limited, because they originate in view
conflicts. If the original aggregate is legal and the actuals are
legal, the aggregate itself is legal.
From-SVN: r178371
Arnaud Charlet [Wed, 31 Aug 2011 09:27:37 +0000 (11:27 +0200)]
[multiple changes]
2011-08-31 Matthew Heaney <heaney@adacore.com>
* a-rbtgbo.adb (Clear_Tree): Assert representation invariant for lock
status.
2011-08-31 Ed Schonberg <schonberg@adacore.com>
* sem_ch5.adb: Set kind of loop parameter properly, to preserve all
errors in B tests.
2011-08-31 Jose Ruiz <ruiz@adacore.com>
* sem_ch13.adb (Analyze_Aspect_Specifications): For the Priority and
Interrupt_Priority aspects, force the analysis of the aspect expression
(when building the equivalent pragma). Otherwise, its analysis is done
too late, after the task or protected object has been created.
* sem_ch9.adb (Analyze_Single_Protected_Declaration,
Analyze_Single_Task_Declaration): Remove the code to move the aspects
to the object declaration because they are needed in the type
declaration.
From-SVN: r178370
Marc Glisse [Wed, 31 Aug 2011 09:26:48 +0000 (11:26 +0200)]
generic.texi (Types for C++): CP_TYPE_QUALS -> cp_type_quals.
2011-08-31 Marc Glisse <marc.glisse@inria.fr>
* doc/generic.texi (Types for C++): CP_TYPE_QUALS -> cp_type_quals.
From-SVN: r178369
Arnaud Charlet [Wed, 31 Aug 2011 09:24:32 +0000 (11:24 +0200)]
[multiple changes]
2011-08-31 Robert Dewar <dewar@adacore.com>
* exp_ch5.adb, exp_ch7.ads, sem_ch5.adb, put_scos.adb, s-rannum.adb,
a-rbtgbo.adb, exp_intr.adb, a-cbdlli.adb, a-cbdlli.ads: Minor
reformatting.
2011-08-31 Hristian Kirtchev <kirtchev@adacore.com>
* exp_util.adb (Find_Protection_Type): Do not look for fields _object
if the corresponding type is malformed due to restriction violations.
2011-08-31 Robert Dewar <dewar@adacore.com>
* s-ransee.ads, s-ransee.adb: Minor reformatting.
2011-08-31 Hristian Kirtchev <kirtchev@adacore.com>
* exp_ch4.adb (Expand_N_Allocator): Correct faulty condition which
would cause the generation of Set_Finalize_Address if the target is a
VM and the designated type is not derived from [Limited_]Controlled.
2011-08-31 Arnaud Charlet <charlet@adacore.com>
* comperr.adb, comperr.ads, gnat1drv.adb (Delete_SCIL_Files): New
subprogram.
(Compiler_Abort, Gnat1drv): Call Delete_SCIL_Files in codepeer mode in
case of a compilation error.
2011-08-31 Hristian Kirtchev <kirtchev@adacore.com>
* init.c (__gnat_error_handler): Standardize the stack overflow or
erroneous memory access message.
* seh_init.c (__gnat_SEH_error_handler): Standardize the stack overflow
or erroneous memory access message.
From-SVN: r178368
Arnaud Charlet [Wed, 31 Aug 2011 09:18:57 +0000 (11:18 +0200)]
[multiple changes]
2011-08-31 Robert Dewar <dewar@adacore.com>
* sem_ch4.adb: Minor reformatting.
* sem_ch6.adb: Minor code reorganization (use Ekind_In).
2011-08-31 Thomas Quinot <quinot@adacore.com>
* scos.ads: Minor documentation clarification.
* put_scos.adb: Do not generate SCO unit header line for a unit that
has no SCO lines.
From-SVN: r178367
Robert Dewar [Wed, 31 Aug 2011 09:17:20 +0000 (09:17 +0000)]
a-rbtgbo.adb, [...]: Minor reformatting.
2011-08-31 Robert Dewar <dewar@adacore.com>
* a-rbtgbo.adb, alfa_test.adb: Minor reformatting.
From-SVN: r178366
Arnaud Charlet [Wed, 31 Aug 2011 09:14:10 +0000 (11:14 +0200)]
[multiple changes]
2011-08-31 Tristan Gingold <gingold@adacore.com>
* exp_ch7.ads, exp_ch7.adb (Finalization_Exception_Data): New type to
hold variables between these following subprograms.
(Build_Exception_Handler, Build_Object_Declarations,
Build_Raise_Statement): Use the above type as parameter.
Make the above adjustments.
* exp_intr.adb (Expand_Unc_Deallocation): Adjust.
2011-08-31 Pascal Obry <obry@adacore.com>
* projects.texi: Minor reformatting.
2011-08-31 Tristan Gingold <gingold@adacore.com>
* s-ransee.ads, s-ransee.adb: Add system.random_seed unit.
* s-rannum.adb (Reset): Use Get_Seed from s-ransee.
2011-08-31 Ed Schonberg <schonberg@adacore.com>
* exp_ch5.adb: Minor code cleanup.
* sem_ch5.adb (Analyze_Iteration_Scheme): Set ekind of loop variable to
prevent cascaded errors.
(Analyze_Loop_Statement): In semantics-only mode, introduce loop
variable of an iterator specification in current scope.
* sem_ch6.adb (Analyze_Return_Statement, Find_what_It_Apples_To): Skip
postconditions on the stack, as they contain no return statements.
2011-08-31 Yannick Moy <moy@adacore.com>
* exp_alfa.adb (Expand_Alfa_N_Package_Declaration,
Expand_Alfa_N_Subprogram_Body): Remove useless procedures which simply
call Qualify_Entity_Names.
(Expand_Alfa): call Qualify_Entity_Names in more cases
* lib-xref-alfa.adb: Take into account system package.
* sem_prag.adb Take into account restrictions in Alfa mode, contrary to
CodePeer mode in which we are interested in finding bugs even if
compiler cannot compile source.
* sem_util.adb, sem_util.ads (Unique_Entity): Take into account case of
deferred constant.
2011-08-31 Gary Dismukes <dismukes@adacore.com>
* sem_ch3.adb (Constrain_Concurrent): Retrieve Base_Type of the subtype
denoted by the subtype mark to ensure getting the concurrent type in
the case where the subtype mark denotes a private subtype of a
concurrent type (needed when using -gnatc).
(Process_Subtype): For the processing specific to type kinds, case on
the Base_Type kind of the Subtype_Mark_Id, to handle cases where the
subtype denotes a private subtype whose base type is nonprivate (needed
for subtypes of private fulfilled by task types when compiling with
-gnatc).
2011-08-31 Gary Dismukes <dismukes@adacore.com>
* sem_disp.adb (Check_Dispatching_Operation): Bypass registration of
late primitives that override interface operations when the full
expander is not active, to avoid blowups in Register_Primitive when
types don't have associated secondary dispatch tables.
2011-08-31 Yannick Moy <moy@adacore.com>
* alfa_test.adb: Code clean up.
2011-08-31 Marc Sango <sango@adacore.com>
* restrict.adb (Check_SPARK_Restriction): Change Comes_From_Source (N)
by Comes_From_Source (Original_Node (N)) in order to treat also the
nodes which have been rewritten.
* sem_ch4.adb (Analyze_Explicit_Dereference, Analyze_Slice): Guard the
explicit dereference and slice violation in spark mode on the nodes
coming only from the source code.
From-SVN: r178365
Arnaud Charlet [Wed, 31 Aug 2011 09:07:20 +0000 (11:07 +0200)]
[multiple changes]
2011-08-31 Robert Dewar <dewar@adacore.com>
* exp_ch5.adb, exp_alfa.ads, prj.ads, sem_attr.adb,
lib-xref-alfa.adb: Minor reformatting.
2011-08-31 Matthew Heaney <heaney@adacore.com>
* a-crbltr.ads (Tree_Type): Default-initialize the Nodes component.
2011-08-31 Javier Miranda <miranda@adacore.com>
* sem_ch4.adb (Try_Object_Operation): Addition of one formal to search
only for class-wide subprograms conflicting with entities of concurrent
tagged types.
2011-08-31 Matthew Heaney <heaney@adacore.com>
* a-rbtgbo.adb (Generic_Allocate): Initialize pointer components of
node to null value.
2011-08-31 Hristian Kirtchev <kirtchev@adacore.com>
* sem_ch12.adb (Insert_Freeze_Node_For_Instance): Provide a more
general description of the routine.
2011-08-31 Ed Schonberg <schonberg@adacore.com>
* a-cbdlli.adb, a-cbdlli.ads: Add iterator machinery to bounded
doubly-linked lists.
From-SVN: r178363
Gary Dismukes [Wed, 31 Aug 2011 09:02:37 +0000 (09:02 +0000)]
exp_ch5.adb (Expand_N_Assignment_Statement): When a discriminant check is needed for a left-hand side that is a dereference...
2011-08-31 Gary Dismukes <dismukes@adacore.com>
* exp_ch5.adb (Expand_N_Assignment_Statement): When a discriminant
check is needed for a left-hand side that is a dereference, and the
base type is private without discriminants (whereas the full type does
have discriminants), an extra retrieval of the underlying type may be
needed in the case where the subtype is a record subtype whose base
type is private. Update comments.
From-SVN: r178362
Arnaud Charlet [Wed, 31 Aug 2011 08:59:01 +0000 (10:59 +0200)]
[multiple changes]
2011-08-31 Javier Miranda <miranda@adacore.com>
* sem_ch4.adb (Try_Object_Operation): When a dispatching primitive is
found check if there is a class-wide subprogram covering the primitive.
2011-08-31 Yannick Moy <moy@adacore.com>
* sem_res.adb: Further cases where full expansion test is needed,
rather than expansion test.
2011-08-31 Pascal Obry <obry@adacore.com>
* prj-attr.adb: Fix Source_File_Switches attribute kind (must be a list)
2011-08-31 Ed Schonberg <schonberg@adacore.com>
* exp_ch5.adb: Handle iterators over derived container types.
2011-08-31 Hristian Kirtchev <kirtchev@adacore.com>
* einfo.ads, einfo.adb: Add new flag Has_Anonymous_Master.
(Has_Anonymous_Master): New routine.
(Set_Has_Anonymous_Master): New routine.
(Write_Entity_Flags): Add an entry for Has_Anonymous_Master.
* exp_ch4.adb: Add with and use clause for Sem_Ch8.
(Current_Anonymous_Master): New routine.
(Current_Unit_First_Declaration): Removed.
(Current_Unit_Scope): Removed.
(Expand_N_Allocator): Anonymous access-to-controlled types now chain
their objects on a per-unit heterogeneous finalization master.
2011-08-31 Matthew Heaney <heaney@adacore.com>
* a-cbhama.adb, a-cbhase.adb (Insert): Check for zero-length buckets
array.
2011-08-31 Jose Ruiz <ruiz@adacore.com>
* s-taprop-linux.adb (Create_Task): Avoid changing the affinity mask
when not needed.
2011-08-31 Gary Dismukes <dismukes@adacore.com>
* sem_disp.adb (Propagate_Tag): Return without propagating in the case
where the actual is an unexpanded call to 'Input.
From-SVN: r178361
Arnaud Charlet [Wed, 31 Aug 2011 08:52:34 +0000 (10:52 +0200)]
[multiple changes]
2011-08-31 Yannick Moy <moy@adacore.com>
* sem_ch4.adb: Code clean up.
2011-08-31 Yannick Moy <moy@adacore.com>
* exp_alfa.adb, exp_alfa.ads: Minor correction of copyright notice.
2011-08-31 Hristian Kirtchev <kirtchev@adacore.com>
* exp_ch7.adb (Build_Array_Deep_Procs): Do not generate Deep_Finalize
and TSS primitive Finalize_Address if finalization is suppressed.
(Build_Record_Deep_Procs): Do not generate Deep_Finalize and TSS
primitive Finalize_Address if finalization is suppressed.
2011-08-31 Jose Ruiz <ruiz@adacore.com>
* s-mudido-affinity.adb, s-taprop-linux.adb, s-taprop-mingw.adb,
s-taprop-solaris.adb, s-taprop-vxworks.adb (Set_Task_Affinity): Make
sure that the underlying task has already been created before trying
to change its affinity.
(Set_CPU): Use the term processor instead of CPU, as we do in
Assign_Task.
2011-08-31 Vincent Celier <celier@adacore.com>
* prj-attr.adb: New Compiler attribute Source_File_Switches.
* prj-nmsc.adb (Process_Compiler): Process attribute
Source_File_Switches.
* prj.ads (Language_Config): New name list component
Name_Source_File_Switches.
* snames.ads-tmpl (Name_Source_File_Switches): New standard name.
2011-08-31 Ed Schonberg <schonberg@adacore.com>
* sem_attr.adb (Analyze_Attribute, case 'Old): If prefix may be a
discriminated component of an actual, expand at once to prevent
ouf-of-order references with generated subtypes.
2011-08-31 Yannick Moy <moy@adacore.com>
* lib-xref-alfa.adb (Add_Alfa_Xrefs): Do not take into account read
reference to operator in Alfa xrefs.
2011-08-31 Hristian Kirtchev <kirtchev@adacore.com>
* sem_ch12.adb (Freeze_Subprogram_Body): Add code to handle the case
where the parent instance was frozen before the current instance due to
the presence of a source body. Update calls to Insert_After_Last_Decl.
(Insert_After_Last_Decl): Renamed to Insert_Freeze_Node_For_Instance.
Update the comment which illustrates the purpose of the routine.
Package instances are now frozen by source bodies which appear after
the instance. This ensures that entities coming from within the
instance are available for use in the said bodies.
(Install_Body): Add code to handle the case where the parent instance
was frozen before the current instance due to the presence of a source
body. Update calls to Insert_After_Last_Decl.
From-SVN: r178360
Arnaud Charlet [Wed, 31 Aug 2011 08:47:22 +0000 (10:47 +0200)]
Complete log of previous change.
From-SVN: r178359
Arnaud Charlet [Wed, 31 Aug 2011 08:43:42 +0000 (10:43 +0200)]
[multiple changes]
2011-08-31 Jose Ruiz <ruiz@adacore.com>
* s-taprop-linux.adb (Set_Task_Affinity): Avoid the use of anonymous
access types.
* affinity.c (__gnat_set_affinity_mask): Declare index variable.
2011-08-31 Yannick Moy <moy@adacore.com>
* sem_ch8.adb (Analyze_Subprogram_Renaming): Refine expander test in
full-expander test.
Minor reformatting, renaming ALFA in Alfa (we dropped acronym)
* einfo.adb (Primitive_Operations): Correctly return list of primitive
operations in a case where it returned previously No_Elist.
From-SVN: r178358
Robert Dewar [Wed, 31 Aug 2011 08:40:22 +0000 (08:40 +0000)]
s-taprop-vxworks.adb, [...]: Minor reformatting.
2011-08-31 Robert Dewar <dewar@adacore.com>
* s-taprop-vxworks.adb, sem_ch5.adb, s-taprop-tru64.adb, exp_alfa.adb,
s-taprop-vms.adb, bindgen.adb, s-mudido.adb, s-mudido.ads,
sem_res.adb, expander.adb, s-taprop-mingw.adb, s-taprop-linux.adb,
s-taprop-solaris.adb, s-mudido-affinity.adb, vms_conv.adb,
s-taprop-irix.adb, s-taprop.ads, s-taskin.adb, s-taskin.ads,
s-taprop-hpux-dce.adb, a-chtgbo.adb, s-taprop-posix.adb: Minor
reformatting.
From-SVN: r178357
Arnaud Charlet [Wed, 31 Aug 2011 08:37:37 +0000 (10:37 +0200)]
[multiple changes]
2011-08-31 Hristian Kirtchev <kirtchev@adacore.com>
* sem_ch12 (Check_Private_View): Revert previous change.
* sem_res.adb (Conversion_Check): Do not emit the error message if the
conversion is in a generic instance.
2011-08-31 Matthew Heaney <heaney@adacore.com>
* a-cbhase.adb (Symmetric_Difference): Dereference correct node array.
* a-chtgbo.adb (Free): Allow 0 as index value.
2011-08-31 Matthew Heaney <heaney@adacore.com>
* a-cborma.adb (Insert): Add comment to explain why no element
assignment.
2011-08-31 Gary Dismukes <dismukes@adacore.com>
* sem_util.adb (Find_Body_Discriminal): Test whether the scope of the
spec discriminant is already a concurrent type, in which case just use
it, otherwise fetch the Corresponding_Concurrent_Type as before.
From-SVN: r178356
Tom de Vries [Wed, 31 Aug 2011 07:05:59 +0000 (07:05 +0000)]
re PR middle-end/43513 (The stack pointer is adjusted twice)
2011-08-31 Tom de Vries <tom@codesourcery.com>
PR middle-end/43513
* gcc.dg/pr43513.c: New test.
From-SVN: r178354
Tom de Vries [Wed, 31 Aug 2011 07:04:25 +0000 (07:04 +0000)]
re PR middle-end/43513 (The stack pointer is adjusted twice)
2011-08-31 Tom de Vries <tom@codesourcery.com>
PR middle-end/43513
* Makefile.in (tree-ssa-ccp.o): Add $(PARAMS_H) to rule.
* tree-ssa-ccp.c (params.h): Include.
(fold_builtin_alloca_for_var): New function.
(ccp_fold_stmt): Use fold_builtin_alloca_for_var.
From-SVN: r178353
GCC Administrator [Wed, 31 Aug 2011 00:19:13 +0000 (00:19 +0000)]
Daily bump.
From-SVN: r178351
Tobias Burnus [Tue, 30 Aug 2011 22:09:49 +0000 (00:09 +0200)]
re PR fortran/45044 (Different named COMMON block size: No warning)
2011-08-30 Tobias Burnus <burnus@net-b.de>
PR fortran/45044
* trans-common.c (build_common_decl): Warn if named common
block's size is not everywhere the same.
2011-08-30 Tobias Burnus <burnus@net-b.de>
PR fortran/45044
* gfortran.dg/common_14.f90: New.
* gfortran.dg/common_resize_1.f: Add two dg-warning.
From-SVN: r178344
Jason Merrill [Tue, 30 Aug 2011 21:27:36 +0000 (17:27 -0400)]
re PR c++/50084 ([C++0x] ICE: decltype + remove_reference + new)
PR c++/50084
* cp-tree.h (cp_decl_specifier_seq): Rename user_defined_type_p
to type_definition_p.
* parser.c (cp_parser_set_decl_spec_type): Likewise.
* decl.c (grokdeclarator): Check it.
From-SVN: r178340
Jason Merrill [Tue, 30 Aug 2011 21:27:27 +0000 (17:27 -0400)]
re PR c++/50089 ([C++0x] ICE when calling a qualified base class member function within a lambda expr without "this->")
PR c++/50089
* semantics.c (finish_id_expression): Use
current_nonlambda_class_type for qualified-ids.
From-SVN: r178339
Jason Merrill [Tue, 30 Aug 2011 21:27:18 +0000 (17:27 -0400)]
re PR c++/50114 (ICE with declaration inside for statement)
PR c++/50114
* decl.c (poplevel): Disable for scope compatibility hack
in C++11 mode.
From-SVN: r178338
Uros Bizjak [Tue, 30 Aug 2011 17:52:29 +0000 (19:52 +0200)]
i386.c (ix86_valid_target_attribute_inner_p): Handle FMA option.
* config/i386/i386.c (ix86_valid_target_attribute_inner_p):
Handle FMA option.
testsuite/ChangeLog:
* gcc.target/i386/fma-compile.c: Escape [ and ] in scan strings.
From-SVN: r178337
Eric Botcazou [Tue, 30 Aug 2011 16:49:06 +0000 (16:49 +0000)]
system-irix-n64.ads, [...]: Remove GCC_ZCX_Support.
* system-irix-n64.ads, system-linux-armeb.ads, system-linux-armel.ads,
system-linux-mips.ads, system-linux-mipsel.ads,
system-linux-mips64el.ads, system-linux-ppc64.ads,
system-linux-sparcv9.ads, system-rtems.ads: Remove GCC_ZCX_Support.
From-SVN: r178336
Eric Botcazou [Tue, 30 Aug 2011 16:46:12 +0000 (16:46 +0000)]
ada-tree.h (TYPE_EXTRA_SUBTYPE_P): Add internal check.
* gcc-interface/ada-tree.h (TYPE_EXTRA_SUBTYPE_P): Add internal check.
* gcc-interface/utils2.c (build_binary_op): Tighten condition.
(build_unary_op): Likewise.
From-SVN: r178335
Eric Botcazou [Tue, 30 Aug 2011 16:35:54 +0000 (16:35 +0000)]
raise-gcc.c: Do not include coretypes.h and tm.h.
* raise-gcc.c: Do not include coretypes.h and tm.h.
(setup_to_install): Remove test for compiler macro.
* targext.c: Document use for the library.
* gcc-interface/Makefile.in: Fix comment on the use of IN_GCC.
From-SVN: r178333
Eric Botcazou [Tue, 30 Aug 2011 16:15:49 +0000 (16:15 +0000)]
decl.c (gnat_to_gnu_entity): Do not convert the expression to the nominal type if...
* gcc-interface/decl.c (gnat_to_gnu_entity) <object>: Do not convert
the expression to the nominal type if the latter is a record type with
a variant part and the type of the former is a record type without one.
Co-Authored-By: Laurent Rougé <laurent.rouge@menta.fr>
From-SVN: r178331
Paolo Carlini [Tue, 30 Aug 2011 15:34:50 +0000 (15:34 +0000)]
stl_iterator.h (make_move_iterator): Implement DR2061.
2011-08-30 Paolo Carlini <paolo.carlini@oracle.com>
* include/bits/stl_iterator.h (make_move_iterator): Implement DR2061.
* testsuite/24_iterators/move_iterator/dr2061.cc: New.
* testsuite/20_util/weak_ptr/comparison/cmp_neg.cc: Adjust dg-warning
line numbers.
From-SVN: r178330
Steven G. Kargl [Tue, 30 Aug 2011 15:34:01 +0000 (15:34 +0000)]
re PR fortran/45170 ([F2003] allocatable character lengths)
2011-08-30 Steven G. Kargl <kargl@gcc.gnu.org>
PR fortran/45170
* trans-stmt.c (gfc_trans_allocate): Evaluate the substring.
2011-08-30 Steven G. Kargl <kargl@gcc.gnu.org>
PR fortran/45170
* gfortran.dg/allocate_with_source_2.f90: New test
From-SVN: r178329
Jason Merrill [Tue, 30 Aug 2011 15:28:40 +0000 (11:28 -0400)]
re PR c++/50220 ([C++0x] [4.7 Regression] ICE when capturing a by-reference template function argument in a lambda)
PR c++/50220
* semantics.c (add_capture): Call complete_type for copy.
From-SVN: r178326
Jason Merrill [Tue, 30 Aug 2011 15:28:30 +0000 (11:28 -0400)]
re PR c++/50234 (internal compiler error: in cxx_eval_component_reference, at cp/semantics.c:6527)
PR c++/50234
* semantics.c (cxx_eval_component_reference): Handle
value-initialization for omitted initializers.
From-SVN: r178325
Andrew Stubbs [Tue, 30 Aug 2011 15:02:26 +0000 (15:02 +0000)]
arm.c (optimal_immediate_sequence_1): Make b1, b2, b3 and b4 unsigned.
2011-08-30 Andrew Stubbs <ams@codesourcery.com>
gcc/
* config/arm/arm.c (optimal_immediate_sequence_1): Make b1, b2,
b3 and b4 unsigned.
From-SVN: r178324
Arnaud Charlet [Tue, 30 Aug 2011 15:01:43 +0000 (17:01 +0200)]
Switch to GPLv3
From-SVN: r178323
Arnaud Charlet [Tue, 30 Aug 2011 14:58:35 +0000 (16:58 +0200)]
Update dependencies
From-SVN: r178322
Arnaud Charlet [Tue, 30 Aug 2011 14:56:45 +0000 (16:56 +0200)]
[multiple changes]
2011-08-30 Jose Ruiz <ruiz@adacore.com>
* s-taskin.ads (Common_ATCB): Add field domain which contains the
dispatching domain to which the task belongs.
* s-taskin.adb (Initialize): Create the default system dispatching
domain and make the environment task part of it.
* s-mudido.ads: Add this new spec for standard Ada 2012 package
Ada.Multiprocessors.Dispatching_Domains.
* s-mudido.adb: Add this new body for targets not supporting
dispatching domains.
* s-mudido-affinity.adb: Add this new body for targets supporting
dispatching domains setting the affinity to a CPU set.
* bindgen.adb (Dispatching_Domain_Used, Check_Dispatching_Domains_Used,
Gen_Adainit): When package System.Multiprocessors.Dispatching_Domains
is used we call the procedure to signal that when we are about to call
the main subprogram no new dispatching domain can be created.
(Check_File_In_Partition): Factor out the common functionality used by
Check_System_Restrictions_Used and Check_Dispatching_Domains_Used.
* s-tassta.adb (Create_Task): Tasks inherit the dispatching domain of
their activators.
* s-taprop.ads (Set_Task_Affinity): Add this new procedure to set task
affinities.
* s-taprop-dummy.adb, s-taprop-hpux-dce.adb, s-taprop-irix.adb,
s-taprop-posix.adb, s-taprop-tru64.adb, s-taprop-vms.adb
(Set_Task_Affinity): Dummy null body for these targets not supporting
task affinities.
s-taprop-linux.adb, s-taprop-mingw.adb, s-taprop-solaris.adb,
s-taprop-vxworks.adb (Create_Task, Enter_Task, Initialize): Handle
dispatching domains and set the affinity of the environment task.
(Set_Task_Affinity): Procedure that uses the underlying CPU set
functionality to handle dispatching domains, pragma CPU and Task_Info.
s-winext.ads (SetThreadAffinityMask): Import this function needed to
set CPU masks.
* s-osinte-solaris.ads (psetit_t, pset_create, pset_assign, pset_bind):
Import the functionality to handle CPU set affinities.
* affinity.c: New file.
* s-osinte-vxworks.ads, s-vxwext.ads, s-vxwext-kernel.ads,
s-vxwext-rtp.ads (taskMaskAffinitySet): Add this new spec for setting
affinity masks.
* s-vxwext.adb, s-vxwext-kernel.adb, s-vxwext-rtp.adb
(taskMaskAffinitySet): Body returning an error indicating that task
affinities are not supported.
Makefile.rtl: Indicate that s-mudido is part of libgnarl.
* gcc-interface/Makefile.in (LIBGNAT_TARGET_PAIRS for VxWorks SMP,
Solaris, Windows, and {x86,PowerPC, ia64,x86_64} Linux): Use the
s-mudido-affinity.adb body which supports task affinities.
2011-08-30 Thomas quinot <quinot@adacore.com>
* sem_ch13.adb: Minor reformatting.
2011-08-30 Vincent Celier <celier@adacore.com>
* vms_conv.adb (Process_Argument): When the qualifier
/UNCHECKED_SHARED_LIB_IMPORTS is for GNAT COMPILE, do not put the
corresponding switch --unchecked-shared-lib-imports after -cargs, as it
is for gnatmake, not for the compiler.
2011-08-30 Ed Schonberg <schonberg@adacore.com>
* sem_ch4.adb (Analyze_Quantified_Expression): Analyze iterator
specification and condition only in Semantics_Only mode. Otherwise the
analysis is done after expression has been rewritten as loop.
* sem_ch5.adb (Analyze_Iterator_Specification): Always generate a
temporary for the iterator name (the domain of iteration) because it
may need finalization actions and these must be generated outside of
the loop.
* sem_res.adb (Resolve_Quantified_Expression): Resolve only in
Semantic_Only mode.
* exp_ch4.adb (Expand_Quantified_Expression): Analyze and resolve once
rewritten as loop.
* exp_ch5.adb (Expand_Iterator_Loop): Code clean-up, now that the
iterator is always an expression.
2011-08-30 Robert Dewar <dewar@adacore.com>
* par-ch4.adb (P_Unparen_Cond_Case_Quant_Expression): New function
(P_Expression_If_OK): New spec checks parens
(P_Expression_Or_Range_Attribute_If_OK): New spec checks parens
* par.adb (P_Expression_If_OK): New spec checks parens
(P_Expression_Or_Range_Attribute_If_OK): New spec checks parens
From-SVN: r178321
Andrew Stubbs [Tue, 30 Aug 2011 14:28:17 +0000 (14:28 +0000)]
arm.c (arm_gen_constant): Set can_negate correctly when code is SET.
2011-08-30 Andrew Stubbs <ams@codesourcery.com>
gcc/
* config/arm/arm.c (arm_gen_constant): Set can_negate correctly
when code is SET.
From-SVN: r178319
Yannick Moy [Tue, 30 Aug 2011 14:23:30 +0000 (14:23 +0000)]
* gcc-interface/Make-lang.in: Update dependencies.
From-SVN: r178318
Arnaud Charlet [Tue, 30 Aug 2011 14:20:11 +0000 (16:20 +0200)]
[multiple changes]
2011-08-30 Yannick Moy <moy@adacore.com>
* exp_aggr.adb, exp_ch11.adb, exp_prag.adb: Remove early exit during
expansion in Alfa mode.
* exp_ch6.adb, exp_ch6.ads (Expand_Actuals): Make subprogram public
* exp_alfa.adb, exp_alfa.ads: New package defining light expansion for
Alfa mode.
* gnat1drv.adb (Adjust_Global_Switches): Update Comment.
* sem_res.adb: Ditto.
2011-08-30 Thomas Quinot <quinot@adacore.com>
* g-socket.ads: Minor documentation adjustment.
From-SVN: r178317
Arnaud Charlet [Tue, 30 Aug 2011 14:16:43 +0000 (16:16 +0200)]
exp_ch9.adb, [...]: Minor reformatting.
2011-08-30 Robert Dewar <dewar@adacore.com>
* exp_ch9.adb, s-tassta.adb, s-secsta.adb: Minor reformatting.
2011-08-30 Yannick Moy <moy@adacore.com>
* exp_ch6_light.adb, exp_ch6_light.ads, exp_attr_light.adb,
exp_attr_light.ads, exp_ch7_light.adb, exp_ch7_light.ads,
exp_light.adb, exp_light.ads, exp_prag.adb, expander.adb,
gnat1drv.adb, exp_ch11.adb, exp_ch6.adb, exp_ch6.ads, exp_aggr.adb:
Revert change which introduced files for "light"
expansion, to be replaced by a single file for Alfa expansion.
From-SVN: r178316
Tristan Gingold [Tue, 30 Aug 2011 14:13:06 +0000 (14:13 +0000)]
Renames Ratio to Percentage, and Sec_Stack_Ratio to Sec_Stack_Percentage
From-SVN: r178315
Robert Dewar [Tue, 30 Aug 2011 14:12:00 +0000 (14:12 +0000)]
opt.ads, [...]: Update comment.
2011-08-30 Robert Dewar <dewar@adacore.com>
* opt.ads, s-soflin.adb, exp_ch9.adb, sem_res.adb: Update comment.
Minor code reorg/reformatting.
From-SVN: r178314
Arnaud Charlet [Tue, 30 Aug 2011 14:06:25 +0000 (16:06 +0200)]
[multiple changes]
2011-08-30 Yannick Moy <moy@adacore.com>
* opt.adb, opt.ads (Full_Expander_Active): New function defines a
common shorthand for (Expander_Active and not ALFA_Mode) that can be
used for testing full expansion, that is active expansion not in the
reduced mode for Alfa
* exp_ch4.adb, exp_ch9.adb, exp_disp.adb, sem_ch10.adb, sem_ch12.adb,
sem_ch6.adb, sem_ch9.adb, sem_res.adb: Use newly defined "flag" instead
of the verbose (Expander_Active and not ALFA_Mode)
2011-08-30 Tristan Gingold <gingold@adacore.com>
* s-parame-vms-alpha.ads, s-parame-hpux.ads, s-tassta.adb,
s-tarest.adb, s-parame-vms-ia64.ads, s-soflin.adb, s-secsta.adb,
s-secsta.ads, s-parame.ads, s-parame-vxworks.ads: Renames Ratio to
Percentage, and Sec_Stack_Ratio to Sec_Stack_Percentage.
From-SVN: r178313
Richard Guenther [Tue, 30 Aug 2011 14:06:00 +0000 (14:06 +0000)]
re PR tree-optimization/48571 (Missed data-dependence for (bogus?) reconstructed array-refs)
2011-08-30 Richard Guenther <rguenther@suse.de>
PR middle-end/48571
* gimple.h (maybe_fold_offset_to_address): Remove.
(maybe_fold_offset_to_reference): Likewise.
(maybe_fold_stmt_addition): Likewise.
(may_propagate_address_into_dereference): Likewise.
* tree-inline.c (remap_gimple_op_r): Do not reconstruct
array references.
* gimple-fold.c (canonicalize_constructor_val): Likewise.
Canonicalize invariant POINTER_PLUS_EXPRs to invariant MEM_REF
addresses instead.
(may_propagate_address_into_dereference): Remove.
(maybe_fold_offset_to_array_ref): Likewise.
(maybe_fold_offset_to_reference): Likewise.
(maybe_fold_offset_to_address): Likewise.
(maybe_fold_stmt_addition): Likewise.
(fold_gimple_assign): Do not reconstruct array references but
instead canonicalize invariant POINTER_PLUS_EXPRs to invariant
MEM_REF addresses.
(gimple_fold_stmt_to_constant_1): Likewise.
* tree-ssa-forwprop.c (forward_propagate_addr_expr_1): Likewise.
* gimplify.c (gimplify_conversion): Likewise.
(gimplify_expr): Likewise.
* gcc.c-torture/execute/pr48571-1.c: New testcase.
* gcc.dg/tree-ssa/ssa-ccp-25.c: Remove.
* gcc.dg/tree-ssa/ssa-ccp-26.c: Likewise.
* gcc.dg/pr36902.c: XFAIL.
From-SVN: r178312
Ilya Tocar [Tue, 30 Aug 2011 14:02:53 +0000 (14:02 +0000)]
Add FMA intrinsics and testcases.
gcc/
2011-08-30 Ilya Tocar <ilya.tocar@intel.com>
* config/i386/fmaintrin.h: New.
* config.gcc: Add fmaintrin.h.
* config/i386/i386.c
(enum ix86_builtins) <IX86_BUILTIN_VFMADDSS3>: New.
<IX86_BUILTIN_VFMADDSD3>: Likewise.
* config/i386/sse.md (fmai_vmfmadd_<mode>): New.
(*fmai_fmadd_<mode>): Likewise.
(*fmai_fmsub_<mode>): Likewise.
(*fmai_fnmadd_<mode>): Likewise.
(*fmai_fnmsub_<mode>): Likewise.
* config/i386/immintrin.h: Add fmaintrin.h.
gcc/testsuite/
2011-08-30 Ilya Tocar <ilya.tocar@intel.com>
* gcc.target/i386/fma-check.h: New.
* gcc.target/i386/fma-256-fmaddXX.c: New testcase.
* gcc.target/i386/fma-256-fmaddsubXX.c: Likewise.
* gcc.target/i386/fma-256-fmsubXX.c: Likewise.
* gcc.target/i386/fma-256-fmsubaddXX.c: Likewise.
* gcc.target/i386/fma-256-fnmaddXX.c: Likewise.
* gcc.target/i386/fma-256-fnmsubXX.c: Likewise.
* gcc.target/i386/fma-fmaddXX.c: Likewise.
* gcc.target/i386/fma-fmaddsubXX.c: Likewise.
* gcc.target/i386/fma-fmsubXX.c: Likewise.
* gcc.target/i386/fma-fmsubaddXX.c: Likewise.
* gcc.target/i386/fma-fnmaddXX.c: Likewise.
* gcc.target/i386/fma-fnmsubXX.c: Likewise.
* gcc.target/i386/fma-compile.c: Likewise.
* gcc.target/i386/i386.exp (check_effective_target_fma): New.
* gcc.target/i386/sse-12.c: Add -mfma.
* gcc.target/i386/sse-13.c: Likewise.
* gcc.target/i386/sse-14.c: Likewise.
* gcc.target/i386/sse-22.c: Likewise.
* gcc.target/i386/sse-23.c: Likewise.
* g++.dg/other/i386-2.C: Likewise.
* g++.dg/other/i386-3.C: Likewise.
From-SVN: r178311
Arnaud Charlet [Tue, 30 Aug 2011 14:01:42 +0000 (16:01 +0200)]
[multiple changes]
2011-08-30 Gary Dismukes <dismukes@adacore.com>
* sem_res.adb (Valid_Conversion): Revise test for implicit anonymous
access conversions to check that the conversion is a rewritten node,
rather than just having Comes_From_Source set to False, which wasn't
sufficient.
2011-08-30 Robert Dewar <dewar@adacore.com>
* exp_ch9.adb, sem_ch9.adb, sem_ch6.adb, exp_disp.adb,
g-socket.ads: Minor reformatting.
2011-08-30 Thomas Quinot <quinot@adacore.com>
* sem_util.adb: Minor reformatting.
2011-08-30 Tristan Gingold <gingold@adacore.com>
* raise-gcc.c: Never catch exception if _UA_FORCE_UNWIND flag is set,
to be compliant with the ABI.
From-SVN: r178310
Arnaud Charlet [Tue, 30 Aug 2011 13:57:09 +0000 (15:57 +0200)]
[multiple changes]
2011-08-30 Hristian Kirtchev <kirtchev@adacore.com>
* sem_ch12.adb (Check_Private_View): Exchange the private and full view
of a designated type when the related access type is an actual in an
instance. This ensures that the full view of designated type is
available when inside the body of the instance.
2011-08-30 Ed Schonberg <schonberg@adacore.com>
* sem_util.adb (May_Be_Lvalue): To determine whether a reference may be
in a position to be modified, a slice is treated like an indexed
component.
From-SVN: r178308
Arnaud Charlet [Tue, 30 Aug 2011 13:53:38 +0000 (15:53 +0200)]
[multiple changes]
2011-08-30 Thomas Quinot <quinot@adacore.com>
* s-oscons-tmplt.c, g-socket.ads: Adjust maximum allowed value for
field tv_sec in struct timeval.
2011-08-30 Yannick Moy <moy@adacore.com>
* exp_ch9.adb, exp_disp.adb, sem_ch9.adb, sem_res.adb: Protect several
blocks of code doing full expansion, previously only guarded by
Expander_Active, by anding the test that ALFA_Mode is not set
2011-08-30 Ed Schonberg <schonberg@adacore.com>
* sem_ch6.adb (Analyze_Return_Type): If the return type is incomplete,
add the function to the list of private dependents, for subsequent
legality check on Taft amendment types.
* sem_ch12.adb (Analyze_Formal_Incomplete_Type): Initialize
Private_Dependents, as for other incomplete types.
* sem_util.adb (Wrong_Type): Avoid cascaded errors when a
Taft-amendment type is used as the return type of a function.
From-SVN: r178307
Javier Miranda [Tue, 30 Aug 2011 13:50:19 +0000 (13:50 +0000)]
sem_ch3.adb (Check_Abstract_Overriding): Code cleanup...
2011-08-30 Javier Miranda <miranda@adacore.com>
* sem_ch3.adb (Check_Abstract_Overriding): Code cleanup: replace code
which emits an error by a call to a new routine which report the error.
* exp_ch9.adb (Build_Wrapper_Spec): Build the wrapper even if the
entity does not cover an existing interface.
* errout.ads, errout.adb (Error_Msg_PT): New routine. Used to factorize
code.
* sem_ch6.adb (Check_Conformance): Add specific error for wrappers of
protected procedures or entries whose mode is not conformant.
(Check_Synchronized_Overriding): Code cleanup: replace code which emits
an error by a call to a new routine which report the error.
From-SVN: r178306
Arnaud Charlet [Tue, 30 Aug 2011 13:44:27 +0000 (15:44 +0200)]
[multiple changes]
2011-08-30 Robert Dewar <dewar@adacore.com>
* gnat_rm.texi: Minor change.
* exp_attr_light.adb: Minor reformatting.
2011-08-30 Ed Schonberg <schonberg@adacore.com>
* sem_ch3.adb: Patch inheritance of aspects in
Complete_Private_Subtype, to avoid infinite loop.
From-SVN: r178305
Arnaud Charlet [Tue, 30 Aug 2011 13:41:24 +0000 (15:41 +0200)]
[multiple changes]
2011-08-30 Javier Miranda <miranda@adacore.com>
* sem_ch3.adb (Add_Internal_Interface_Entities): If serious errors have
been reported and a subprogram covering an interface primitive is not
found then skip generating the internal entity. Done to avoid crashing
the frontend.
(Check_Abstract_Overriding): Change text of error of wrong formal of
protected subprogram or entry. Done for consistency to emit exactly the
same error reported by Check_Synchronized_Overriding. In addition, the
error is restricted to protected types (bug found working on AI05-0090)
2011-08-30 Yannick Moy <moy@adacore.com>
* exp_aggr.adb, exp_ch11.adb, exp_prag.adb: Remove early exit during
expansion in Alfa mode.
* exp_ch6.adb, exp_ch6.ads (Expand_Actuals): Make subprogram public.
* exp_light.adb, exp_light.ads: New package defining light expansion.
* expander.adb (Expand): Call light expansion in Alfa mode
* exp_ch6_light.adb, exp_ch6_light.ads: Light expansion of chapter 6
constructs.
* exp_ch7_light.adb, exp_ch7_light.ads: Light expansion of chapter 7
constructs.
* exp_attr_light.adb, exp_attr_light.ads: Light expansion of attributes
* gnat1drv.adb (Adjust_Global_Switches): Comment
2011-08-30 Yannick Moy <moy@adacore.com>
* lib-xref-alfa.adb: Minor refactoring.
2011-08-30 Yannick Moy <moy@adacore.com>
* exp_ch9.adb (Expand_Entry_Barrier): Do not perform expansion in Alfa
mode.
* sem_ch6.adb (Analyze_Subprogram_Body_Helper): Do not perform
expansion in Alfa mode.
* sem_ch9.adb (Analyze_Entry_Body): Do not perform expansion in Alfa
mode.
2011-08-30 Robert Dewar <dewar@adacore.com>
* debug_a.adb: Update comment.
From-SVN: r178304
Robert Dewar [Tue, 30 Aug 2011 13:31:38 +0000 (13:31 +0000)]
exp_ch5.adb, [...]: Minor reformatting
2011-08-30 Robert Dewar <dewar@adacore.com>
* exp_ch5.adb, sem_ch3.adb, sem_ch5.adb, einfo.adb, checks.adb,
sem_util.adb, sem_util.ads, sem_res.adb, s-stposu.adb, sem_attr.adb,
exp_ch4.adb, exp_ch6.adb, s-bbthre.adb, lib-xref-alfa.adb,
sem_ch8.adb, sem_disp.adb, exp_ch3.adb: Minor reformatting
2011-08-30 Robert Dewar <dewar@adacore.com>
* gnat_rm.texi: Add section on C.6(16) implementation advice for pragma
volatile.
From-SVN: r178303
Kirill Yukhin [Tue, 30 Aug 2011 13:28:21 +0000 (13:28 +0000)]
Rename avx2-vmovmskb-2.c to avx2-vpmovmskb-2.c.
2011-08-30 Kirill Yukhin <kirill.yukhin@intel.com>
PR testsuite/50185
* gcc.target/i386/avx2-vmovmskb-2.c: Rename to ...
* gcc.target/i386/avx2-vpmovmskb-2.c: ... this. Update.
From-SVN: r178302
Hristian Kirtchev [Tue, 30 Aug 2011 13:28:16 +0000 (13:28 +0000)]
exp_ch3.adb (Expand_Freeze_Array_Type): Correct the call to Build_Finalization_Master by supplying an insertion node...
2011-08-30 Hristian Kirtchev <kirtchev@adacore.com>
* exp_ch3.adb (Expand_Freeze_Array_Type): Correct the call to
Build_Finalization_Master by supplying an insertion node and enclosing
scope. In its old version, the call did not generate a finalization
master.
(Expand_Freeze_Record_Type): Add local variable Has_AACC. Add code to
recognize anonymous access-to-controlled components. Rewrite the
machinery which creates finalization masters to service anonymous
access-to-controlled components of a record type. In its current state,
only one heterogeneous master is necessary to handle multiple anonymous
components.
(Freeze_Type): Comment reformatting.
* rtsfind.ads: Add RE_Set_Is_Heterogeneous to tables RE_Id and
RE_Unit_Table.
* s-stposu.adb (Allocate_Any_Controlled): Rewrite the machinery which
associates TSS primitive Finalize_Address with either the master itself
or with the internal hash table depending on the mode of operation of
the master.
From-SVN: r178301
Javier Miranda [Tue, 30 Aug 2011 13:23:33 +0000 (13:23 +0000)]
exp_ch3.adb (Make_Eq_If): If the etype of the _parent component is an interface type then do not generate...
2011-08-30 Javier Miranda <miranda@adacore.com>
* exp_ch3.adb (Make_Eq_If): If the etype of the _parent component is an
interface type then do not generate code to compare this component.
Required since they have no components and their equality operator is
abstract.
From-SVN: r178300
Arnaud Charlet [Tue, 30 Aug 2011 13:22:13 +0000 (15:22 +0200)]
[multiple changes]
2011-08-30 Steve Baird <baird@adacore.com>
* sem_util.ads (Deepest_Type_Access_Level): New function; for the type
of a saooaaat (i.e, a stand-alone object of an anonymous access type),
returns the (static) accessibility level of the object. Otherwise, the
same as Type_Access_Level.
(Dynamic_Accessibility_Level): New function; given an expression which
could occur as the rhs of an assignment to a saooaaat (i.e., an
expression of an access-to-object type), return the new value for the
saooaaat's associated Extra_Accessibility object.
(Effective_Extra_Accessibility): New function; same as
Einfo.Extra_Accessibility except that object renames are looked through.
* sem_util.adb
(Deepest_Type_Access_Level): New function; see sem_util.ads description.
(Dynamic_Accessibility_Level): New function; see sem_util.ads
description.
(Effective_Extra_Accessibility): New function; see sem_util.ads
description.
* einfo.ads (Is_Local_Anonymous_Access): Update comments.
(Extra_Accessibility): Update comments.
(Init_Object_Size_Align): New procedure; same as Init_Size_Align
except RM_Size field (which is only for types) is unaffected.
* einfo.adb
(Extra_Accessibility): Expand domain to allow objects, not just formals.
(Set_Extra_Accessibility): Expand domain to allow objects, not just
formals.
(Init_Size): Add assertion that we are not trashing the
Extra_Accessibility attribute of an object.
(Init_Size_Align): Add assertion that we are not trashing the
Extra_Accessibility attribute of an object.
(Init_Object_Size_Align): New procedure; see einfo.ads description.
* sem_ch3.adb (Find_Type_Of_Object): Set Is_Local_Anonymous_Access
differently for the type of a (non-library-level) saooaaat depending
whether Ada_Version < Ada_2012. This is the only point where Ada_Version
is queried in this set of changes - everything else (in particular,
setting of the Extra_Accessibility attribute in exp_ch3.adb) is
driven off of the setting of the Is_Local_Anonymous_Access attribute.
The special treatment of library-level saooaaats is an optimization,
not required for correctnesss. This is based on the observation that the
Ada2012 rules (static and dynamic) for saooaaats turn out to be
equivalent to the
Ada2005 rules in the case of a library-level saooaaat.
* exp_ch3.adb
(Expand_N_Object_Declaration): If Is_Local_Anonymous_Access is
false for the type of a saooaaat, declare and initialize its
accessibility level object and set the Extra_Accessibility attribute
of the saooaaat to refer to this object.
* checks.adb (Apply_Accessibility_Check): Add Ada 2012 saooaaat support.
* exp_ch4.adb (Expand_N_In): Replace some Extra_Accessibility calls with
calls to Effective_Extra_Accessibility in order to support
renames of saooaaats.
(Expand_N_Type_Conversion): Add new local function,
Has_Extra_Accessibility, and call it when determining whether an
accessibility check is needed.
It returns True iff Present (Effective_Extra_Accessibility (Id)) would
evaluate to True (without raising an exception).
* exp_ch5.adb
(Expand_N_Assignment_Statement): When assigning to an
Ada2012
saooaaat, update its associated Extra_Accessibility object (if
it has one). This includes an accessibility check.
* exp_ch6.adb (Add_Call_By_Copy_Code): When parameter copy-back updates
a saooaaat, update its Extra_Accessibility object too (if it
has one).
(Expand_Call): Replace a couple of calls to Type_Access_Level
with calls to Dynamic_Access_Level to handle cases where
passing a literal (any literal) is incorrect.
* sem_attr.adb (Resolve_Attribute): Handle the static accessibility
checks associated with "Saooaat := Some_Object'Access;"; this must
be rejected if Some_Object is declared in a more nested scope
than Saooaat.
* sem_ch5.adb (Analyze_Assignment): Force accessibility checking for an
assignment to a saooaaat even if Is_Local_Anonymous_Access
returns False for its type (indicating a 2012-style saooaaat).
* sem_ch8.adb
(Analyze_Object_Renaming): Replace a call to Init_Size_Align
(which is only appropriate for objects, not types) with a call
of Init_Object_Size_Align in order to avoid trashing the
Extra_Accessibility attribute of a rename (the two attributes
share storage).
* sem_res.adb
(Valid_Conversion) Replace six calls to Type_Access_Level with
calls to Deepest_Type_Access_Level. This is a bit tricky. For an
Ada2012 non-library-level saooaaat, the former returns library level
while the latter returns the (static) accessibility level of the
saooaaat. A type conversion to the anonymous type of a saooaaat
can only occur as part of an assignment to the saooaaat, so we
know that such a conversion must be in a lhs context, so Deepest
yields the result that we need. If such a conversion could occur,
say, as the operand of an equality operator, then this might not
be right. Also add a test so that static accessibilty checks are
performed for converting to a saooaaat's type even if
Is_Local_Anonymous_Access yields False for the type.
2011-08-30 Javier Miranda <miranda@adacore.com>
* sem_disp.adb (Check_Dispatching_Operation): Complete condition that
controls generation of a warning associated with late declaration of
dispatching functions. Required to avoid generating spurious
warnings.
From-SVN: r178299
Bernd Schmidt [Tue, 30 Aug 2011 13:17:40 +0000 (13:17 +0000)]
bb-reorder.c (insert_section_boundary_note): Only do it if we reordered the blocks; i.e.
* bb-reorder.c (insert_section_boundary_note): Only do it if
we reordered the blocks; i.e. not if !optimize_function_for_speed_p.
From-SVN: r178298
Bernd Schmidt [Tue, 30 Aug 2011 13:16:44 +0000 (13:16 +0000)]
i386.c (get_pc_thunk_name): Change prefix to "__x86.get_pc_thunk".
* config/i386/i386.c (get_pc_thunk_name): Change prefix to
"__x86.get_pc_thunk".
his line, and those below, will be ignored--
M ChangeLog
M config/i386/i386.c
From-SVN: r178297
Gary Dismukes [Tue, 30 Aug 2011 13:16:22 +0000 (13:16 +0000)]
sem_ch6.adb (Check_Return_Subtype_Indication): Issue error if the return object has an anonymous access type and the...
2011-08-30 Gary Dismukes <dismukes@adacore.com>
* sem_ch6.adb (Check_Return_Subtype_Indication): Issue error if the
return object has an anonymous access type and the function's type is
a named access type.
* sem_ch8.adb (Analyze_Object_Renaming): Suppress error about renaming
conversions on implicit conversions, since such conversions can occur
for anonymous access cases due to expansion. Issue error for attempt
to rename an anonymous expression as an object of a named access type.
* sem_res.ads (Valid_Conversion): Add defaulted parameter Report_Errs,
to indicate whether this function should report errors on invalid
conversions.
* sem_res.adb (Resolve): For Ada 2012, in the case where the type of
the expression is of an anonymous access type and the expected type is
a named general access type, rewrite the expression as a type
conversion, unless this is an expression of a membership test.
(Valid_Conversion.Error_Msg_N): New procedure that conditions the
calling of Error_Msg_N on new formal Report_Errs.
(Valid_Conversion.Error_Msg_NE): New procedure that conditions the
calling of Error_Msg_NE on new formal Report_Errs.
(Valid_Conversion): Move declaration of this function to the package
spec, to allow calls from membership test processing. For Ada 2012,
enforce legality restrictions on implicit conversions of anonymous
access values to general access types, disallowing such conversions in
cases where the expression has a dynamic accessibility level (access
parameters, stand-alone anonymous access objects, or a component of a
dereference of one of the first two cases).
* sem_type.adb (Covers): For Ada 2012, allow an anonymous access type
in the context of a named general access expected type.
* exp_ch4.adb Add with and use of Exp_Ch2.
(Expand_N_In): Add processing for membership tests applied to
expressions of an anonymous access type. First, Valid_Conversion is
called to check whether the test is statically False, and then the
conversion is expanded to test that the expression's accessibility
level is no deeper than that of the tested type. In the case of
anonymous access-to-tagged types, a tagged membership test is applied
as well.
(Tagged_Membership): Extend to handle access type cases, applying the
test to the designated types.
* exp_ch6.adb (Expand_Call): When creating an extra actual for an
accessibility level, and the actual is a 'Access applied to a current
instance, pass the accessibility level of the type of the current
instance rather than applying Object_Access_Level to the prefix. Add a
??? comment, since this level isn't quite right either (will eventually
need to pass an implicit level parameter to init procs).
From-SVN: r178296
Bernd Schmidt [Tue, 30 Aug 2011 13:15:38 +0000 (13:15 +0000)]
genautomata.c (NO_COMB_OPTION): New macro.
* genautomata.c (NO_COMB_OPTION): New macro.
(no_comb_flag): New static variable.
(gen_automata_option): Handle NO_COMB_OPTION.
(comb_vect_p): False if no_comb_flag.
(add_vect): Move computation of min/max values. Return early if
no_comb_flag.
* doc/md.texi (automata_option): Document no-comb-vect.
From-SVN: r178295
Arnaud Charlet [Tue, 30 Aug 2011 13:12:16 +0000 (15:12 +0200)]
[multiple changes]
2011-08-30 Bob Duff <duff@adacore.com>
* s-taskin.ads: Minor comment fix.
2011-08-30 Ed Schonberg <schonberg@adacore.com>
* sem_res.adb (Check_Parameterless_Call): If the node is a selected
component and the selector is a dispatching operation, check if it is
a prefixed call before rewriting as a parameterless function call.
From-SVN: r178294
Christian Bruel [Tue, 30 Aug 2011 11:18:49 +0000 (13:18 +0200)]
fix -fbranch-probabilities and adjust testsuite
From-SVN: r178289
Paolo Carlini [Tue, 30 Aug 2011 08:57:52 +0000 (08:57 +0000)]
error_constants.h: Fix commas.
2011-08-30 Paolo Carlini <paolo.carlini@oracle.com>
* config/os/mingw32/error_constants.h: Fix commas.
From-SVN: r178286
Jason Merrill [Tue, 30 Aug 2011 04:30:58 +0000 (00:30 -0400)]
re PR c++/50224 ([C++0x] bogus unused parameter warning (it is referenced in lambda))
PR c++/50224
* semantics.c (finish_id_expression): Mark captured variables used.
From-SVN: r178277
Jakub Jelinek [Tue, 30 Aug 2011 04:30:42 +0000 (06:30 +0200)]
re PR c++/50207 (G++ segv's on reduced test case)
PR c++/50207
* class.c (finish_struct_1): Complain if the first field is
artificial.
Co-Authored-By: Jason Merrill <jason@redhat.com>
From-SVN: r178276
Jason Merrill [Tue, 30 Aug 2011 04:30:27 +0000 (00:30 -0400)]
re PR c++/50209 ([C++0x] Braced-init-lists are rejected as function default arguments)
PR c++/50209
Core DR 994
* parser.c (cp_parser_default_argument): Use
cp_parser_initializer_clause.
(cp_parser_late_parsing_default_args): Likewise.
From-SVN: r178275
GCC Administrator [Tue, 30 Aug 2011 00:18:54 +0000 (00:18 +0000)]
Daily bump.
From-SVN: r178269
Benjamin Kosnik [Mon, 29 Aug 2011 23:25:17 +0000 (23:25 +0000)]
constexpr.cc: Add.
2011-08-29 Benjamin Kosnik <bkoz@redhat.com>
* testsuite/20_util/duration/arithmetic/constexpr.cc: Add.
From-SVN: r178265