2014-07-29 Robert Dewar <dewar@adacore.com>
* sem_aggr.adb (Resolve_Array_Aggregate): Change Is_Static_Range
to Is_OK_Static_Range.
* sem_attr.adb (Eval_Attribute): Make sure we properly flag
static attributes (Eval_Attribute, case Size): Handle size of
zero properly (Eval_Attribute, case Value_Size): Handle size of
zero properly.
* sem_ch13.adb: Minor reformatting.
* sem_ch3.adb (Process_Range_Expr_In_Decl): Change
Is_Static_Range to Is_OK_Static_Range.
* sem_eval.adb (Eval_Case_Expression): Total rewrite, was
wrong in several ways (Is_Static_Range): Moved here from spec
(Is_Static_Subtype): Moved here from spec Change some incorrect
Is_Static_Subtype calls to Is_OK_Static_Subtype.
* sem_eval.ads: Add comments to section on
Is_Static_Expression/Raises_Constraint_Error (Is_OK_Static_Range):
Add clarifying comments (Is_Static_Range): Moved to body
(Is_Statically_Unevaluated): New function.
* sem_util.ads, sem_util.adb (Is_Preelaborable_Expression): Change
Is_Static_Range to Is_OK_Static_Range.
* sinfo.ads: Additional commments for Is_Static_Expression noting
that clients should almost always use Is_OK_Static_Expression
instead. Many other changes throughout front end units to obey
this rule.
* tbuild.ads, tbuild.adb (New_Occurrence_Of): Set Is_Static_Expression
for enumeration literal.
* exp_ch5.adb, sem_intr.adb, sem_ch5.adb, exp_attr.adb, exp_ch9.adb,
lib-writ.adb, sem_ch9.adb, einfo.ads, checks.adb, checks.ads,
sem_prag.adb, sem_ch12.adb, freeze.adb, sem_res.adb, exp_ch4.adb,
exp_ch6.adb, sem_ch4.adb, sem_ch6.adb, exp_aggr.adb, sem_cat.adb:
Replace all occurrences of Is_Static_Expression by
Is_OK_Static_Expression.
From-SVN: r213159
2014-07-29 Hristian Kirtchev <kirtchev@adacore.com>
* exp_ch4.adb (Process_Transient_Object): Remove constant
In_Cond_Expr, use its initialization expression in place.
* exp_ch7.adb (Process_Declarations): There is no need to check
that a transient object being hooked is controlled as it would
not have been hooked in the first place.
* exp_util.adb (Is_Aliased): 'Reference-d or renamed transient
objects are not considered aliased when the related context is
a Boolean expression_with_actions.
(Requires_Cleanup_Actions): There is no need to check that a transient
object being hooked is controlled as it would not have been hooked in
the first place.
From-SVN: r213158
2014-07-29 Robert Dewar <dewar@adacore.com>
* errout.adb: Minor reformatting.
2014-07-29 Hristian Kirtchev <kirtchev@adacore.com>
* exp_ch3.adb (Default_Initialize_Object): Add new variables
Abrt_Blk and Dummy. Generate a dummy temporary when aborts are
not allowed to ensure the symmetrical generation of symbols.
* exp_ch7.adb (Build_Object_Declarations): Remove variables A_Expr
and E_Decl. Add new variables Decl and Dummy. Generate a dummy
temporary when aborts are not allowed to ensure symmertrical
generation of symbols.
* exp_intr.adb (Expand_Unc_Deallocation): Add new variable
Dummy. Generate a dummy temporary when aborts are not allowed
to ensure symmertrical generation of symbols.
2014-07-29 Ed Schonberg <schonberg@adacore.com>
* exp_dbug.adb (Debug_Renaming_Declaration): For an object
renaming, indicate that the renamed entity itself needs debug
information. This is necessary if that entity is a temporary,
e.g. part of the expansion of an explicit dereference in an
iterator.
From-SVN: r213157
2014-07-29 Thomas Quinot <quinot@adacore.com>
* errout.adb (Set_Error_Posted): When propagating flag to
an enclosing named association, also propagate to the parent
of that node, so that named and positional associations are
treated consistently.
2014-07-29 Ed Schonberg <schonberg@adacore.com>
* sem_attr.adb (Resolve_Attribute, case 'Update): Set
Do_Range_Check properly on array component expressions that
have a scalar type. In GNATprove mode, only checks on scalar
components must be marked by the front-end.
2014-07-29 Ed Schonberg <schonberg@adacore.com>
* sem_res.adb (Resolve_Type_Conversion): If the type of the
expression is a limited view, use the non-limited view when
available.
2014-07-29 Hristian Kirtchev <kirtchev@adacore.com>
* exp_ch4.adb (Expand_N_Case_Expression): Mark the generated
case statement as coming from a conditional expression.
(Expand_N_If_Expression): Mark the generated if statement as
coming from a conditional expression.
* exp_ch5.adb (Expand_N_Case_Statement): Do not process controlled
objects found in case statement alternatives when the case
statement is actually a case expression.
(Expand_N_If_Statement):
Do not process controlled objects found in an if statement when
the if statement is actually an if expression.
* sinfo.adb (From_Conditional_Expression): New routine.
(Set_From_Conditional_Expression): New routine.
* sinfo.ads Add new semantic flag From_Conditional_Expression and
update related nodes.
(From_Conditional_Expression): New routine along with pragma Inline.
(Set_From_Conditional_Expression): New routine along with pragma Inline.
From-SVN: r213156
2014-07-29 Hristian Kirtchev <kirtchev@adacore.com>
* exp_ch7.adb (Build_BIP_Cleanup_Stmts): Remove
formal parameter Obj_Id and update the comment on usage. Renamed
Obj_Typ to Func_Typ and update all occurrences.
(Find_Last_Init): Remove formal parameter Decl and update the comment
on usage.
Remove local constants Obj_Id and Obj_Typ. Remove local variables
Init_Typ and Is_Conc. Remove the extraction of the initialization type.
(Find_Last_Init_In_Block): Remove formal parameter
Init_Typ and update the comment on usage.
(Is_Init_Call): Remove formal parameter Init_Typ and update the comment
on usage. Check whether the procedure call is an initialization
procedure of either the object type or the initialization type.
(Is_Init_Proc_Of): New routine.
(Process_Object_Declaration): Obj_Id and Obj_Typ are now global to this
routine. Add new variable Init_Typ. Add circuitry to extract the object
type as well as the initialization type.
2014-07-29 Robert Dewar <dewar@adacore.com>
* sem_case.adb: Minor reformatting.
* sem_aux.ads: Minor reformatting.
2014-07-29 Ed Schonberg <schonberg@adacore.com>
* sinfo.adb (Set_Else_Actions, Set_Then_Actions): Set parent
pointer on these fields, even though they are semantic, because
subsequent analysis and expansion of action nades may require
exploring the tree, for example to locate a node to be wrapped
when a function with controlled result is called.
2014-07-29 Claire Dross <dross@adacore.com>
* sem_aux.adb (Get_Binary_Nkind): Use case on
Name_Id instead of an intermediate string.
(Get_Unary_Nkind): Use case on Name_Id instead of an intermediate
string.
2014-07-29 Sergey Rybin <rybin@adacore.com frybin>
* gnat_ugn.texi (gnatelim, gnatstub, gnatmetric): Add note
about processing sources with preprocessor directives.
From-SVN: r213155
2014-07-29 Martin Jambor <mjambor@suse.cz>
* tree-sra.c (sra_modify_constructor_assign): Change type of stmt
parameter to gimple.
(sra_modify_assign): Likewise.
From-SVN: r213154
2014-07-28 Trevor Saunders <tsaunders@mozilla.com>
* config/i386/i386.c (ix86_return_in_memory): replace one
ATTRIBUTE_UNUSED where the attribute can actually sometimes be unused.
From-SVN: r213142
PR middle-end/61734
* fold-const.c (fold_comparison): Disable X - Y CMP 0 to X CMP Y for
operators other than the equality operators.
From-SVN: r213118
2014-07-28 Richard Biener <rguenther@suse.de>
PR middle-end/52478
* optabs.c (gen_int_libfunc): For -ftrapv libfuncs make
sure to register SImode ones, not only >= word_mode ones.
* expr.c (expand_expr_real_2): When expanding -ftrapv
binops do not use OPTAB_LIB_WIDEN.
* gcc.dg/torture/ftrapv-1.c: New testcase.
From-SVN: r213117
2014-07-28 Richard Biener <rguenther@suse.de>
PR tree-optimization/61921
* tree-ssa-structalias.c (create_variable_info_for_1): Check
if there is a varpool node before dereferencing it.
* gfortran.dg/pr61921.f90: New testcase.
From-SVN: r213114
* graphite-sese-to-poly.c:
(new_pbb_from_pbb): Set a new id of pbb1->domain (instead of using the
id of the pbb), which contains pointer to the pbb1.
[gcc/testsuite]
* gcc.dg/graphite/isl-ast-gen-if-2.c: New testcase.
From-SVN: r213110
* graphite-isl-ast-to-gimple.c:
(graphite_create_new_guard): New function.
(translate_isl_ast_node_if): New function.
(translate_isl_ast): Add calling of translate_isl_ast_node_if.
[gcc/testsuite]
* gcc.dg/graphite/isl-ast-gen-if-1.c: New testcase.
From-SVN: r213109
PR c/61861
* macro.c (builtin_macro): Add location parameter. Set
location of builtin macro to the expansion point.
(enter_macro_context): Pass location to builtin_macro.
* gcc.dg/pr61861.c: New test.
From-SVN: r213102
gcc/testsuite/
PR rtl-optimization/61926
* gcc.target/mips/const-anchor-1.c, gcc.target/mips/const-anchor-2.c:
Reverse argument order.
* gcc.target/mips/const-anchor-3.c, gcc.target/mips/const-anchor-4.c:
New XFAILed tests that match the original order.
From-SVN: r213093
2014-07-26 Andrew Pinski <apinski@cavium.com>
* config/aarch64/aarch64.md (*extr_insv_lower_reg<mode>): Remove +
from the read only register.
From-SVN: r213090
2014-07-26 Tobias Burnus <burnus@net-b.de>
* check.c (gfc_check_sizeof): Permit for assumed type if and
only if it has an array descriptor.
* intrinsic.c (do_ts29113_check): Permit SIZEOF.
(add_functions): SIZEOF is an Inquiry function.
* intrinsic.texi (SIZEOF): Add note that only contiguous
arrays are permitted.
* trans-expr.c (gfc_conv_intrinsic_to_class): Handle assumed
rank.
* trans-intrinsic.c (gfc_conv_intrinsic_sizeof): Handle
assumed type + array descriptor, CLASS and assumed rank.
(gfc_conv_intrinsic_storage_size): Handle class arrays.
2014-07-26 Tobias Burnus <burnus@net-b.de>
* gfortran.dg/sizeof_2.f90: Change dg-error.
* gfortran.dg/sizeof_4.f90: New.
* gfortran.dg/storage_size_1.f08: Correct expected
value.
From-SVN: r213079
cp/
2014-07-25 Edward Smith-Rowland <3dw4rd@verizon.net>
Implement N4051 - Allow typename in a template template parameter
* parser.c (cp_parser_type_parameter_key): New funtion;
(cp_parser_token_is_type_parameter_key): Ditto;
(cp_parser_type_parameter): Look for type-parameter-key for all versions
but pedwarn for less than cxx1z.
testsuite/
2014-07-25 Edward Smith-Rowland <3dw4rd@verizon.net>
Implement N4051 - Allow typename in a template template parameter
* lib/target-supports.exp (check_effective_target_c++1y): Now
means C++1y and up.
(check_effective_target_c++1y_down): New.
(check_effective_target_c++1z_only): New.
(check_effective_target_c++1z): New.
* g++.dg/cpp1z/typename-tmpl-tmpl-parm.C: New.
* g++.dg/cpp1z/typename-tmpl-tmpl-parm-neg.C: New.
* g++.dg/cpp1z/typename-tmpl-tmpl-parm-.C: New.
From-SVN: r213065
* lto-streamer-out.c (struct sccs): Turn to ...
(class DFS): ... this one; refactor the DFS walk so it can
be re-done on per-SCC basis.
(DFS::DFS): New constructor.
(DFS::~DFS): New destructor.
(hash_tree): Add new MAP argument holding in-SCC hash values;
remove POINTER_TYPE hashing hack.
(scc_entry_compare): Rename to ...
(DFS::scc_entry_compare): ... this one.
(hash_scc): Rename to ...
(DFS::hash_scc): ... this one; pass output_block instead
of streamer_cache; work harder to get unique and stable SCC
hashes.
(DFS_write_tree): Rename to ...
(DFS::DFS_write_tree): ... this one; add SINGLE_P parameter.
(lto_output_tree): Update.
Co-Authored-By: Richard Biener <rguenther@suse.de>
From-SVN: r213059
No substantial changes, although the hash values will be slightly
different.
v2: add_flag moved to inchash. Some minor changes.
gcc/:
2014-07-25 Andi Kleen <ak@linux.intel.com>
* lto-streamer-out.c (hash_tree): Convert to inchash.
From-SVN: r213057
