Jan Hubicka [Sat, 9 Aug 2014 22:07:55 +0000 (00:07 +0200)]
devirt-35.C: Fix template.
* g++.dg/ipa/devirt-35.C: Fix template.
* g++.dg/ipa/devirt-36.C: Likewise.
* g++.dg/ipa/devirt-37.C: New testcase.
* ipa-devirt.c (get_dynamic_type): Handle case when instance is in
DECL correctly; do not give up on types in static storage.
Ulrich Drepper [Sat, 9 Aug 2014 17:56:57 +0000 (17:56 +0000)]
random.tcc (uniform_on_sphere_helper): Define.
* include/ext/random.tcc (uniform_on_sphere_helper): Define.
(uniform_on_sphere_distribution::operator()): Use the new helper
class for the implementation.
* testsuite/ext/random/uniform_on_sphere_distribution/operators/
equal.cc: Remove bogus part of comment.
* testsuite/ext/random/uniform_on_sphere_distribution/operators/
inequal.cc: Likewise.
* testsuite/ext/random/uniform_on_sphere_distribution/operators/
serialize.cc: Add check to verify result of serialzation and
deserialization.
* testsuite/ext/random/uniform_on_sphere_distribution/operators/
generate.cc: New file.
François Dumont [Sat, 9 Aug 2014 08:00:59 +0000 (08:00 +0000)]
re PR libstdc++/61667 (setting max_load_factor of unordered_map cause buckets shrink)
2014-08-09 François Dumont <fdumont@gcc.gnu.org>
PR libstdc++/61667
* include/bits/hashtable.h (_Hashtable<>::__rehash_policy): Use
_M_need_rehash to initialize the rehash policy and check if a rehash is
needed.
* testsuite/23_containers/unordered_map/modifiers/61667.cc: New.
Cary Coutant [Fri, 8 Aug 2014 20:33:26 +0000 (20:33 +0000)]
Remove skeleton type units that were being produced with -gsplit-dwarf.
These sections were originally intended as targets for .gdb_index
entries that needed to point to type units. Because of the limitations
of the .debug_gnu_pubnames/pubtypes sections with split DWARF, we were
not able to pass along enough information to the gold linker to generate
those index entries properly, and they had to point to the CU instead.
GDB had to deal with that, and was updated a while ago to no longer
depend on the skeleton TU sections at all. This allows us to reduce
object file sizes with split DWARF by about 30%.
* calls.c (precompute_arguments): Check
promoted_for_signed_and_unsigned_p and set the promoted mode.
(promoted_for_signed_and_unsigned_p): New function.
(expand_expr_real_1): Check promoted_for_signed_and_unsigned_p
and set the promoted mode.
* expr.h (promoted_for_signed_and_unsigned_p): New function definition.
* cfgexpand.c (expand_gimple_stmt_1): Call emit_move_insn if
SUBREG is promoted with SRP_SIGNED_AND_UNSIGNED.
* calls.c (precompute_arguments): Use new SUBREG_PROMOTED_SET
instead of SUBREG_PROMOTED_UNSIGNED_SET.
(expand_call): Likewise.
* cfgexpand.c (expand_gimple_stmt_1): Use SUBREG_PROMOTED_SIGN
to get promoted mode.
* combine.c (record_promoted_value): Skip > 0 comparison with
SUBREG_PROMOTED_UNSIGNED_P as it now returns only 0 or 1.
* expr.c (convert_move): Use SUBREG_CHECK_PROMOTED_SIGN instead
of SUBREG_PROMOTED_UNSIGNED_P.
(convert_modes): Likewise.
(store_expr): Use SUBREG_PROMOTED_SIGN to get promoted mode.
Use SUBREG_CHECK_PROMOTED_SIGN instead of SUBREG_PROMOTED_UNSIGNED_P.
(expand_expr_real_1): Use new SUBREG_PROMOTED_SET instead of
SUBREG_PROMOTED_UNSIGNED_SET.
* function.c (assign_parm_setup_reg): Use new SUBREG_PROMOTED_SET
instead of SUBREG_PROMOTED_UNSIGNED_SET.
* ifcvt.c (noce_emit_cmove): Updated to use SUBREG_PROMOTED_GET and
SUBREG_PROMOTED_SET.
* internal-fn.c (ubsan_expand_si_overflow_mul_check): Use
SUBREG_PROMOTED_SET instead of SUBREG_PROMOTED_UNSIGNED_SET.
* optabs.c (widen_operand): Use SUBREG_CHECK_PROMOTED_SIGN instead
of SUBREG_PROMOTED_UNSIGNED_P.
* rtl.h (SUBREG_PROMOTED_UNSIGNED_SET): Remove.
(SUBREG_PROMOTED_SET): New define.
(SUBREG_PROMOTED_GET): Likewise.
(SUBREG_PROMOTED_SIGN): Likewise.
(SUBREG_PROMOTED_SIGNED_P): Likewise.
(SUBREG_CHECK_PROMOTED_SIGN): Likewise.
(SUBREG_PROMOTED_UNSIGNED_P): Updated.
* rtlanal.c (unsigned_reg_p): Use new SUBREG_PROMOTED_GET
instead of SUBREG_PROMOTED_UNSIGNED_GET.
(nonzero_bits1): Skip > 0 comparison with the results as
SUBREG_PROMOTED_UNSIGNED_P now returns only 0 or 1.
(num_sign_bit_copies1): Use SUBREG_PROMOTED_SIGNED_P instead
of !SUBREG_PROMOTED_UNSIGNED_P.
* simplify-rtx.c (simplify_unary_operation_1): Use new
SUBREG_PROMOTED_SIGNED_P instead of !SUBREG_PROMOTED_UNSIGNED_P.
(simplify_subreg): Use new SUBREG_PROMOTED_SIGNED_P,
SUBREG_PROMOTED_UNSIGNED_P and SUBREG_PROMOTED_SET instead of
SUBREG_PROMOTED_UNSIGNED_P and SUBREG_PROMOTED_UNSIGNED_SET.
Jan Hubicka [Thu, 7 Aug 2014 20:58:17 +0000 (22:58 +0200)]
ipa-devirt.c: Include gimple-pretty-print.h
* ipa-devirt.c: Include gimple-pretty-print.h
(referenced_from_vtable_p): Exclude DECL_EXTERNAL from
further tests.
(decl_maybe_in_construction_p): Fix conditional on cdtor check
(get_polymorphic_call_info): Fix return value
(type_change_info): New sturcture based on ipa-prop
variant.
(noncall_stmt_may_be_vtbl_ptr_store): New predicate
based on ipa-prop variant.
(extr_type_from_vtbl_ptr_store): New function
based on ipa-prop variant.
(record_known_type): New function.
(check_stmt_for_type_change): New function.
(get_dynamic_type): New function.
* ipa-prop.c (ipa_analyze_call_uses): Use get_dynamic_type.
* tree-ssa-pre.c: ipa-utils.h
(eliminate_dom_walker::before_dom_children): Use ipa-devirt
machinery; sanity check with ipa-prop devirtualization.
* trans-mem.c (ipa_tm_insert_gettmclone_call): Clear
polymorphic flag.
* g++.dg/ipa/devirt-35.C: New testcase.
* g++.dg/ipa/devirt-36.C: New testcase.
Jonathan Wakely [Thu, 7 Aug 2014 20:13:52 +0000 (21:13 +0100)]
future (_State_baseV2::_M_set_result): Pass pointers to _M_do_set.
* include/std/future (_State_baseV2::_M_set_result): Pass pointers to
_M_do_set.
(_State_baseV2::_M_do_set): Change parameters to pointers.
(_State_baseV2::_Setter): Change _M_arg from reference to pointer.
(_State_baseV2::__setter): Initialize _Setter with pointers.
(_State_baseV2::__setter(promise<void>*)): Remove overload.
(promise::set_value, promise::set_exception): Pass setter directly
to _M_set_result.
(_State_baseV2::_Task_setter): Add template parameter for callable
type and replace std::function member with pointer to that type.
Change _M_result member from reference to pointer.
(_State_baseV2::_S_task_setter): Change parameter to lvalue reference
and initialize _Task_setter with pointers.
(__location_invariant): Specialize for _Setter and _Task_setter.
Paolo Carlini [Thu, 7 Aug 2014 19:51:28 +0000 (19:51 +0000)]
re PR c++/51312 ([C++0x] Wrong interpretation of converted constant expressions (for enumerator initializers))
/cp
2014-08-07 Paolo Carlini <paolo.carlini@oracle.com>
PR c++/51312
* decl.c (build_enumerator): Handle class types with conversion
operators via perform_implicit_conversion_flags and
build_expr_type_conversion.
* cvt.c (build_expr_type_conversion): Replace pair of errors
with error + inform.
/testsuite
2014-08-07 Paolo Carlini <paolo.carlini@oracle.com>
Maxim Kuvyrkov [Thu, 7 Aug 2014 11:16:29 +0000 (11:16 +0000)]
Add statistical printout of rank_for_schedule decisions
* haifa-sched.c (SCHED_SORT): Delete. Macro used exactly once.
(enum rfs_decition:RFS_*): New constants wrapped in an enum.
(rfs_str): String corresponding to RFS_* constants.
(rank_for_schedule_stats_t): New typedef.
(rank_for_schedule_stats): New static variable.
(rfs_result): New static function.
(rank_for_schedule): Track statistics for deciding heuristics.
(rank_for_schedule_stats_diff, print_rank_for_schedule_stats): New
static functions.
(ready_sort): Use them for debug printouts.
(schedule_block): Init statistics state. Print statistics on
rank_for_schedule decisions.
Tom de Vries [Thu, 7 Aug 2014 07:22:24 +0000 (07:22 +0000)]
For cproj test-case, use glibc version xfail
2014-08-04 Tom de Vries <tom@codesourcery.com>
* gcc.dg/cproj-fails-with-broken-glibc.c: Use xfail for broken glibc
version instead of required-target.
* lib/target-supports.exp
(check_effective_target_not_glibc_2_11_or_earlier): Replace by ...
(check_effective_target_glibc_2_11_or_earlier): ... this.
Jason Merrill [Thu, 7 Aug 2014 01:44:40 +0000 (21:44 -0400)]
init.c (build_vec_init): Fix constant initialization of trailing elements.
* init.c (build_vec_init): Fix constant initialization of
trailing elements.
(build_value_init_noctor): Call maybe_constant_init.
* semantics.c (maybe_constant_init): See through EXPR_STMT and
conversion to void.
David Malcolm [Tue, 5 Aug 2014 21:01:47 +0000 (21:01 +0000)]
gdbhooks.py: add "break-on-pass" command
gcc/
2014-08-05 David Malcolm <dmalcolm@redhat.com>
* gdbhooks.py (find_gcc_source_dir): New helper function.
(class PassNames): New class, locating and parsing passes.def.
(class BreakOnPass): New command "break-on-pass".
Trevor Saunders [Tue, 5 Aug 2014 19:52:08 +0000 (19:52 +0000)]
fix pr62009 use after free in redirect_edge_var_map_dup
The change to get the entry for the old edge before inserting the new
one was incorrect because if inserting the new one resized the table
then the pointer to the entry for the old one would become invalid.
gcc/
* tree-ssa.c (redirect_edge_var_map_dup): insert newe before
getting olde.
* config/aarch64/aarch64-builtins.c
(aarch64_simd_builtin_type_mode): Delete.
(v8qi_UP): Remap to V8QImode.
(v4hi_UP): Remap to V4HImode.
(v2si_UP): Remap to V2SImode.
(v2sf_UP): Remap to V2SFmode.
(v1df_UP): Remap to V1DFmode.
(di_UP): Remap to DImode.
(df_UP): Remap to DFmode.
(v16qi_UP):V16QImode.
(v8hi_UP): Remap to V8HImode.
(v4si_UP): Remap to V4SImode.
(v4sf_UP): Remap to V4SFmode.
(v2di_UP): Remap to V2DImode.
(v2df_UP): Remap to V2DFmode.
(ti_UP): Remap to TImode.
(ei_UP): Remap to EImode.
(oi_UP): Remap to OImode.
(ci_UP): Map to CImode.
(xi_UP): Remap to XImode.
(si_UP): Remap to SImode.
(sf_UP): Remap to SFmode.
(hi_UP): Remap to HImode.
(qi_UP): Remap to QImode.
(aarch64_simd_builtin_datum): Make mode a machine_mode.
(VAR1): Build builtin name.
(aarch64_init_simd_builtins): Remove dead code.
Jakub Jelinek [Tue, 5 Aug 2014 08:09:00 +0000 (10:09 +0200)]
loop-unroll.c (struct iv_to_split): Remove n_loc and loc fields.
* loop-unroll.c (struct iv_to_split): Remove n_loc and loc fields.
(analyze_iv_to_split_insn): Don't initialize them.
(get_ivts_expr): Removed.
(allocate_basic_variable, insert_base_initialization): Use
SET_SRC instead of *get_ivts_expr.
(split_iv): Use &SET_SRC instead of get_ivts_expr.
Roman Gareev [Tue, 5 Aug 2014 03:41:29 +0000 (03:41 +0000)]
graphite-isl-ast-to-gimple.c: Add a new struct ast_build_info.
[gcc/]
* graphite-isl-ast-to-gimple.c: Add a new struct ast_build_info.
(translate_isl_ast_for_loop): Add checking of the
flag_loop_parallelize_all.
(ast_build_before_for): New function.
(scop_to_isl_ast): Add checking of the
flag_loop_parallelize_all.
* graphite-dependences.c: Move the defenition of the
scop_get_dependences from graphite-optimize-isl.c to this file.
(apply_schedule_on_deps): Add checking of the ux's emptiness.
(carries_deps): Add checking of the x's value.
* graphite-optimize-isl.c: Move the defenition of the
scop_get_dependences to graphite-dependences.c.
* graphite-poly.h: Add declarations of scop_get_dependences
and carries_deps.
re PR go/61308 (gccgo: ICE in Expression::check_bounds [GoSmith])
PR go/61308
compiler: Handle enclosing vars for function type in function lit.
This fixes a dumb bug in which the enclosing vars were
incorrectly cleared when a function literal contains a
reference to a function type. The test for this will go into
the master repository in the change at
http://codereview.appspot.com/121200043 .
re PR go/61308 (gccgo: ICE in Expression::check_bounds [GoSmith])
PR go/61308
PR go/61866
compiler: Don't cast index expr to int before bounds check.
This fixes http://gcc.gnu.org/PR61866 : on a 32-bit system,
casting an int64 index to int drops the upper 32 bits of the
value, and thus can cause an out-of-range index to appear to
be in range.
This undoes part of change 1318:fa6e0c716dba
(https://codereview.appspot.com/104610044) and therefore
breaks http://gcc.gnu.org/PR61308 again. I have a separate
patch for that (http://codereview.appspot.com/122020043). In
addition to undoing part of that change, this patch adds code
to avoid a compiler crash. This changes PR61308 from a
compiler crash to an incorrect error message.
re PR middle-end/60102 (powerpc fp-bit ices at dwf_regno)
PR target/60102
[libgcc]
2014-07-31 Rohit <rohitarulraj@freescale.com>
* config/rs6000/linux-unwind.h (ppc_fallback_frame_state): Update
based on change in SPE high register numbers and 3 HTM registers.
[gcc]
2014-07-31 Rohit <rohitarulraj@freescale.com>
* config/rs6000/rs6000.c
(rs6000_reg_names) : Add SPE high register names.
(alt_reg_names) : Likewise.
(rs6000_dwarf_register_span) : For SPE high registers, replace
dwarf register numbers with GCC hard register numbers.
(rs6000_init_dwarf_reg_sizes_extra) : Likewise.
(rs6000_dbx_register_number): For SPE high registers, return dwarf
register number for the corresponding GCC hard register number.
* config/rs6000/rs6000.h
(FIRST_PSEUDO_REGISTER) : Update based on 32 newly added GCC hard
register numbers for SPE high registers.
(DWARF_FRAME_REGISTERS) : Likewise.
(DWARF_REG_TO_UNWIND_COLUMN) : Likewise.
(DWARF_FRAME_REGNUM) : Likewise.
(FIXED_REGISTERS) : Likewise.
(CALL_USED_REGISTERS) : Likewise.
(CALL_REALLY_USED_REGISTERS) : Likewise.
(REG_ALLOC_ORDER) : Likewise.
(enum reg_class) : Likewise.
(REG_CLASS_NAMES) : Likewise.
(REG_CLASS_CONTENTS) : Likewise.
(SPE_HIGH_REGNO_P) : New macro to identify SPE high registers.
Jakub Jelinek [Mon, 4 Aug 2014 15:45:50 +0000 (17:45 +0200)]
task.c (GOMP_taskgroup_end): If taskgroup->num_children is not zero...
* task.c (GOMP_taskgroup_end): If taskgroup->num_children
is not zero, but taskgroup->children is NULL and there are
any task->children, schedule those instead of waiting.
* testsuite/libgomp.c/depend-6.c: New test.
* testsuite/libgomp.c/depend-7.c: New test.
* testsuite/libgomp.c/depend-8.c: New test.
* testsuite/libgomp.c/depend-9.c: New test.
* testsuite/libgomp.c/depend-10.c: New test.
projects / thirdparty / gcc.git / log
