2004-08-13 Paolo Carlini <pcarlini@suse.de>
* include/std/std_fstream.h (class basic_ifstream,
class basic_ofstream, class basic_fstream): Add const overloads
of is_open, as per DR 365 [WP].
* docs/html/ext/howto.html: Add an entry for DR 365.
From-SVN: r85948
* genattrtab.c (insn_ent): Replace insn_code, insn_index and lineno
fields with a pointer to the instruction definition.
(get_attr_value, fill_attr, make_length_attrs, remove_insn_ent)
(insert_insn_ent, simplify_test_exp, optimize_attrs, write_attr_get)
(write_attr_case, write_const_num_delay_slots): Update accordingly.
(write_attr_case, write_const_num_delay_slots): Write the name of
an insn next to its case statement.
* genoutput.c (data): Add a filename field.
(gen_insn, gen_peephole, gen_expand, gen_split): Set it.
(output_insn_data): Print the location of each insn definition.
* genrecog.c (write_action): Print the name of an insn above
the statement that returns its code.
From-SVN: r85947
2004-08-13 Olivier Hainque <hainque@act-europe.fr>
* decl.c (gnat_to_gnu_entity) <E_Variable>: When building an allocator
for a global aliased object with a variable size and an unconstrained
nominal subtype, pretend there is no initializer if the one we have is
incomplete, and avoid referencing an inexistant component in there. The
part we have will be rebuilt anyway and the reference may confuse
further operations.
2004-08-13 Thomas Quinot <quinot@act-europe.fr>
* einfo.ads: Minor reformatting
* lib-writ.adb (Output_Main_Program_Line): Do not set parameter
restrictions in the ALI if we only want to warn about violations.
2004-08-13 Vincent Celier <celier@gnat.com>
* ali.adb (Scan_ALI): Initialize component Body_Needed_For_SAL to False
when creating a new Unit_Record in table Units.
* gnatls.adb (Output_Unit): In verbose mode, output the restrictions
that are violated, if any.
* prj-nmsc.adb (Ada_Check.Get_Path_Names_And_Record_Sources): Do not
add directory separator if path already ends with a directory separator.
2004-08-13 Ed Schonberg <schonberg@gnat.com>
* rtsfind.adb (Entity_Not_Defined): If the error ocurrs in a predefined
unit, this is an attempt to inline a construct that is not available in
the current restricted mode, so abort rather than trying to continue.
* sem_ch3.adb (Build_Underlying_Full_View): If the new type has
discriminants that rename those of the parent, recover names of
original discriminants for the constraint on the full view of the
parent.
(Complete_Private_Subtype): Do not create a subtype declaration if the
subtype is an itype.
* gnat_rm.texi: Added section on implementation of discriminated
records with default values for discriminants.
2004-08-13 Ed Schonberg <schonberg@gnat.com>
PR ada/15601
* sem_res.adb (Make_Call_Into_Operator): Handle properly the case where
the second operand is overloaded.
From-SVN: r85934
* genrecog.c (add_to_sequence): When processing a MATCH_PARALLEL,
if pred->singleton != PARALLEL, issue a warning and pretend it was.
Also issue a warning for any predicate we don't know about.
From-SVN: r85933
* dbxout.c (dbxout_begin_prologue): New function.
(dbx_debug_hooks): Use new begin prologue hook.
(dbxout_function_end): Emit N_ENSYM.
* stab.def (N_BNSYM, N_ENSYM): Define and document these two new stabs.
testsuite:
gcc.dg/darwin-20040809-2.c: New test.
From-SVN: r85908
2004-08-12 Janis Johnson <janis187@us.ibm.com>
* doc/extend.texi (AltiVec builtins): Fix description for recent
changes. Update operation list to match current support.
From-SVN: r85902
2004-08-12 Janis Johnson <janis187@us.ibm.com>
* g++.dg/ext/altivec-2.C: Check for hardware support before
executing any VMX instructions.
From-SVN: r85890
PR c++/16273
* class.c (count_depth_data): New type.
(dfs_depth_post): New function.
(dfs_depth_q): Likewise.
(find_final_overrider_data_s): Change type of vpath.
Add vpath_list.
(dfs_find_final_overrider_1): New function.
(dfs_find_final_overrider): Use it.
(dfs_find_final_overrider_q): Adjust use of vpath.
(dfs_find_final_overrider_post): Likewise.
(find_final_overrider): Use dfs_depth. Allocate and deallocate
vpath_list.
From-SVN: r85886
PR libgcj/16662
* java/lang/natObject.cc (LOCK_LOG, LOG): Add debug tracing.
(Almost everywhere): add LOG calls, fix, add comments.
(_Jv_MonitorEnter): Replace masking of LOCKED bit with assertion.
Add explicit check for LOCKED bit in slow case (PR 16662).
(_Jv_MonitorExit): Add casts in debug-only code.
Always release LOCKED bit before throwing exception.
(_Jv_ObjectCheckMonitor): Lock may be held if lightweight lock
isn't. Handle easy cases without lock acquisition.
(Object::wait): Use NotifyAll for lock inflation.
From-SVN: r85884
2004-08-12 Andreas Krebbel <krebbel1@de.ibm.com>
* config/s390/s390.c (struct s390_frame_layout): New struct as element
of struct machine_function.
(cfun->machine->frame_size): Moved into cfun->machine->frame_layout and
changed all uses.
(cfun->machine->save_fprs_p): Replaced by cfun_save_high_fprs and
changed all uses.
(cfun_frame_layout, cfun_save_high_fprs_p, cfun_gprs_save_area_size)
(cfun_set_fpr_bit, cfun_fpr_bit_p): New macros.
(s390_frame_area, s390_register_info): New functions.
(s390_optimize_prolog): Renamed to s390_optimize_prologue. Added check
for base register.
(s390_return_addr_rtx, s390_return_address_offset)
(s390_va_start, s390_gimplify_va_arg)
(s390_emit_prologue, s390_emit_epilogue): Adjusted for new stack
layouts.
(s390_frame_info): Functionality partly moved to s390_register_info.
Made adaptions for new stack layout.
(save_gprs, restore_gprs): Changed meaning of second parameter and
adapted all callers.
* config/s390/s390.h (s390_backchain_string): New global variable.
(MASK_BACKCHAIN): Removed definition.
(TARGET_BACKCHAIN): Changed check.
(TARGET_KERNEL_BACKCHAIN): New macro.
(TARGET_SWITCHES): Removed entries of "backchain" and "no-backchain".
(TARGET_OPTIONS): Added "backchain", "no-backchain" and
"kernel-backchain".
(DYNAMIC_CHAIN_ADDRESS): Adjusted for new stack layouts.
* config/s390/s390.md ("allocate_stack"): Added TARGET_KERNEL_BACKCHAIN
as condition. Adjusted for new stack layout.
* doc/invoke.texi: Added documentation for new option
"-mkernel-backchain" and adjusted documentation of "-mbackchain" and
"-mno-backchain".
From-SVN: r85882
* expr.c (move_by_pieces): Set alignment for move to minimum of
MOVE_MAX_PIECES mode alignment and the largest non-slow mode
alignment, but not less than the original alignment.
(move_by_pieces_ninsns): Same.
(can_store_by_pieces): Similar for store with STORE_MAX_PIECES.
(store_by_pieces_1): Same.
From-SVN: r85875
PR tree-optimization/16867
* tree.c (is_global_var): New function.
(needs_to_live_in_memory): Check for TREE_ADDRESSABLE.
Call is_global_var.
* tree.h (DECL_NEEDS_TO_LIVE_IN_MEMORY_INTERNAL): Remove.
Update all users.
(is_global_var): Declare.
* tree-dfa.c (dump_variable): Display global and addressable
attributes.
(add_referenced_var): Clarify documentation when marking
variables call-clobbered.
* tree-flow-inline.h (is_call_clobbered): Call is_global_var
instead of needs_to_live_in_memory.
(mark_call_clobbered): If the variable is a tag, mark it
DECL_EXTERNAL.
* tree-gimple.c (is_gimple_reg): Don't check for
TREE_ADDRESSABLE.
(is_gimple_non_addressable): Likewise.
* tree-ssa-alias.c (get_nmt_for): Always check whether the tag
needs to be marked call-clobbered.
(setup_pointers_and_addressables): Call is_global_var instead
of needs_to_live_in_memory.
* tree-ssa-dce.c (need_to_preserve_store): Remove.
Update all users with is_global_var.
(mark_stmt_if_obviously_necessary): Fix processing of aliased
stores. Don't check the virtual definitions. Rather, check
whether the store is going into global memory.
(find_obviously_necessary_stmts): Get the symbol from the PHI
result.
* tree-ssa-operands.c (get_call_expr_operands): Do not add
clobbering may-defs if the call does not have side effects.
libjava/ChangeLog
PR tree-optimization/16867
* testsuite/libjava.lang/PR16867.java: New test.
From-SVN: r85874
