Richard Guenther [Fri, 19 Mar 2010 10:18:25 +0000 (10:18 +0000)]
re PR tree-optimization/43415 (Consumes large amounts of memory and time in PRE at -O3)
2010-03-19 Richard Guenther <rguenther@suse.de>
PR tree-optimization/43415
* tree-ssa-pre.c (phi_translate): Split out worker to ...
(phi_translate_1): ... this.
(phi_translate): Move all caching here. Cache all NARY
and REFERENCE translations.
David S. Miller [Fri, 19 Mar 2010 05:34:12 +0000 (05:34 +0000)]
sparc.c: Include dwarf2out.h.
* config/sparc/sparc.c: Include dwarf2out.h.
(emit_pic_helper): Delete.
(pic_helper_symbol_name): Delete.
(pic_helper_emitted_p): Delete.
(pic_helper_needed): New.
(USE_HIDDEN_LINKONCE): Define to '1' if HAVE_GAS_HIDDEN else '0'.
(get_pc_thunk_name): New.
(load_pic_register): Remove 'delay_pic_helper' arg. Use
get_thunk_pc_name and ggc_strdup to generate PIC thunk symbol.
Set pic_helper_needed to true. Don't call emit_pic_helper.
(sparc_expand_prologue): Update load_pic_register call.
(sparc_output_mi_thunk): Likewise.
(sparc_file_end): Emit a hidden comdat symbol for the PIC
thunk if possible. Output CFI information as needed.
2010-03-11 Francois-Xavier Coudert <fxcoudert@gcc.gnu.org>
Jack Howarth <howarth@bromo.med.uc.edu>
PR target/36399
* gcc/config/i386/i386.h: Fix ABI on darwin x86-32.
* gcc/testsuite/gcc.target/i386/push-1.c: Don't xfail
scan-assembler-not "movups" on darwin.
Co-Authored-By: Jack Howarth <howarth@bromo.med.uc.edu>
From-SVN: r157553
Jakub Jelinek [Thu, 18 Mar 2010 20:15:05 +0000 (21:15 +0100)]
re PR rtl-optimization/43058 (var-tracking uses up all virtual memory)
PR debug/43058
* var-tracking.c (non_suitable_const): New function.
(add_uses): For DEBUG_INSNs with constants, don't record any
value, instead just the constant value itself.
(compute_bb_dataflow) <case MO_VAL_LOC>: If PAT_VAR_LOCATION_LOC
is not VAR_LOC_UNKNOWN_P, set var to the constant.
(emit_notes_in_bb): Likewise.
(emit_note_insn_var_location): For onepart variables if
cur_loc is a VOIDmode constant, use DECL_MODE.
Martin Jambor [Thu, 18 Mar 2010 20:07:13 +0000 (21:07 +0100)]
re PR middle-end/42450 (another GCC 4.5 ICE on C++ templated code)
2010-03-18 Martin Jambor <mjambor@suse.cz>
PR middle-end/42450
* cgraph.h (cgraph_redirect_edge_call_stmt_to_callee): Declare.
* cgraphunit.c (cgraph_materialize_all_clones): Update calls in
all non-clones. Moved call redirection...
(cgraph_redirect_edge_call_stmt_to_callee): ...to this new
function.
(cgraph_materialize_all_clones): Dispose of all
combined_args_to_skip bitmaps.
(verify_cgraph_node): Do not check for edges pointing to wrong
nodes in inline clones.
* tree-inline.c (copy_bb): Call
cgraph_redirect_edge_call_stmt_to_callee.
* ipa.c (cgraph_remove_unreachable_nodes): Call
cgraph_node_remove_callees even when there are used clones.
Peter Bergner [Thu, 18 Mar 2010 03:10:04 +0000 (22:10 -0500)]
re PR target/42427 (invalid assembly code for 301.apsi for -fnon-call-exceptions)
gcc/
PR target/42427
* config/rs6000/rs6000.c (rs6000_split_multireg_move): Add support for
non-offsettable and pre_modify update addressing.
* config/rs6000/dfp.md (*movdd_hardfloat32): Make the "0", "1"
and "2" alternatives "#".
(*movdd_softfloat32): Make all alternatives "#";
* config/rs6000/rs6000.md (DIFD): New define_mode_iterator.
(*movdf_hardfloat32): Make the "0", "1" and "2" alternatives "#".
(*movdf_softfloat32): Make all alternatives "#";
(movdi): Use the new DIFD mode iterator to create a common splitter
for movdi, movdf and movdd patterns.
gcc/testsuite/
PR target/42427
* gcc.dg/pr42427.c: New test.
Jerry DeLisle [Thu, 18 Mar 2010 02:38:17 +0000 (02:38 +0000)]
re PR fortran/43265 (No EOF condition if reading with '(x)' from an empty file)
2010-03-17 Jerry DeLisle <jvdelisle@gcc.gnu.org>
PR libfortran/43265
* io/io.h: Delete prototype for read_sf, making it static.
* io/read.c (read_x): Modify to call hit_eof if PAD="no".
* io/transfer.c (read_sf_internal): New static function extracted from
read_sf for use on internal units only. Handle empty string case.
(read_sf): New factoring of this function, make it static. Add special
conditions for EOF based on ADVANCE="no", PAD="no", and whether any
bytes have been previously read from the record.
(read_block_form): Modify to call read_sf or read_sf_internal.
(next_record_r): Add a done flag similar to next_record_w. Call hit_eof
if internal array unit next record returns finished, meaning an EOF was
found and not done, ie not the last record expected. For external
units call hit_eof if item_count is 1 or there are no pending spaces.
(next_record): Update call to next_record_r.
Bernd Schmidt [Wed, 17 Mar 2010 09:25:35 +0000 (09:25 +0000)]
re PR rtl-optimization/42216 (changes in scheduling regress 464.h264ref 20%)
PR rtl-optimization/42216
* regrename.c (create_new_chain): New function, broken out from...
(scan_rtx_reg): ... here. Call it. Handle the case where we are
appending a use to an empty chain.
(build_def_use): Remove previous changes that convert OP_INOUT to
OP_OUT operands; instead detect the case where an OP_INOUT operand
uses a previously untracked register and create an empty chain for
it.
re PR middle-end/43365 (Destructor not called when returning in exception handler)
PR middle-end/43365
* tree-eh.c (replace_goto_queue): Also replace in the eh_seq.
(lower_try_finally): Save and restore eh_seq around the expansion
of the try-finally.
* doc/sourcebuild.texi (Testsuites): Fix markup.
Use pathnames relative to gcc/testsuite.
(Test Directives): Move description of how timeout is determined.
(Ada Tests): Favor gnat.exp over ada/acats/tests/gcc.
(C Tests): Correct gcc.misc-tests directory.
Framework tests now live in gcc.test-framework.
Jakub Jelinek [Tue, 16 Mar 2010 10:50:42 +0000 (11:50 +0100)]
re PR debug/43051 (VTA causes a stack living parameter unavailable in most of the function)
PR debug/43051
PR debug/43092
* cselib.c (cselib_preserve_constants,
cfa_base_preserved_val): New static variables.
(preserve_only_constants): New function.
(cselib_reset_table): If cfa_base_preserved_val is non-NULL, don't
clear its REG_VALUES. If cselib_preserve_constants, don't
empty the whole hash table, but preserve there VALUEs with constants,
cfa_base_preserved_val and cfa_base_preserved_val plus constant.
(cselib_preserve_cfa_base_value): New function.
(cselib_invalidate_regno): Don't invalidate cfa_base_preserved_val.
(cselib_init): Change argument to int bitfield. Set
cselib_preserve_constants to whether CSELIB_PRESERVE_CONSTANTS
is in it.
(cselib_finish): Clear cselib_preserve_constants and
cfa_base_preserved_val.
* cselib.h (enum cselib_record_what): New enum.
(cselib_init): Change argument to int.
(cselib_preserve_cfa_base_value): New prototype.
* postreload.c (reload_cse_regs_1): Adjust cselib_init caller.
* dse.c (dse_step1): Likewise.
* cfgcleanup.c (thread_jump): Likewise.
* sched-deps.c (sched_analyze): Likewise.
* gcse.c (local_cprop_pass): Likewise.
* simplify-rtx.c (simplify_replace_fn_rtx): Add argument to callback.
If FN is non-NULL, call the callback always and whenever it returns
non-NULL just return that. Only do rtx_equal_p if FN is NULL.
* rtl.h (simplify_replace_fn_rtx): Add argument to callback.
* combine.c (propagate_for_debug_subst): Add old_rtx argument,
compare from with old_rtx and if it isn't rtx_equal_p, return NULL.
* Makefile.in (var-tracking.o): Depend on $(RECOG_H).
* var-tracking.c: Include recog.h.
(bb_stack_adjust_offset): Remove.
(vt_stack_adjustments): Don't call it, instead just gather the
adjustments using insn_stack_adjust_offset_pre_post on each bb insn.
(adjust_stack_reference): Remove.
(compute_cfa_pointer): New function.
(hard_frame_pointer_adjustment, cfa_base_rtx): New static variables.
(struct adjust_mem_data): New type.
(adjust_mems, adjust_mem_uses, adjust_mem_stores, adjust_insn): New
functions.
(get_address_mode): New function.
(replace_expr_with_values): Use it.
(use_type): Don't do cselib_lookup for VAR_LOC_UNKNOWN_P.
Use get_address_mode. For cfa_base_rtx return MO_CLOBBER.
(adjust_sets): Remove.
(add_uses): Don't add extra MO_VAL_USE for cfa_base_rtx plus constant.
Use get_address_mode.
(get_adjusted_src): Remove.
(add_stores): Don't call it. Never reuse expr SET. Don't add extra
MO_VAL_USE for cfa_base_rtx plus constant. Use get_address_mode.
(add_with_sets): Don't call adjust_sets.
(fp_setter, vt_init_cfa_base): New functions.
(vt_initialize): Change return type to bool. Move most of pool etc.
initialization to the beginning of the function from end. Pass
CSELIB_RECORD_MEMORY | CSELIB_PRESERVE_CONSTANTS to cselib_init.
If !frame_pointer_needed, call vt_stack_adjustment before mos
vector is filled, call vt_init_cfa_base if argp/framep has been
eliminated to sp. If frame_pointer_needed and argp/framep has
been eliminated to hard frame pointer, set
hard_frame_pointer_adjustment and call vt_init_cfa_base after
encountering fp setter in the prologue. For MO_ADJUST, call
log_op_type before pusing the op into mos vector, not afterwards.
Call adjust_insn before cselib_process_insn/add_with_sets,
call cancel_changes (0) afterwards.
(variable_tracking_main_1): Adjust for vt_initialize calling
vt_stack_adjustments and returning whether it succeeded or not.
Jakub Jelinek [Mon, 15 Mar 2010 17:03:02 +0000 (18:03 +0100)]
dwarf2out.c (dwarf2out_frame_debug): Don't assert drap_reg has been set.
* dwarf2out.c (dwarf2out_frame_debug): Don't assert drap_reg
has been set.
(based_loc_descr): Use DW_OP_fbreg for vdrap_reg even when
drap_reg has not been set.
Michael Matz [Mon, 15 Mar 2010 16:13:28 +0000 (16:13 +0000)]
re PR middle-end/43300 (ICE: in emit_move_insn, at expr.c:3432)
PR middle-end/43300
* tree-outof-ssa.c (emit_partition_copy): New argument sizeexp,
use it to expand block copies.
(insert_partition_copy_on_edge, insert_rtx_to_part_on_edge,
insert_part_to_rtx_on_edge): Adjust callers of emit_partition_copy.
(insert_value_copy_on_edge): Use store_expr for BLKmode values.
PR middle-end/43354
* graphite-sese-to-poly.c (rewrite_close_phi_out_of_ssa): Do not
call insert_out_of_ssa_copy for default definitions.
* gfortran.dg/graphite/id-pr43354.f: New.
Sebastian Pop [Sat, 13 Mar 2010 17:35:29 +0000 (17:35 +0000)]
Use ssizetype when long_long_integer_type_node is NULL.
2010-03-11 Sebastian Pop <sebastian.pop@amd.com>
* graphite-clast-to-gimple.c (my_long_long): Defined.
(gcc_type_for_cloog_iv): Use it instead of long_long_integer_type_node.
* graphite-sese-to-poly.c (my_long_long): Defined.
(scop_ivs_can_be_represented): Use it.
PR middle-end/43306
* tree-chrec.c (evolution_function_right_is_integer_cst): CHREC_RIGHT
should be an INTEGER_CST. Also handle CASE_CONVERT.
* gcc.dg/graphite/pr43306.c: New.
Sebastian Pop [Sat, 13 Mar 2010 17:34:51 +0000 (17:34 +0000)]
Add PARAM_GRAPHITE_MAX_BBS_PER_FUNCTION.
2010-03-09 Sebastian Pop <sebastian.pop@amd.com>
* graphite.c (graphite_initialize): To bound the number of bbs per
function, use PARAM_GRAPHITE_MAX_BBS_PER_FUNCTION.
* params.def (PARAM_GRAPHITE_MAX_BBS_PER_FUNCTION): Declared.
* doc/invoke.texi: Document it.
Sebastian Pop [Sat, 13 Mar 2010 17:34:38 +0000 (17:34 +0000)]
Limit the number of parameters per SCoP.
2010-03-09 Sebastian Pop <sebastian.pop@amd.com>
* graphite-sese-to-poly.c (build_poly_scop): Limit scops following
the number of parameters in the scop. Use as an upper bound
PARAM_GRAPHITE_MAX_NB_SCOP_PARAMS.
* params.def (PARAM_GRAPHITE_MAX_NB_SCOP_PARAMS): Declared.
* doc/invoke.texi: Document it.
Paul Thomas [Fri, 12 Mar 2010 22:00:52 +0000 (22:00 +0000)]
re PR fortran/43291 ([OOP] Type mismatch in argument; passed CLASS(t1) to CLASS(t2))
2010-03-12 Paul Thomas <pault@gcc.gnu.org>
PR fortran/43291
PR fortran/43326
* resolve.c (resolve_compcall): Add new boolean dummy argument
'class_members'. Only resolve expression at end if false.
Remove redundant, static variable 'class_object'.
(check_class_members): Add extra argument to call of
resolve_compcall.
(resolve_typebound_function): Renamed resolve_class_compcall.
Do all the detection of class references here. Correct calls to
resolve_compcall for extra argument.
(resolve_typebound_subroutine): resolve_class_typebound_call
renamed. Otherwise same as resolve_typebound_function.
(gfc_resolve_expr): Call resolve_typebound_function.
(resolve_code): Call resolve_typebound_subroutine.
2010-03-12 Paul Thomas <pault@gcc.gnu.org>
PR fortran/43291
PR fortran/43326
* gfortran.dg/dynamic_dispatch_7.f03: New test.
projects / thirdparty / gcc.git / log
