- Jan 20, 2006
-
-
Volker Reichelt authored
* parser.c (cp_lexer_next_token_is_keyword): Simplify. * parser.c (clear_decl_specs): Remove prototype. * parser.c (cp_parser_expression_fn): Remove. From-SVN: r110039
-
Joseph Myers authored
PR libstdc++/25524 * cppdefault.h (struct default_include): Add multilib flag. * cppdefault.c (cpp_include_defaults): Set it. * c.opt (-imultilib): New option. * c-opts.c (imultilib): New. (c_common_handle_option): Handle -imultilib. (c_common_post_options): Likewise. * c-incpath.c (add_standard_paths, register_include_chains): Likewise. * c-incpath.h (register_include_chains): Add extra parameter. * gcc.c (do_spec_1): Generate -imultilib option. (The Specs Language): Update %I description. (process_command): Update copyright notice. * doc/cppopts.texi (-imultilib): Document. * doc/invoke.texi (-imultilib): Include in option summary. (%I): Update specs documentation. libstdc++-v3: * include/Makefile.am: Install host-specific headers in multilib subdirectory. * include/Makefile.in: Regenerate. From-SVN: r110037
-
Steve Ellcey authored
PR testsuite/24962 * gcc.target/ia64/20030811-1.c: Change 'long' to 'long long'. From-SVN: r110034
-
Steve Ellcey authored
PR libgomp/25877 * team.c: Add include of alloca.h. * configure.ac: Add check for alloca.h. * configure: Regenerate. * config.h.in: Regenerate. From-SVN: r110031
-
Zdenek Dvorak authored
* loop-iv.c (iv_analysis_loop_init): Use df in more conservative way. From-SVN: r110030
-
John David Anglin authored
PR ada/24533 * s-osinte-linux-hppa.ads: Reduce alignment of atomic_lock_t to 8. From-SVN: r110025
-
Diego Novillo authored
* configure: Regenerate. From-SVN: r110023
-
Volker Reichelt authored
* call.c (add_builtin_candidates): Remove superfluous return. * name-lookup.c (do_toplevel_using_decl): Likewise. * parser.c (cp_parser_type_specifier_seq): Likewise. (cp_parser_save_default_args): Likewise. From-SVN: r110022
-
Richard Sandiford authored
* stmt.c (case_bit_test_cmp): Stabilize the sort using code label numbers. From-SVN: r110021
-
Richard Guenther authored
2006-01-20 Richard Guenther <rguenther@suse.de> Steven Bosscher <stevenb.gcc@gmail.com> PR rtl-optimization/24626 * cfgloopmanip.c (lv_adjust_loop_entry_edge): Don't set EDGE_TRUE_VALUE if in RTL mode. Revert 2005-03-30 Mostafa Hagog <mustafa@il.ibm.com> * cfgrtl.c (rtl_verify_flow_info_1): Fix. * gcc.dg/torture/pr24626-1.c: New testcase. * gcc.dg/torture/pr24626-2.c: Likewise. * gcc.dg/torture/pr24626-3.c: Likewise. * gcc.dg/torture/pr24626-4.c: Likewise. Co-Authored-By:
Steven Bosscher <stevenb.gcc@gmail.com> From-SVN: r110020
-
Dirk Mueller authored
2006-01-20 Dirk Mueller <dmueller@suse.com> PR c++/5520 * c-parser.c (c_parser_if_body): Use build_empty_stmt() instead of a special NOP marker. * c-typeck.c (c_finish_if_stmt): Remove obsoleted special NOP marker handling. * c-common.h (empty_body_warning): Add forward declaration. * c-common.c (empty_body_warning): Add (from c_finish_if_stmt). Now uses IS_EMPTY_STMT() instead of special NOP markers. * semantics.c (finish_if_stmt): Call empty_body_warning. * parser.c (cp_parser_implicitly_scoped_statement): Mark empty statement with an empty stmt. * g++.dg/warn/empty-body.C: New. From-SVN: r110019
-
Mark Mitchell authored
PR c++/22136 * name-lookup.c (do_class_using_decl): Don't try to look up base classes in templates with dependent base types. PR c++/22136 * g++.dg/template/using10.C: New test. * g++.dg/temlpate/using11.C: Likewise. * g++.dg/inherit/using5.C: Tweak error messages. From-SVN: r110017
-
Paul Brook authored
2005-01-20 Paul Brook <paul@codesourcery.com> * Makefile.in (lang_checks): Keep going if mkdir fails. From-SVN: r110012
-
Paolo Carlini authored
2006-01-19 Paolo Carlini <pcarlini@suse.de> Implement list::splice (and merge) bits of N1599 * include/bits/stl_list.h (list<>::_M_check_equal_allocators): New. (list<>::splice(iterator, list&), splice(iterator, list&, iterator), splice(iterator, list&, iterator, iterator)): Use it. * include/bits/stl_list.h (list<>::merge(list&), merge(list&, _StrictWeakOrdering)): Likewise. * testsuite/23_containers/list/operators/5.cc: New. From-SVN: r110011
-
Kenneth Zadeck authored
df-scan.c (df_scan_free_bb_info): Added basic block parameter to be able to clean out basic block when not necessary. 2005-01-19 Kenneth Zadeck <zadeck@naturalbridge.com> * df-scan.c (df_scan_free_bb_info): Added basic block parameter to be able to clean out basic block when not necessary. (df_scan_free): Fixed to properly delete information if df is unused before calling df_finish. (df_scan_alloc, df_rescan_blocks, df_reg_chain_unlink, df_insn_create_insn_record, df_bb_refs_record): Fixed formatting or comment typos. (df_bb_refs_delete): New function. (df_refs_delete): Moved per block code to df_bb_refs_delete. * df-core.c (df_set_blocks): Added code to properly clean out unused blocks if they are not part of the blocks to consider. Added additional null check. (df_compact_blocks): Added basic block parameter to free_bb_fun to be able to clean out basic block when not necessary * df.h (df_free_bb_function): Ditto. (df_bb_refs_delete): New function. * df-problems.c (df_ru_free_bb_info, df_rd_set_bb_info, df_lr_set_bb_info, df_ur_free_bb_info, df_urec_free_bb_info):Added basic block parameter to be able to clean out basic block when not necessary. (df_ru_alloc, df_rd_alloc): Fixed dyslexic overflow test. (df_ru_free, df_rd_free, df_lr_free, df_ur_free, df_urec_free): Fixed to properly delete information if df is unused before calling df_finish. From-SVN: r110009
-
Kenneth Zadeck authored
2005-01-19 Kenneth Zadeck <zadeck@naturalbridge.com> PR rtl-optimization/25799 * df-problems.c (df_ru_confluence_n, df_rd_confluence_n): Corrected confluence operator to remove bits from op2 before oring with op1 rather than removing bits from op1. (df_ru_transfer_function): Corrected test on wrong bitmap which caused infinite loop. Both of these problems were introduced in the dataflow rewrite. From-SVN: r110008
-
Kenneth Zadeck authored
2005-01-19 Kenneth Zadeck <zadeck@naturalbridge.com> PR rtl-optimization/25799 * df-problems.c (df_ru_confluence_n, df_rd_confluence_n): Corrected confluence operator to remove bits from op2 before oring with op1 rather than removing bits from op1. * (df_ru_transfer_function): Corrected test on wrong bitmap which caused infinite loop. Both of these problems were introduced in the dataflow rewrite. From-SVN: r110007
-
DJ Delorie authored
* reload1.c (find_reload_regs): Note the details of reload failures in the dump file. (spill_failure): Likewise. From-SVN: r110006
-
Zdenek Dvorak authored
* loop-iv.c: Include df.h and hashtab.h. (enum iv_grd_result): New enum. (DF_REF_IV, DF_REF_IV_SET): New macros. (struct biv_entry): New. (df, bivs): New global variables. (struct insn_info, insn_info, last_def, bivs, max_insn_no, max_reg_no, assign_luids, mark_sets, kill_sets, mark_single_set, simple_set_p): Removed. (clear_iv_info, latch_dominating_def, record_iv, iv_analyze_expr, iv_analyze_result, iv_analyze_def, biv_hash, biv_eq, analyzed_for_bivness_p, record_biv): New functions. (iv_analysis_loop_init, iv_get_reaching_def, simple_reg_p, get_biv_step_1, get_biv_step, iv_analyze_biv, iv_analyze_op, iv_analyze, biv_p, iv_analysis_done): Work with df representation of UD chains. (iv_constant, iv_subreg, iv_extend, iv_mult, iv_shift): Do not set analysed. (iv_number_of_iterations): Use new interface to iv analysis. * loop-unroll.c: Do not include varray.h. (analyze_iv_to_split_insn): Use new interface to iv analysis. * loop-unswitch.c (may_unswitch_on): Ditto. * df.c (df_bitmaps_free): Only work for bbs for that structures are allocated. (df_bb_modify): Realloc tables to the new index. (df_find_use): New function. * df.h (df_find_use): Declare. * optabs.c (expand_unop): Make the mode of the REG_EQUAL node be outmode. * cfgloop.h (struct rtx_iv): Remove analysed field. (iv_get_reaching_def): Removed. (iv_analyze_result, iv_analyze_expr, iv_current_loop_df): Declare. * Makefile.in (loop-unroll.o): Remove VARRAY_H dependency. (loop-iv.o): Add df.h and hashtab.h dependency. * df-problems.c (df_ru_alloc, df_rd_alloc): Fix memory reallocation. From-SVN: r110005
-
Alan Modra authored
* libgcc2.c (__floatdisf, __floatdidf): Don't use IBM Extended Double TFmode. (__floatundisf, __floatundidf): Likewise. * libgcc2.h (IS_IBM_EXTENDED): Define. From-SVN: r110004
-
DJ Delorie authored
From-SVN: r110003
-
GCC Administrator authored
From-SVN: r110001
-
Alan Modra authored
* gcc.target/powerpc/rs6000-ldouble-1.c: Tweak powerpc linux target test. * gcc.target/powerpc/rs6000-ldouble-2.c: Likewise. From-SVN: r109995
-
Paul Brook authored
2006-01-19 Paul Brook <paul@codesourcery.com> * gcc/config/arm/arm.c (arm_compute_func_type): Treat all functions as nothrow when unwinding tables are disabled. From-SVN: r109994
-
- Jan 19, 2006
-
-
John David Anglin authored
PR testsuite/25171 * c-decl.c (check_bitfield_type_and_width): Don't issue pedwarn "type of bit-field" when in system header. From-SVN: r109990
-
DJ Delorie authored
* config/m32c/m32c.c (m32c_prepare_shift): Add code to deal with the 16 bit shift limit of the m16c. From-SVN: r109987
-
H.J. Lu authored
2006-01-19 H.J. Lu <hongjiu.lu@intel.com> PR libstdc++/25797 * acinclude.m4 (GLIBCXX_CHECK_LINKER_FEATURES): Remove -Werror from CFLAGS. Check if linker really supports --gc-sections. * configure: Regenerated. From-SVN: r109985
-
Jan Hubicka authored
From-SVN: r109982
-
Volker Reichelt authored
PR c++/25854 * pt.c (maybe_process_partial_specialization): Return early on error_mark_node. * g++.dg/template/spec28.C: New test. From-SVN: r109975
-
Andrew Pinski authored
2006-01-19 Andrew Pinski <pinskia@physics.uc.edu> PR target/22099 * config/i386/i386.md (builtin_setjmp_receiver): Don't emit the label. * config/i386/i386.c (output_set_got): Output the label if we have one for the TARGET_DEEP_BRANCH_PREDICTION case. From-SVN: r109974
-
Andrew Pinski authored
2006-01-19 Andrew Pinski <pinskia@physics.uc.edu> PR target/15642 * inclhack.def (AAB_darwin7_9_long_double_funcs [replace]): Define __APPLE_CC_ as 1345. (broken_nan): New. * fixincl.x: Regenerate. From-SVN: r109973
-
Jan Hubicka authored
* invoke.texi (generic): Document (i686) Update. * config.gcc: Make x86_64-* and i686-* default to generic tunning. * i386.h (TARGET_GENERIC32, TARGET_GENERIC64, TARGET_GENERIC, TARGET_USE_INCDEC, TARGET_PAD_RETURNS): New macros. (x86_use_incdec, x86_pad_returns): New variables (TARGET_CPU_DEFAULT_generic): New constant (TARGET_CPU_DEFAULT_NAMES): Add generic. (enum processor_type): Add generic32 and generic64. * i386.md (cpu attribute): Add generic32/generic64 (movhi splitter): Behave sanely when both partial_reg_dependency and partial_reg_stall are set. (K8 splitters): Enable for generic as well. * predicates.md (incdec_operand): Use TARGET_INCDEC (aligned_operand): Avoid memory mismatch stalls. * athlon.md: Enable for generic64, new patterns for 128bit moves. * ppro.md: Enable for generic32 * i386.c (generic64_cost, generic32_cost): New. (m_GENERIC32, m_GENERIC64, m_GENERIC): New macros. (x86_use_leave): Enable for generic64. (x86_use_sahf, x86_ext_80387_constants): Enable for generic32. (x86_push_memory, x86_movx, x86_unroll_strlen, x86_deep_branch, x86_use_simode_fiop, x86_use_cltd, x86_promote_QImode, x86_sub_esp_4, x86_sub_esp_8, x86_add_esp_4, x86_add_esp_8, x86_integer_DFmode_moves, x86_partial_reg_dependency, x86_memory_mismatch_stall, x86_accumulate_outgoing_args, x86_prologue_using_move, x86_epilogue_using_move, x86_arch_always_fancy_math_387, x86_sse_partial_reg_dependency, x86_four_jump_limit, x86_schedule): Enable for generic. (x86_use_incdec, x86_pad_returns): New. (override_options): Add generic32 and generic64, translate "generic" to generic32/generic64 and "i686" to "generic32", refuse "generic32"/"generic64" as arch target. (ix86_issue_rate, ix86_adjust_cost): Handle generic as athlon. (ix86_reorg): Honor PAD_RETURNS. From-SVN: r109972
-
Jan Hubicka authored
* invoke.texi (generic): Document (i686) Update. * config.gcc: Make x86_64-* and i686-* default to generic tunning. * i386.h (TARGET_GENERIC32, TARGET_GENERIC64, TARGET_GENERIC, TARGET_USE_INCDEC, TARGET_PAD_RETURNS): New macros. (x86_use_incdec, x86_pad_returns): New variables (TARGET_CPU_DEFAULT_generic): New constant (TARGET_CPU_DEFAULT_NAMES): Add generic. (enum processor_type): Add generic32 and generic64. * i386.md (cpu attribute): Add generic32/generic64 (movhi splitter): Behave sanely when both partial_reg_dependency and partial_reg_stall are set. (K8 splitters): Enable for generic as well. * predicates.md (incdec_operand): Use TARGET_INCDEC (aligned_operand): Avoid memory mismatch stalls. * athlon.md: Enable for generic64, new patterns for 128bit moves. * ppro.md: Enable for generic32 * i386.c (generic64_cost, generic32_cost): New. (m_GENERIC32, m_GENERIC64, m_GENERIC): New macros. (x86_use_leave): Enable for generic64. (x86_use_sahf, x86_ext_80387_constants): Enable for generic32. (x86_push_memory, x86_movx, x86_unroll_strlen, x86_deep_branch, x86_use_simode_fiop, x86_use_cltd, x86_promote_QImode, x86_sub_esp_4, x86_sub_esp_8, x86_add_esp_4, x86_add_esp_8, x86_integer_DFmode_moves, x86_partial_reg_dependency, x86_memory_mismatch_stall, x86_accumulate_outgoing_args, x86_prologue_using_move, x86_epilogue_using_move, x86_arch_always_fancy_math_387, x86_sse_partial_reg_dependency, x86_four_jump_limit, x86_schedule): Enable for generic. (x86_use_incdec, x86_pad_returns): New. (override_options): Add generic32 and generic64, translate "generic" to generic32/generic64 and "i686" to "generic32", refuse "generic32"/"generic64" as arch target. (ix86_issue_rate, ix86_adjust_cost): Handle generic as athlon. (ix86_reorg): Honor PAD_RETURNS. Co-Authored-By:
Evandro Menezes <evandro.menezes@amd.com> Co-Authored-By:
H.J. Lu <hongjiu.lu@intel.com> From-SVN: r109971
-
Andrew Pinski authored
2006-01-19 Andrew Pinski <pinskia@physics.uc.edu> * objc.dg/gnu-encoding/struct-layout-encoding-1_generate.c (switchfiles): Fix fprintf's by adding srcdir or removing it. From-SVN: r109970
-
Diego Novillo authored
tree-pretty-print.c (dump_generic_node): Handle OMP_PARALLEL_FN, OMP_PARALLEL_DATA_ARG and OMP_RETURN_EXPR. * tree-pretty-print.c (dump_generic_node): Handle OMP_PARALLEL_FN, OMP_PARALLEL_DATA_ARG and OMP_RETURN_EXPR. * cgraph.c (cgraph_expand_queue): Rename from cgraph_analyze_queue. Update all users. * cgraphunit.c (cgraph_assemble_pending_functions): Process cgraph_expand_queue. (cgraph_expand_all_functions): Likewise. (cgraph_finalize_pending_functions): Remove. Update callers. * tree.h (OMP_DIRECTIVE_P): Define. (OMP_PARALLEL_FN): Define. (OMP_PARALLEL_DATA_ARG): Define. (OMP_SECTIONS_SECTIONS): Define. * tree-pass.h (pass_expand_omp): Declare. * omp-low.c (struct omp_region): Declare. (struct omp_context): Remove fields 'parallel_type', 'parallel_start_ix' and 'parallel_start_additional_args'. Update all users. (struct omp_for_data): Rename from struct expand_omp_for_data. (omp_regions): New static variable. (root_omp_region): New static variable. (find_omp_clause): Make static. (is_in_combined_parallel_ctx): Remove. (is_combined_parallel): New. (extract_omp_for_data): Move earlier in the file. (workshare_safe_to_combine_p): New. (get_ws_args_for): New. (determine_parallel_type): Move earlier in the file. (omp_copy_decl_2): Do not set DECL_CONTEXT of new local to the child function. (omp_copy_decl): Likewise. (create_omp_child_function): Likewise. (lookup_omp_region): New. (dump_omp_region): New. (debug_omp_region): New. (debug_all_omp_regions): New. (new_omp_region): New. (scan_omp_parallel): If parallel_nesting_level > 1, the directive is nested within another parallel directive. Set OMP_PARALLEL_FN. (scan_omp_for): Do not try to handle combined parallel+for cases. Remove FIXME comment. (scan_omp_nested): Remove. (scan_omp_1): Do not call scan_omp_nested when parallel_nesting_level is > 1. Do not change the DECL_CONTEXT of local variables found. (lookup_decl_in_outer_ctx): New. (lower_rec_input_clauses): Rename from expand_rec_input_clauses. (lower_lastprivate_clauses): Rename from expand_lastprivate_clauses. (lower_reduction_clauses): Rename from expand_reduction_clauses. (lower_copyprivate_clauses): Rename from expand_copyprivate_clauses. If CTX is nested, lookup VAR in the outer context when building copy assignment. (lower_send_clauses): Rename from expand_send_clauses. If CTX is nested, lookup VAR in the outer context when building copy assignments. (lower_send_shared_vars): Rename from expand_send_shared_vars. If CTX is nested, lookup VAR in the outer context when building copy assignments. (expand_parallel_call): Rename from build_parallel_call. Handle combined parallel+workshare cases. Re-implement to emit code into the CFG. (list2chain): New. (expand_omp_parallel): Re-implement to emit code into the CFG. Call move_sese_region_to_fn to outline the sub-graph containing the parallel region. (expand_omp_for_1): Remove. (expand_omp_for_generic): Re-implement to emit code into the CFG. (expand_omp_for_static_nochunk): Likewise. (expand_omp_for_static_chunk): Likewise. (expand_omp_for): Likewise. (expand_omp_sections): Likewise. (remove_exit_barriers): New. (expand_omp_synch): New. (expand_omp): New. (build_omp_regions_1): New. (build_omp_regions): New. (execute_expand_omp): New. (gate_expand_omp): New. (pass_expand_omp): Define. (lower_omp_sections): Rename from expand_omp_sections. Set OMP_SECTIONS_SECTIONS. (lower_omp_single_simple): Rename from expand_omp_single_simple. (lower_omp_single_copy): Rename from expand_omp_single_copy. (lower_omp_single): Rename from expand_omp_simple. (lower_omp_master): Rename from expand_omp_master. (lower_omp_ordered): Rename from expand_omp_ordered. (lower_omp_critical): Rename from expand_omp_critical. (lower_omp_for_lastprivate): Rename from expand_omp_for_lastprivate. (lower_omp_for): Re-implement. (lower_omp_parallel): Re-implement. (lower_regimplify): Rename from expand_regimplify. (lower_omp_1): Rename from expand_omp_1. If there are syntax errors in the program, replace every OpenMP directive with NOP. Call lower_omp_* instead of expand_omp_*. (lower_omp): Rename from expand_omp. * tree-gimple.c (is_gimple_stmt): Handle OMP_RETURN_EXPR. * tree-gimple.h (enum omp_parallel_type): Remove. (gimple_boolify): Declare extern. (find_omp_clause, determine_parallel_type): Remove. * gimple-low.c (lower_omp_directive): New. (lower_stmt): Call it. (record_vars_into): Move from ... (record_vars): ... here. Call record_vars_into with current_function_decl. * gimplify.c (struct gimplify_ctx): Remove fields combined_pre_p and combined_ctxp. Update users. (get_formal_tmp_var): Add documentation. (gimple_boolify): Make extern. (gimplify_expr_in_ctx): Remove. Update callers. (gimplify_omp_parallel): Do not assume that OMP_PARALLEL_BODY will always be a BIND_EXPR. (gimplify_expr): Handle OMP_RETURN_EXPR. * tree.def (BLOCK): Remove documentation about BLOCK_TYPE_TAGS. (OMP_PARALLEL): Add 3 operands. (OMP_SECTIONS): Add 1 operand. (OMP_RETURN_EXPR): Define. * tree-inline.c (estimate_num_insns_1): Handle OpenMP directives. (copy_tree_r): Restore TREE_CHAIN in OMP_CLAUSE_*. * tree-iterator.c (alloc_stmt_list): Assert that we are not creating a circular free list. (free_stmt_list): Assert that we are not freeing stmt_list_cache. * tree-flow.h (move_sese_region_to_fn): Declare. (record_vars_into): Declare. * tree-cfg.c (make_omp_sections_edges): New. (make_exit_edges): Handle OMP_PARALLEL, OMP_FOR, OMP_SINGLE, OMP_MASTER, OMP_ORDERED, OMP_CRITICAL, OMP_RETURN_EXPR, OMP_SECTIONS and OMP_SECTION. (is_ctrl_altering_stmt): Return true for OMP_DIRECTIVE_P. (set_bb_for_stmt): Undo change to check currently_expanding_to_rtl. (verify_stmt): Do not handle OMP_DIRECTIVE_P. (gather_blocks_in_sese_region): New. (struct move_stmt_d): Declare. (move_stmt_r): New. (move_block_to_fn): New. (move_sese_region_to_fn): New. * passes.c (init_optimization_passes): Schedule pass_expand_omp after pass_init_datastructures. * tree-ssa-operands.c (get_expr_operands): Handle OMP_PARALLEL, OMP_SECTIONS, OMP_FOR, OMP_RETURN_EXPR, OMP_SINGLE, OMP_MASTER, OMP_ORDERED, OMP_CRITICAL. testsuite/ * testsuite/gcc.dg/gomp/for-13.c: Use -fdump-tree-ompexp. * testsuite/gcc.dg/gomp/critical-1.c: Likewise. * testsuite/gcc.dg/gomp/critical-3.c: Likewise. * testsuite/gcc.dg/gomp/empty.c: Likewise. * testsuite/gcc.dg/gomp/ordered-1.c: Likewise. * testsuite/gcc.dg/gomp/for-4.c: Likewise. * testsuite/gcc.dg/gomp/for-6.c: Likewise. * testsuite/gcc.dg/gomp/master-3.c: Likewise. * testsuite/gcc.dg/gomp/for-8.c: Likewise. * testsuite/gcc.dg/gomp/for-10.c: Likewise. * testsuite/gcc.dg/gomp/for-18.c: Likewise. * testsuite/gcc.dg/gomp/for-5.c: Likewise. * testsuite/gcc.dg/gomp/for-7.c: Likewise. * testsuite/gcc.dg/gomp/for-9.c: Likewise. From-SVN: r109969
-
Jeff Law authored
tree-vrp.c (extract_range_from_assert): Refine the result range if the variable referenced by the ASSERT_EXPR has a... * tree-vrp.c (extract_range_from_assert): Refine the result range if the variable referenced by the ASSERT_EXPR has a range and either the tentative result range or the previous range is a VR_ANTI_RANGE. * gcc.dg/tree-ssa/vrp25.c: New test. From-SVN: r109968
-
Richard Sandiford authored
From-SVN: r109967
-
Richard Sandiford authored
From-SVN: r109966
-
Jan Hubicka authored
From-SVN: r109962
-
Paolo Bonzini authored
2006-01-19 Paolo Bonzini <bonzini@gnu.org> * combine.c (try_combine): Do not worry about MEMs wrapped by USEs. (expand_compound_operation, expand_field_assignment): Fail if the bit lengths of an extract operation are out of range. (make_extraction): Compute wanted_inner_mode based on the position and length of the extraction. Make it extraction_mode for non-constant positions, and do not modify offset in that case. When generating a new MEM, use a mode that can hold the extraction while keeping correct alignment. Remove code that supported MEMs wrapped by USEs. (simplify_shift_const_1, force_to_mode) <case USE>: Remove. From-SVN: r109961
-