* var-tracking.c (track_expr_p): Do not track declarations for parts
of tracked parameters.
(add_stores): Do not track values for tracked parameters passed in
multiple locations.
(vt_get_decl_and_offset): Handle PARALLEL.
(vt_add_function_parameter): Handle parameters with incoming PARALLEL.
From-SVN: r205461
* tree-ssa-threadupdate.c (thread_through_all_blocks): Do not
clobber the loop structure thread_block was unsuccessful. If
thread_block was unsuccessful, cleanup appropriately.
From-SVN: r205459
PR sanitizer/59306
* ubsan.c (instrument_null): Use gimple_store_p/gimple_assign_load_p
instead of walk_gimple_op.
(ubsan_pass): Adjust. Call instrument_null only if SANITIZE_NULL.
testsuite/
* g++.dg/ubsan/pr59306.C: New test.
From-SVN: r205443
* cgraph.h (enum cgraph_simd_clone_arg_type): New.
(struct cgraph_simd_clone_arg, struct cgraph_simd_clone): New.
(struct cgraph_node): Add simdclone and simd_clones fields.
* config/i386/i386.c (ix86_simd_clone_compute_vecsize_and_simdlen,
ix86_simd_clone_adjust, ix86_simd_clone_usable): New functions.
(TARGET_SIMD_CLONE_COMPUTE_VECSIZE_AND_SIMDLEN,
TARGET_SIMD_CLONE_ADJUST, TARGET_SIMD_CLONE_USABLE): Define.
* doc/tm.texi.in (TARGET_SIMD_CLONE_COMPUTE_VECSIZE_AND_SIMDLEN,
TARGET_SIMD_CLONE_ADJUST, TARGET_SIMD_CLONE_USABLE): Add.
* doc/tm.texi: Regenerated.
* ggc.h (ggc_alloc_cleared_simd_clone_stat): New function.
* ipa-cp.c (determine_versionability): Fail if "omp declare simd"
attribute is present.
* omp-low.c: Include pretty-print.h, ipa-prop.h and tree-eh.h.
(simd_clone_vector_of_formal_parm_types): New function.
(simd_clone_struct_alloc, simd_clone_struct_copy,
simd_clone_vector_of_formal_parm_types, simd_clone_clauses_extract,
simd_clone_compute_base_data_type, simd_clone_mangle,
simd_clone_create, simd_clone_adjust_return_type,
create_tmp_simd_array, simd_clone_adjust_argument_types,
simd_clone_init_simd_arrays): New functions.
(struct modify_stmt_info): New type.
(ipa_simd_modify_stmt_ops, ipa_simd_modify_function_body,
simd_clone_adjust, expand_simd_clones, ipa_omp_simd_clone): New
functions.
(pass_data_omp_simd_clone): New variable.
(pass_omp_simd_clone): New class.
(make_pass_omp_simd_clone): New function.
* passes.def (pass_omp_simd_clone): New.
* target.def (TARGET_SIMD_CLONE_COMPUTE_VECSIZE_AND_SIMDLEN,
TARGET_SIMD_CLONE_ADJUST, TARGET_SIMD_CLONE_USABLE): New target
hooks.
* target.h (struct cgraph_node, struct cgraph_simd_node): Declare.
* tree-core.h (OMP_CLAUSE_LINEAR_VARIABLE_STRIDE): Document.
* tree.h (OMP_CLAUSE_LINEAR_VARIABLE_STRIDE): Define.
* tree-pass.h (make_pass_omp_simd_clone): New prototype.
* tree-vect-data-refs.c: Include cgraph.h.
(vect_analyze_data_refs): Inline by hand find_data_references_in_loop
and find_data_references_in_bb, if find_data_references_in_stmt
fails, still allow calls to #pragma omp declare simd functions
in #pragma omp simd loops unless they contain data references among
the call arguments or in lhs.
* tree-vect-loop.c (vect_determine_vectorization_factor): Handle
calls with no lhs.
(vect_transform_loop): Allow NULL STMT_VINFO_VECTYPE for calls without
lhs.
* tree-vectorizer.h (enum stmt_vec_info_type): Add
call_simd_clone_vec_info_type.
(struct _stmt_vec_info): Add simd_clone_fndecl field.
(STMT_VINFO_SIMD_CLONE_FNDECL): Define.
* tree-vect-stmts.c: Include tree-ssa-loop.h,
tree-scalar-evolution.h and cgraph.h.
(vectorizable_call): Handle calls without lhs. Assert
!stmt_can_throw_internal instead of failing for it. Don't update
EH stuff.
(struct simd_call_arg_info): New.
(vectorizable_simd_clone_call): New function.
(vect_transform_stmt): Call it.
(vect_analyze_stmt): Likewise. Allow NULL STMT_VINFO_VECTYPE for
calls without lhs.
* ipa-prop.c (ipa_add_new_function): Only call ipa_analyze_node
if cgraph_function_with_gimple_body_p is true.
c/
* c-decl.c (c_builtin_function_ext_scope): Avoid binding if
external_scope is NULL.
cp/
* semantics.c (finish_omp_clauses): For #pragma omp declare simd
linear clause step call maybe_constant_value.
testsuite/
* g++.dg/gomp/declare-simd-1.C (f38): Make sure
simdlen is a power of two.
* gcc.dg/gomp/simd-clones-2.c: Compile on all targets.
Remove -msse2. Adjust regexps for name mangling changes.
* gcc.dg/gomp/simd-clones-3.c: Likewise.
* gcc.dg/vect/vect-simd-clone-1.c: New test.
* gcc.dg/vect/vect-simd-clone-2.c: New test.
* gcc.dg/vect/vect-simd-clone-3.c: New test.
* gcc.dg/vect/vect-simd-clone-4.c: New test.
* gcc.dg/vect/vect-simd-clone-5.c: New test.
* gcc.dg/vect/vect-simd-clone-6.c: New test.
* gcc.dg/vect/vect-simd-clone-7.c: New test.
* gcc.dg/vect/vect-simd-clone-8.c: New test.
* gcc.dg/vect/vect-simd-clone-9.c: New test.
* gcc.dg/vect/vect-simd-clone-10.c: New test.
* gcc.dg/vect/vect-simd-clone-10.h: New file.
* gcc.dg/vect/vect-simd-clone-10a.c: New file.
* gcc.dg/vect/vect-simd-clone-11.c: New test.
Co-Authored-By: Jakub Jelinek <jakub@redhat.com>
From-SVN: r205442
2013-11-27 Richard Biener <rguenther@suse.de>
PR tree-optimization/59288
* tree-vect-loop.c (get_initial_def_for_induction): Do not
re-analyze the PHI but use STMT_VINFO_LOOP_PHI_EVOLUTION_PART.
* gcc.dg/torture/pr59288.c: New testcase.
From-SVN: r205434
2013-11-27 Marek Polacek <polacek@redhat.com>
* ubsan.c (ubsan_type_descriptor): If varpool_get_node returns NULL
for a decl, recreate that decl. Save into the hash table VAR_DECLs
rather than ADDR_EXPRs.
testsuite/
* c-c++-common/ubsan/undefined-1.c: New test.
From-SVN: r205433
* gengtype.c (struct seen_tag): New.
(already_seen_tag): New.
(mark_tag_as_seen): New.
(walk_subclasses): Support having multiple subclasses using the
same tag by tracking which tags have already been seen, and using
this to avoid adding duplicate cases to the "switch" statement.
The call to already_seen_tag introduces an O(N^2) when running
gengtype on N, the number of tags, due to the repeated linear
search, but currently max(N) is relatively small (the number of
GSS codes, which is 26).
(walk_type): Pass in a seen_tag for use by the walk_subclasses
recursion.
* gimple.def (GIMPLE_OMP_ATOMIC_STORE, GIMPLE_OMP_RETURN): Rename
underlying GSS values for these codes (from GSS_OMP_ATOMIC_STORE to
GSS_OMP_ATOMIC_STORE_LAYOUT) to make clear that although
GIMPLE_OMP_RETURN happens to share the data layout of
GIMPLE_OMP_ATOMIC_STORE, they are not otherwise related.
(GIMPLE_OMP_PARALLEL, GIMPLE_OMP_TARGET): Likewise, rename
underlying GSS value from GSS_OMP_PARALLEL to
GSS_OMP_PARALLEL_LAYOUT to make clear that these gimple codes are
not directly related; they merely share in-memory layout.
(GIMPLE_OMP_SINGLE, GIMPLE_OMP_TEAMS): Likewise, rename GSS values
for these two codes from GSS_OMP_SINGLE to GSS_OMP_SINGLE_LAYOUT.
* gsstruct.def (GSS_OMP_PARALLEL, gimple_statement_omp_parallel):
Rename to...
(GSS_OMP_PARALLEL_LAYOUT, gimple_statement_omp_parallel_layout):
...these.
(GSS_OMP_SINGLE, gimple_statement_omp_single): Rename to...
(GSS_OMP_SINGLE_LAYOUT, gimple_statement_omp_single_layout):
...these.
(GSS_OMP_ATOMIC_STORE, gimple_statement_omp_atomic_store): Rename
to...
(GSS_OMP_ATOMIC_STORE_LAYOUT, gimple_statement_omp_atomic_store):
...these.
* gimple.h (gimple_statement_resx): New subclass of
gimple_statement_eh_ctrl, with the invariant that
stmt->code == GIMPLE_RESX.
(gimple_statement_eh_dispatch): New subclass of
gimple_statement_eh_ctrl, with the invariant that
stmt->code == GIMPLE_EH_DISPATH.
(gimple_statement_omp_parallel): The existing class expressed
a layout (GSS_OMP_PARALLEL), but the codes with that layout
are not all related, so it makes more sense for this class to
express a *code* (GIMPLE_OMP_PARALLEL). GSS_OMP_PARALLEL has
been renamed to GSS_OMP_PARALLEL_LAYOUT to express this, so
rename the existing gimple_statement_omp_parallel class to...
(gimple_statement_omp_parallel_layout): ...this, expressing
a statement of structure layout GSS_OMP_PARALLEL_LAYOUT.
(gimple_statement_omp_taskreg): New subclass of
gimple_statement_omp_parallel_layout, expressing the invariant
that the code is one of GIMPLE_OMP_PARALLEL or GIMPLE_OMP_TASK,
as used by the various gimple_omp_taskreg_ accessors.
(gimple_statement_omp_parallel): Reintroduce this class, this time
as a subclass of gimple_statement_omp_taskreg to express the
invariant stmt->code == GIMPLE_OMP_PARALLEL.
(gimple_statement_omp_target) New class, subclassing
gimple_statement_omp_parallel_layout, to express the invariant
stmt->code == GIMPLE_OMP_TARGET.
(gimple_statement_omp_task): Update to inherit from
gimple_statement_omp_taskreg rather than
gimple_statement_omp_parallel.
(gimple_statement_omp_single): Rename to...
(gimple_statement_omp_single_layout): ...this, expressing the
invariant that the layout is GSS_OMP_SINGLE_LAYOUT.
(gimple_statement_omp_single): ...and reintroduce this name as
a subclass of gimple_statement_omp_single_layout, expressing
the invariant that code == GIMPLE_OMP_SINGLE.
(gimple_statement_omp_teams): New class, subclassing
gimple_statement_omp_single_layout, for the code GIMPLE_OMP_TEAMS.
(gimple_statement_omp_atomic_store): Rename to...
(gimple_statement_omp_atomic_store_layout): ...this, expressing
the invariant that the layout is GSS_OMP_ATOMIC_STORE_LAYOUT.
(gimple_statement_omp_atomic_store): ...and reintroduce this
name as a subclass of gimple_statement_omp_atomic_store_layout
with code == GIMPLE_OMP_ATOMIC_STORE.
(gimple_statement_omp_return): New class, subclassing
gimple_statement_omp_atomic_store_layout for the code
GIMPLE_OMP_RETURN.
(is_a_helper <gimple_statement_eh_ctrl>::test): Delete.
(is_a_helper <gimple_statement_resx>::test): New.
(is_a_helper <gimple_statement_eh_dispatch>::test): New.
(is_a_helper <gimple_statement_omp_atomic_store>::test): Only
check for GIMPLE_OMP_ATOMIC_STORE, not for GIMPLE_OMP_RETURN.
(is_a_helper <gimple_statement_omp_return>::test): New.
(is_a_helper <gimple_statement_omp_taskreg>::test): New.
(is_a_helper <gimple_statement_omp_parallel>::test): Only check
for GIMPLE_OMP_PARALLEL, not for GIMPLE_OMP_TASK or
GIMPLE_OMP_TARGET.
(is_a_helper <gimple_statement_omp_target>::test): New.
(is_a_helper <gimple_statement_omp_single>::test): Only check
for GIMPLE_OMP_SINGLE, not for GIMPLE_OMP_TEAMS.
(is_a_helper <gimple_statement_omp_teams>::test): New.
(is_a_helper <const gimple_statement_eh_ctrl>::test): Delete.
(is_a_helper <const gimple_statement_resx>::test): New.
(is_a_helper <const gimple_statement_eh_dispatch>::test): New.
(is_a_helper <const gimple_statement_omp_atomic_store>::test): Only
check for GIMPLE_OMP_ATOMIC_STORE, not for GIMPLE_OMP_RETURN.
(is_a_helper <const gimple_statement_omp_return>::test): New.
(is_a_helper <const gimple_statement_omp_taskreg>::test): New.
(is_a_helper <const gimple_statement_omp_parallel>::test): Only
check for GIMPLE_OMP_PARALLEL, not for GIMPLE_OMP_TASK or
GIMPLE_OMP_TARGET.
(is_a_helper <const gimple_statement_omp_target>::test): New.
(is_a_helper <const gimple_statement_omp_single>::test): Only
check for GIMPLE_OMP_SINGLE, not for GIMPLE_OMP_TEAMS.
(is_a_helper <const gimple_statement_omp_teams>::test): New.
(gimple_omp_return_set_lhs, gimple_omp_return_lhs,
gimple_omp_return_lhs_ptr): Replace bogus downcasts to
gimple_statement_omp_atomic_store with downcasts to
gimple_statement_omp_return, thus requiring that the code be
GIMPLE_OMP_RETURN.
(gimple_resx_region, gimple_resx_set_region): Replace bogus
downcasts to gimple_statement_eh_ctrl with downcasts to
gimple_statement_resx, thus requiring that the code be
GIMPLE_RESX.
(gimple_eh_dispatch_region, gimple_eh_dispatch_set_region):
Replace bogus downcasts to const gimple_statement_eh_ctrl with
downcasts to gimple_statement_eh_dispatch, thus requiring that
the code be GIMPLE_EH_DISPATCH.
(gimple_omp_taskreg_clauses, gimple_omp_taskreg_clauses_ptr)
gimple_omp_taskreg_set_clauses, gimple_omp_taskreg_child_fn,
gimple_omp_taskreg_child_fn_ptr, gimple_omp_taskreg_set_child_fn,
gimple_omp_taskreg_data_arg, gimple_omp_taskreg_data_arg_ptr,
gimple_omp_taskreg_set_data_arg): Replace bogus downcasts to
gimple_statement_omp_parallel with downcasts to
gimple_statement_omp_taskreg, thus requiring that the code be
either GIMPLE_OMP_PARALLEL or GIMPLE_OMP_TASK.
(gimple_omp_target_clauses, gimple_omp_target_clauses_ptr
gimple_omp_target_set_clauses, gimple_omp_target_child_fn
gimple_omp_target_child_fn_ptr, gimple_omp_target_set_child_fn
gimple_omp_target_data_arg, gimple_omp_target_data_arg_ptr
gimple_omp_target_set_data_arg): Replace bogus downcasts to
gimple_statement_omp_parallel with downcasts to
gimple_statement_omp_target, thus requiring that the code be
GIMPLE_OMP_TARGET.
(gimple_omp_teams_clauses, gimple_omp_teams_clauses_ptr
gimple_omp_teams_set_clauses): Replace bogus downcasts to
gimple_statement_omp_single with downcasts to
gimple_statement_omp_teams, thus requiring that the code be
GIMPLE_OMP_TEAMS.
* gimple.c (gimple_build_resx): Fix bogus as_a<> to use
gimple_statement_resx.
(gimple_build_eh_dispatch): Fix bogus as_a<> to use
gimple_statement_eh_dispatch.
From-SVN: r205428
PR tree-optimization/59014
* tree-vrp.c (register_edge_assert_for_1): Don't look
through conversions from non-integral types or through
narrowing conversions.
* gcc.c-torture/execute/pr59014.c: New test.
From-SVN: r205417
PR sanitizer/59258
* ubsan.c (ubsan_source_location): Don't add any location
to ADDR_EXPR in the ctor. Revert 2013-11-22 change.
(ubsan_create_data): Strip block info from LOC.
From-SVN: r205415
PR middle-end/59273
* tree-vect-generic.c (optimize_vector_constructor): Don't optimize
if there isn't optab handler for the corresponding vector PLUS_EXPR.
From-SVN: r205414
PR middle-end/59150
* omp-low.c (lower_rec_input_clause): For reduction with placeholder
of references to constant size types in simd loops, defer emitting
initializer for the new_var, emit it later on only if not using
SIMD arrays for it.
* g++.dg/gomp/pr59150.C: New test.
From-SVN: r205411
PR middle-end/59152
* omp-low.c (expand_omp_for_static_chunk): Don't set loop->latch
for the inner loop if collapse_bb is non-NULL.
(expand_omp_simd): Use cont_bb rather than e->dest as latch.
* c-c++-common/gomp/pr59152.c: New test.
From-SVN: r205410
gcc/
* config/arm/arm.c (arm_legitimize_address): Check xop1 is not
a constant immediate before force_reg.
gcc/testsuite/
* gcc.target/arm/20131120.c: New test.
From-SVN: r205397
2013-11-26 Richard Biener <rguenther@suse.de>
PR tree-optimization/59245
* tree-vrp.c (set_value_range): Assert that we don't have
overflowed constants (but our infinities).
(set_value_range_to_value): Drop all overflow flags.
(vrp_visit_phi_node): Likewise.
(vrp_visit_assignment_or_call): Use set_value_range_to_value
to set a constant range.
* gcc.dg/torture/pr59245.c: New testcase.
From-SVN: r205395
PR bootstrap/55552
* configure.ac (install_gold_as_default): New. Set to yes for
--disable-ld or --enable-gold=default.
(gcc_cv_ld_gold_srcdir): New.
(gcc_cv_ld): Also check in-tree gold if install_gold_as_default
is yes.
(ORIGINAL_LD_BFD_FOR_TARGET): New AC_SUBST.
(ORIGINAL_LD_GOLD_FOR_TARGET): Likewise.
* configure: Regenerated.
* exec-tool.in (ORIGINAL_LD_BFD_FOR_TARGET): New variable.
(ORIGINAL_LD_GOLD_FOR_TARGET): Likewise.
(original) [collect-ld && -fuse-ld=bfd]: Set to
$ORIGINAL_LD_BFD_FOR_TARGET.
(original) [collect-ld && -fuse-ld=gold]: Set to
$ORIGINAL_LD_GOLD_FOR_TARGET.
(dir) [collect-ld && ../gold/ld-new]: Set to gold.
(fast_install) [collect-ld && ../gold/ld-new]: Set to yes.
From-SVN: r205392
gcc/ChangeLog
2013-11-26 Terry Guo <terry.guo@arm.com>
* config/arm/arm.c (require_pic_register): Handle high pic base
register for thumb-1.
(arm_load_pic_register): Also initialize high pic base register.
* doc/invoke.texi: Update documentation for option -mpic-register.
gcc/testsuite/ChangeLog
2013-11-26 Terry Guo <terry.guo@arm.com>
* gcc.target/arm/thumb1-pic-high-reg.c: New case.
* gcc.target/arm/thumb1-pic-single-base.c: New case.
From-SVN: r205391
PR target/58314
PR target/50751
* config/sh/sh.c (max_mov_insn_displacement, disp_addr_displacement):
Prefix function names with 'sh_'. Make them non-static.
* config/sh/sh-protos.h (sh_disp_addr_displacement,
sh_max_mov_insn_displacement): Add declarations.
* config/sh/constraints.md (Q): Reject QImode.
(Sdd): Use match_code "mem".
(Snd): Fix erroneous matching of non-memory operands.
* config/sh/predicates.md (short_displacement_mem_operand): New
predicate.
(general_movsrc_operand): Disallow PC relative QImode loads.
* config/sh/sh.md (*mov<mode>_reg_reg): Remove it.
(*movqi, *movhi): Merge both insns into...
(*mov<mode>): ... this new insn. Replace generic 'm' constraints with
'Snd' and 'Sdd' constraints. Calculate insn length dynamically based
on the operand types.
From-SVN: r205390
