/* Vectorizer
- Copyright (C) 2003-2013 Free Software Foundation, Inc.
+ Copyright (C) 2003-2015 Free Software Foundation, Inc.
Contributed by Dorit Naishlos <dorit@il.ibm.com>
This file is part of GCC.
#include "tree-data-ref.h"
#include "target.h"
-#include "hash-table.h"
/* Used for naming of new temporaries. */
enum vect_var_kind {
scalar elements in one scalar iteration (GROUP_SIZE) multiplied by VF
divided by vector size. */
unsigned int vec_stmts_size;
+ /* Whether the scalar computations use two different operators. */
+ bool two_operators;
};
/* The unrolling factor required to vectorized this SLP instance. */
unsigned int unrolling_factor;
- /* Vectorization costs associated with SLP instance. */
- stmt_vector_for_cost body_cost_vec;
-
/* The group of nodes that contain loads of this SLP instance. */
vec<slp_tree> loads;
-
- /* The first scalar load of the instance. The created vector loads will be
- inserted before this statement. */
- gimple first_load;
} *slp_instance;
#define SLP_INSTANCE_TREE(S) (S)->root
#define SLP_INSTANCE_GROUP_SIZE(S) (S)->group_size
#define SLP_INSTANCE_UNROLLING_FACTOR(S) (S)->unrolling_factor
-#define SLP_INSTANCE_BODY_COST_VEC(S) (S)->body_cost_vec
#define SLP_INSTANCE_LOADS(S) (S)->loads
-#define SLP_INSTANCE_FIRST_LOAD_STMT(S) (S)->first_load
#define SLP_TREE_CHILDREN(S) (S)->children
#define SLP_TREE_SCALAR_STMTS(S) (S)->stmts
#define SLP_TREE_VEC_STMTS(S) (S)->vec_stmts
#define SLP_TREE_NUMBER_OF_VEC_STMTS(S) (S)->vec_stmts_size
#define SLP_TREE_LOAD_PERMUTATION(S) (S)->load_permutation
+#define SLP_TREE_TWO_OPERATORS(S) (S)->two_operators
/* This structure is used in creation of an SLP tree. Each instance
corresponds to the same operand in a group of scalar stmts in an SLP
enum vect_def_type first_dt;
tree first_op_type;
bool first_pattern;
+ bool second_pattern;
} *slp_oprnd_info;
/* Peeling hashtable helpers. */
-struct peel_info_hasher : typed_free_remove <_vect_peel_info>
+struct peel_info_hasher : free_ptr_hash <_vect_peel_info>
{
- typedef _vect_peel_info value_type;
- typedef _vect_peel_info compare_type;
- static inline hashval_t hash (const value_type *);
- static inline bool equal (const value_type *, const compare_type *);
+ static inline hashval_t hash (const _vect_peel_info *);
+ static inline bool equal (const _vect_peel_info *, const _vect_peel_info *);
};
inline hashval_t
-peel_info_hasher::hash (const value_type *peel_info)
+peel_info_hasher::hash (const _vect_peel_info *peel_info)
{
return (hashval_t) peel_info->npeel;
}
inline bool
-peel_info_hasher::equal (const value_type *a, const compare_type *b)
+peel_info_hasher::equal (const _vect_peel_info *a, const _vect_peel_info *b)
{
return (a->npeel == b->npeel);
}
values unknown at compile time. */
int min_profitable_iters;
+ /* Threshold of number of iterations below which vectorzation will not be
+ performed. It is calculated from MIN_PROFITABLE_ITERS and
+ PARAM_MIN_VECT_LOOP_BOUND. */
+ unsigned int th;
+
/* Is the loop vectorizable? */
bool vectorizable;
vec<gimple> reduction_chains;
/* Hash table used to choose the best peeling option. */
- hash_table <peel_info_hasher> peeling_htab;
+ hash_table<peel_info_hasher> *peeling_htab;
+
+ /* Cost vector for a single scalar iteration. */
+ vec<stmt_info_for_cost> scalar_cost_vec;
+
+ /* Cost of a single scalar iteration. */
+ int single_scalar_iteration_cost;
/* Cost data used by the target cost model. */
void *target_cost_data;
fix it up. */
bool operands_swapped;
+ /* True if there are no loop carried data dependencies in the loop.
+ If loop->safelen <= 1, then this is always true, either the loop
+ didn't have any loop carried data dependencies, or the loop is being
+ vectorized guarded with some runtime alias checks, or couldn't
+ be vectorized at all, but then this field shouldn't be used.
+ For loop->safelen >= 2, the user has asserted that there are no
+ backward dependencies, but there still could be loop carried forward
+ dependencies in such loops. This flag will be false if normal
+ vectorizer data dependency analysis would fail or require versioning
+ for alias, but because of loop->safelen >= 2 it has been vectorized
+ even without versioning for alias. E.g. in:
+ #pragma omp simd
+ for (int i = 0; i < m; i++)
+ a[i] = a[i + k] * c;
+ (or #pragma simd or #pragma ivdep) we can vectorize this and it will
+ DTRT even for k > 0 && k < m, but without safelen we would not
+ vectorize this, so this field would be false. */
+ bool no_data_dependencies;
+
/* If if-conversion versioned this loop before conversion, this is the
loop version without if-conversion. */
struct loop *scalar_loop;
cost model. */
#define LOOP_VINFO_NITERS_UNCHANGED(L) (L)->num_iters_unchanged
#define LOOP_VINFO_COST_MODEL_MIN_ITERS(L) (L)->min_profitable_iters
+#define LOOP_VINFO_COST_MODEL_THRESHOLD(L) (L)->th
#define LOOP_VINFO_VECTORIZABLE_P(L) (L)->vectorizable
#define LOOP_VINFO_VECT_FACTOR(L) (L)->vectorization_factor
#define LOOP_VINFO_PTR_MASK(L) (L)->ptr_mask
#define LOOP_VINFO_PEELING_FOR_GAPS(L) (L)->peeling_for_gaps
#define LOOP_VINFO_OPERANDS_SWAPPED(L) (L)->operands_swapped
#define LOOP_VINFO_PEELING_FOR_NITER(L) (L)->peeling_for_niter
+#define LOOP_VINFO_NO_DATA_DEPENDENCIES(L) (L)->no_data_dependencies
#define LOOP_VINFO_SCALAR_LOOP(L) (L)->scalar_loop
+#define LOOP_VINFO_SCALAR_ITERATION_COST(L) (L)->scalar_cost_vec
+#define LOOP_VINFO_SINGLE_SCALAR_ITERATION_COST(L) (L)->single_scalar_iteration_cost
#define LOOP_REQUIRES_VERSIONING_FOR_ALIGNMENT(L) \
- (L)->may_misalign_stmts.length () > 0
+ ((L)->may_misalign_stmts.length () > 0)
#define LOOP_REQUIRES_VERSIONING_FOR_ALIAS(L) \
- (L)->may_alias_ddrs.length () > 0
+ ((L)->may_alias_ddrs.length () > 0)
#define LOOP_VINFO_NITERS_KNOWN_P(L) \
(tree_fits_shwi_p ((L)->num_iters) && tree_to_shwi ((L)->num_iters) > 0)
of this stmt. */
vec<dr_p> same_align_refs;
- /* Selected SIMD clone's function decl. */
- tree simd_clone_fndecl;
+ /* Selected SIMD clone's function info. First vector element
+ is SIMD clone's function decl, followed by a pair of trees (base + step)
+ for linear arguments (pair of NULLs for other arguments). */
+ vec<tree> simd_clone_info;
/* Classify the def of this stmt. */
enum vect_def_type def_type;
is 1. */
unsigned int gap;
+ /* The minimum negative dependence distance this stmt participates in
+ or zero if none. */
+ unsigned int min_neg_dist;
+
/* Not all stmts in the loop need to be vectorized. e.g, the increment
of the loop induction variable and computation of array indexes. relevant
indicates whether the stmt needs to be vectorized. */
/* For loads only, true if this is a gather load. */
bool gather_p;
- bool stride_load_p;
+
+ /* True if this is an access with loop-invariant stride. */
+ bool strided_p;
/* For both loads and stores. */
bool simd_lane_access_p;
#define STMT_VINFO_VECTORIZABLE(S) (S)->vectorizable
#define STMT_VINFO_DATA_REF(S) (S)->data_ref_info
#define STMT_VINFO_GATHER_P(S) (S)->gather_p
-#define STMT_VINFO_STRIDE_LOAD_P(S) (S)->stride_load_p
+#define STMT_VINFO_STRIDED_P(S) (S)->strided_p
#define STMT_VINFO_SIMD_LANE_ACCESS_P(S) (S)->simd_lane_access_p
#define STMT_VINFO_DR_BASE_ADDRESS(S) (S)->dr_base_address
#define STMT_VINFO_RELATED_STMT(S) (S)->related_stmt
#define STMT_VINFO_PATTERN_DEF_SEQ(S) (S)->pattern_def_seq
#define STMT_VINFO_SAME_ALIGN_REFS(S) (S)->same_align_refs
-#define STMT_VINFO_SIMD_CLONE_FNDECL(S) (S)->simd_clone_fndecl
+#define STMT_VINFO_SIMD_CLONE_INFO(S) (S)->simd_clone_info
#define STMT_VINFO_DEF_TYPE(S) (S)->def_type
#define STMT_VINFO_GROUP_FIRST_ELEMENT(S) (S)->first_element
#define STMT_VINFO_GROUP_NEXT_ELEMENT(S) (S)->next_element
#define STMT_VINFO_GROUP_SAME_DR_STMT(S) (S)->same_dr_stmt
#define STMT_VINFO_GROUPED_ACCESS(S) ((S)->first_element != NULL && (S)->data_ref_info)
#define STMT_VINFO_LOOP_PHI_EVOLUTION_PART(S) (S)->loop_phi_evolution_part
+#define STMT_VINFO_MIN_NEG_DIST(S) (S)->min_neg_dist
#define GROUP_FIRST_ELEMENT(S) (S)->first_element
#define GROUP_NEXT_ELEMENT(S) (S)->next_element
conversion. */
#define MAX_INTERM_CVT_STEPS 3
-/* The maximum vectorization factor supported by any target (V32QI). */
-#define MAX_VECTORIZATION_FACTOR 32
+/* The maximum vectorization factor supported by any target (V64QI). */
+#define MAX_VECTORIZATION_FACTOR 64
/* Avoid GTY(()) on stmt_vec_info. */
typedef void *vec_void_p;
static inline bool
unlimited_cost_model (loop_p loop)
{
- if (loop != NULL && loop->force_vect
+ if (loop != NULL && loop->force_vectorize
&& flag_simd_cost_model != VECT_COST_MODEL_DEFAULT)
return flag_simd_cost_model == VECT_COST_MODEL_UNLIMITED;
return (flag_vect_cost_model == VECT_COST_MODEL_UNLIMITED);
extern stmt_vec_info new_stmt_vec_info (gimple stmt, loop_vec_info,
bb_vec_info);
extern void free_stmt_vec_info (gimple stmt);
-extern tree vectorizable_function (gimple, tree, tree);
+extern tree vectorizable_function (gcall *, tree, tree);
extern void vect_model_simple_cost (stmt_vec_info, int, enum vect_def_type *,
stmt_vector_for_cost *,
stmt_vector_for_cost *);
extern bool vect_supportable_shift (enum tree_code, tree);
extern void vect_get_vec_defs (tree, tree, gimple, vec<tree> *,
vec<tree> *, slp_tree, int);
-extern tree vect_gen_perm_mask (tree, unsigned char *);
+extern tree vect_gen_perm_mask_any (tree, const unsigned char *);
+extern tree vect_gen_perm_mask_checked (tree, const unsigned char *);
/* In tree-vect-data-refs.c. */
extern bool vect_can_force_dr_alignment_p (const_tree, unsigned int);
extern bool vect_prune_runtime_alias_test_list (loop_vec_info);
extern tree vect_check_gather (gimple, loop_vec_info, tree *, tree *,
int *);
-extern bool vect_analyze_data_refs (loop_vec_info, bb_vec_info, int *);
+extern bool vect_analyze_data_refs (loop_vec_info, bb_vec_info, int *,
+ unsigned *);
extern tree vect_create_data_ref_ptr (gimple, tree, struct loop *, tree,
tree *, gimple_stmt_iterator *,
- gimple *, bool, bool *);
+ gimple *, bool, bool *,
+ tree = NULL_TREE);
extern tree bump_vector_ptr (tree, gimple, gimple_stmt_iterator *, gimple, tree);
extern tree vect_create_destination_var (tree, tree);
extern bool vect_grouped_store_supported (tree, unsigned HOST_WIDE_INT);
extern void vect_record_grouped_load_vectors (gimple, vec<tree> );
extern tree vect_get_new_vect_var (tree, enum vect_var_kind, const char *);
extern tree vect_create_addr_base_for_vector_ref (gimple, gimple_seq *,
- tree, struct loop *);
+ tree, struct loop *,
+ tree = NULL_TREE);
/* In tree-vect-loop.c. */
/* FORNOW: Used in tree-parloops.c. */
extern bool vectorizable_induction (gimple, gimple_stmt_iterator *, gimple *);
extern tree get_initial_def_for_reduction (gimple, tree, tree *);
extern int vect_min_worthwhile_factor (enum tree_code);
-extern int vect_get_known_peeling_cost (loop_vec_info, int, int *, int,
+extern int vect_get_known_peeling_cost (loop_vec_info, int, int *,
+ stmt_vector_for_cost *,
stmt_vector_for_cost *,
stmt_vector_for_cost *);
-extern int vect_get_single_scalar_iteration_cost (loop_vec_info);
/* In tree-vect-slp.c. */
extern void vect_free_slp_instance (slp_instance);
extern bool vect_transform_slp_perm_load (slp_tree, vec<tree> ,
gimple_stmt_iterator *, int,
slp_instance, bool);
+extern bool vect_slp_analyze_operations (vec<slp_instance> slp_instances,
+ void *);
extern bool vect_schedule_slp (loop_vec_info, bb_vec_info);
-extern void vect_update_slp_costs_according_to_vf (loop_vec_info);
-extern bool vect_analyze_slp (loop_vec_info, bb_vec_info);
+extern bool vect_analyze_slp (loop_vec_info, bb_vec_info, unsigned);
extern bool vect_make_slp_decision (loop_vec_info);
extern void vect_detect_hybrid_slp (loop_vec_info);
extern void vect_get_slp_defs (vec<tree> , slp_tree,
Additional pattern recognition functions can (and will) be added
in the future. */
typedef gimple (* vect_recog_func_ptr) (vec<gimple> *, tree *, tree *);
-#define NUM_PATTERNS 11
+#define NUM_PATTERNS 12
void vect_pattern_recog (loop_vec_info, bb_vec_info);
/* In tree-vectorizer.c. */
projects / gcc.git / blobdiff