URL
https://opencores.org/ocsvn/openrisc/openrisc/trunk
Subversion Repositories openrisc
[/] [openrisc/] [trunk/] [gnu-stable/] [gcc-4.5.1/] [gcc/] [tree-ssa-dom.c] - Rev 856
Go to most recent revision | Compare with Previous | Blame | View Log
/* SSA Dominator optimizations for trees Copyright (C) 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010 Free Software Foundation, Inc. Contributed by Diego Novillo <dnovillo@redhat.com> This file is part of GCC. GCC is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 3, or (at your option) any later version. GCC is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with GCC; see the file COPYING3. If not see <http://www.gnu.org/licenses/>. */ #include "config.h" #include "system.h" #include "coretypes.h" #include "tm.h" #include "tree.h" #include "flags.h" #include "rtl.h" #include "tm_p.h" #include "ggc.h" #include "basic-block.h" #include "cfgloop.h" #include "output.h" #include "expr.h" #include "function.h" #include "diagnostic.h" #include "timevar.h" #include "tree-dump.h" #include "tree-flow.h" #include "domwalk.h" #include "real.h" #include "tree-pass.h" #include "tree-ssa-propagate.h" #include "langhooks.h" #include "params.h" /* This file implements optimizations on the dominator tree. */ /* Representation of a "naked" right-hand-side expression, to be used in recording available expressions in the expression hash table. */ enum expr_kind { EXPR_SINGLE, EXPR_UNARY, EXPR_BINARY, EXPR_CALL }; struct hashable_expr { tree type; enum expr_kind kind; union { struct { tree rhs; } single; struct { enum tree_code op; tree opnd; } unary; struct { enum tree_code op; tree opnd0; tree opnd1; } binary; struct { tree fn; bool pure; size_t nargs; tree *args; } call; } ops; }; /* Structure for recording known values of a conditional expression at the exits from its block. */ struct cond_equivalence { struct hashable_expr cond; tree value; }; /* Structure for recording edge equivalences as well as any pending edge redirections during the dominator optimizer. Computing and storing the edge equivalences instead of creating them on-demand can save significant amounts of time, particularly for pathological cases involving switch statements. These structures live for a single iteration of the dominator optimizer in the edge's AUX field. At the end of an iteration we free each of these structures and update the AUX field to point to any requested redirection target (the code for updating the CFG and SSA graph for edge redirection expects redirection edge targets to be in the AUX field for each edge. */ struct edge_info { /* If this edge creates a simple equivalence, the LHS and RHS of the equivalence will be stored here. */ tree lhs; tree rhs; /* Traversing an edge may also indicate one or more particular conditions are true or false. The number of recorded conditions can vary, but can be determined by the condition's code. So we have an array and its maximum index rather than use a varray. */ struct cond_equivalence *cond_equivalences; unsigned int max_cond_equivalences; }; /* Hash table with expressions made available during the renaming process. When an assignment of the form X_i = EXPR is found, the statement is stored in this table. If the same expression EXPR is later found on the RHS of another statement, it is replaced with X_i (thus performing global redundancy elimination). Similarly as we pass through conditionals we record the conditional itself as having either a true or false value in this table. */ static htab_t avail_exprs; /* Stack of available expressions in AVAIL_EXPRs. Each block pushes any expressions it enters into the hash table along with a marker entry (null). When we finish processing the block, we pop off entries and remove the expressions from the global hash table until we hit the marker. */ typedef struct expr_hash_elt * expr_hash_elt_t; DEF_VEC_P(expr_hash_elt_t); DEF_VEC_ALLOC_P(expr_hash_elt_t,heap); static VEC(expr_hash_elt_t,heap) *avail_exprs_stack; /* Structure for entries in the expression hash table. */ struct expr_hash_elt { /* The value (lhs) of this expression. */ tree lhs; /* The expression (rhs) we want to record. */ struct hashable_expr expr; /* The stmt pointer if this element corresponds to a statement. */ gimple stmt; /* The hash value for RHS. */ hashval_t hash; /* A unique stamp, typically the address of the hash element itself, used in removing entries from the table. */ struct expr_hash_elt *stamp; }; /* Stack of dest,src pairs that need to be restored during finalization. A NULL entry is used to mark the end of pairs which need to be restored during finalization of this block. */ static VEC(tree,heap) *const_and_copies_stack; /* Track whether or not we have changed the control flow graph. */ static bool cfg_altered; /* Bitmap of blocks that have had EH statements cleaned. We should remove their dead edges eventually. */ static bitmap need_eh_cleanup; /* Statistics for dominator optimizations. */ struct opt_stats_d { long num_stmts; long num_exprs_considered; long num_re; long num_const_prop; long num_copy_prop; }; static struct opt_stats_d opt_stats; /* Local functions. */ static void optimize_stmt (basic_block, gimple_stmt_iterator); static tree lookup_avail_expr (gimple, bool); static hashval_t avail_expr_hash (const void *); static hashval_t real_avail_expr_hash (const void *); static int avail_expr_eq (const void *, const void *); static void htab_statistics (FILE *, htab_t); static void record_cond (struct cond_equivalence *); static void record_const_or_copy (tree, tree); static void record_equality (tree, tree); static void record_equivalences_from_phis (basic_block); static void record_equivalences_from_incoming_edge (basic_block); static void eliminate_redundant_computations (gimple_stmt_iterator *); static void record_equivalences_from_stmt (gimple, int); static void dom_thread_across_edge (struct dom_walk_data *, edge); static void dom_opt_leave_block (struct dom_walk_data *, basic_block); static void dom_opt_enter_block (struct dom_walk_data *, basic_block); static void remove_local_expressions_from_table (void); static void restore_vars_to_original_value (void); static edge single_incoming_edge_ignoring_loop_edges (basic_block); /* Given a statement STMT, initialize the hash table element pointed to by ELEMENT. */ static void initialize_hash_element (gimple stmt, tree lhs, struct expr_hash_elt *element) { enum gimple_code code = gimple_code (stmt); struct hashable_expr *expr = &element->expr; if (code == GIMPLE_ASSIGN) { enum tree_code subcode = gimple_assign_rhs_code (stmt); expr->type = NULL_TREE; switch (get_gimple_rhs_class (subcode)) { case GIMPLE_SINGLE_RHS: expr->kind = EXPR_SINGLE; expr->ops.single.rhs = gimple_assign_rhs1 (stmt); break; case GIMPLE_UNARY_RHS: expr->kind = EXPR_UNARY; expr->type = TREE_TYPE (gimple_assign_lhs (stmt)); expr->ops.unary.op = subcode; expr->ops.unary.opnd = gimple_assign_rhs1 (stmt); break; case GIMPLE_BINARY_RHS: expr->kind = EXPR_BINARY; expr->type = TREE_TYPE (gimple_assign_lhs (stmt)); expr->ops.binary.op = subcode; expr->ops.binary.opnd0 = gimple_assign_rhs1 (stmt); expr->ops.binary.opnd1 = gimple_assign_rhs2 (stmt); break; default: gcc_unreachable (); } } else if (code == GIMPLE_COND) { expr->type = boolean_type_node; expr->kind = EXPR_BINARY; expr->ops.binary.op = gimple_cond_code (stmt); expr->ops.binary.opnd0 = gimple_cond_lhs (stmt); expr->ops.binary.opnd1 = gimple_cond_rhs (stmt); } else if (code == GIMPLE_CALL) { size_t nargs = gimple_call_num_args (stmt); size_t i; gcc_assert (gimple_call_lhs (stmt)); expr->type = TREE_TYPE (gimple_call_lhs (stmt)); expr->kind = EXPR_CALL; expr->ops.call.fn = gimple_call_fn (stmt); if (gimple_call_flags (stmt) & (ECF_CONST | ECF_PURE)) expr->ops.call.pure = true; else expr->ops.call.pure = false; expr->ops.call.nargs = nargs; expr->ops.call.args = (tree *) xcalloc (nargs, sizeof (tree)); for (i = 0; i < nargs; i++) expr->ops.call.args[i] = gimple_call_arg (stmt, i); } else if (code == GIMPLE_SWITCH) { expr->type = TREE_TYPE (gimple_switch_index (stmt)); expr->kind = EXPR_SINGLE; expr->ops.single.rhs = gimple_switch_index (stmt); } else if (code == GIMPLE_GOTO) { expr->type = TREE_TYPE (gimple_goto_dest (stmt)); expr->kind = EXPR_SINGLE; expr->ops.single.rhs = gimple_goto_dest (stmt); } else gcc_unreachable (); element->lhs = lhs; element->stmt = stmt; element->hash = avail_expr_hash (element); element->stamp = element; } /* Given a conditional expression COND as a tree, initialize a hashable_expr expression EXPR. The conditional must be a comparison or logical negation. A constant or a variable is not permitted. */ static void initialize_expr_from_cond (tree cond, struct hashable_expr *expr) { expr->type = boolean_type_node; if (COMPARISON_CLASS_P (cond)) { expr->kind = EXPR_BINARY; expr->ops.binary.op = TREE_CODE (cond); expr->ops.binary.opnd0 = TREE_OPERAND (cond, 0); expr->ops.binary.opnd1 = TREE_OPERAND (cond, 1); } else if (TREE_CODE (cond) == TRUTH_NOT_EXPR) { expr->kind = EXPR_UNARY; expr->ops.unary.op = TRUTH_NOT_EXPR; expr->ops.unary.opnd = TREE_OPERAND (cond, 0); } else gcc_unreachable (); } /* Given a hashable_expr expression EXPR and an LHS, initialize the hash table element pointed to by ELEMENT. */ static void initialize_hash_element_from_expr (struct hashable_expr *expr, tree lhs, struct expr_hash_elt *element) { element->expr = *expr; element->lhs = lhs; element->stmt = NULL; element->hash = avail_expr_hash (element); element->stamp = element; } /* Compare two hashable_expr structures for equivalence. They are considered equivalent when the the expressions they denote must necessarily be equal. The logic is intended to follow that of operand_equal_p in fold-const.c */ static bool hashable_expr_equal_p (const struct hashable_expr *expr0, const struct hashable_expr *expr1) { tree type0 = expr0->type; tree type1 = expr1->type; /* If either type is NULL, there is nothing to check. */ if ((type0 == NULL_TREE) ^ (type1 == NULL_TREE)) return false; /* If both types don't have the same signedness, precision, and mode, then we can't consider them equal. */ if (type0 != type1 && (TREE_CODE (type0) == ERROR_MARK || TREE_CODE (type1) == ERROR_MARK || TYPE_UNSIGNED (type0) != TYPE_UNSIGNED (type1) || TYPE_PRECISION (type0) != TYPE_PRECISION (type1) || TYPE_MODE (type0) != TYPE_MODE (type1))) return false; if (expr0->kind != expr1->kind) return false; switch (expr0->kind) { case EXPR_SINGLE: return operand_equal_p (expr0->ops.single.rhs, expr1->ops.single.rhs, 0); case EXPR_UNARY: if (expr0->ops.unary.op != expr1->ops.unary.op) return false; if ((CONVERT_EXPR_CODE_P (expr0->ops.unary.op) || expr0->ops.unary.op == NON_LVALUE_EXPR) && TYPE_UNSIGNED (expr0->type) != TYPE_UNSIGNED (expr1->type)) return false; return operand_equal_p (expr0->ops.unary.opnd, expr1->ops.unary.opnd, 0); case EXPR_BINARY: { if (expr0->ops.binary.op != expr1->ops.binary.op) return false; if (operand_equal_p (expr0->ops.binary.opnd0, expr1->ops.binary.opnd0, 0) && operand_equal_p (expr0->ops.binary.opnd1, expr1->ops.binary.opnd1, 0)) return true; /* For commutative ops, allow the other order. */ return (commutative_tree_code (expr0->ops.binary.op) && operand_equal_p (expr0->ops.binary.opnd0, expr1->ops.binary.opnd1, 0) && operand_equal_p (expr0->ops.binary.opnd1, expr1->ops.binary.opnd0, 0)); } case EXPR_CALL: { size_t i; /* If the calls are to different functions, then they clearly cannot be equal. */ if (! operand_equal_p (expr0->ops.call.fn, expr1->ops.call.fn, 0)) return false; if (! expr0->ops.call.pure) return false; if (expr0->ops.call.nargs != expr1->ops.call.nargs) return false; for (i = 0; i < expr0->ops.call.nargs; i++) if (! operand_equal_p (expr0->ops.call.args[i], expr1->ops.call.args[i], 0)) return false; return true; } default: gcc_unreachable (); } } /* Compute a hash value for a hashable_expr value EXPR and a previously accumulated hash value VAL. If two hashable_expr values compare equal with hashable_expr_equal_p, they must hash to the same value, given an identical value of VAL. The logic is intended to follow iterative_hash_expr in tree.c. */ static hashval_t iterative_hash_hashable_expr (const struct hashable_expr *expr, hashval_t val) { switch (expr->kind) { case EXPR_SINGLE: val = iterative_hash_expr (expr->ops.single.rhs, val); break; case EXPR_UNARY: val = iterative_hash_object (expr->ops.unary.op, val); /* Make sure to include signedness in the hash computation. Don't hash the type, that can lead to having nodes which compare equal according to operand_equal_p, but which have different hash codes. */ if (CONVERT_EXPR_CODE_P (expr->ops.unary.op) || expr->ops.unary.op == NON_LVALUE_EXPR) val += TYPE_UNSIGNED (expr->type); val = iterative_hash_expr (expr->ops.unary.opnd, val); break; case EXPR_BINARY: val = iterative_hash_object (expr->ops.binary.op, val); if (commutative_tree_code (expr->ops.binary.op)) val = iterative_hash_exprs_commutative (expr->ops.binary.opnd0, expr->ops.binary.opnd1, val); else { val = iterative_hash_expr (expr->ops.binary.opnd0, val); val = iterative_hash_expr (expr->ops.binary.opnd1, val); } break; case EXPR_CALL: { size_t i; enum tree_code code = CALL_EXPR; val = iterative_hash_object (code, val); val = iterative_hash_expr (expr->ops.call.fn, val); for (i = 0; i < expr->ops.call.nargs; i++) val = iterative_hash_expr (expr->ops.call.args[i], val); } break; default: gcc_unreachable (); } return val; } /* Print a diagnostic dump of an expression hash table entry. */ static void print_expr_hash_elt (FILE * stream, const struct expr_hash_elt *element) { if (element->stmt) fprintf (stream, "STMT "); else fprintf (stream, "COND "); if (element->lhs) { print_generic_expr (stream, element->lhs, 0); fprintf (stream, " = "); } switch (element->expr.kind) { case EXPR_SINGLE: print_generic_expr (stream, element->expr.ops.single.rhs, 0); break; case EXPR_UNARY: fprintf (stream, "%s ", tree_code_name[element->expr.ops.unary.op]); print_generic_expr (stream, element->expr.ops.unary.opnd, 0); break; case EXPR_BINARY: print_generic_expr (stream, element->expr.ops.binary.opnd0, 0); fprintf (stream, " %s ", tree_code_name[element->expr.ops.binary.op]); print_generic_expr (stream, element->expr.ops.binary.opnd1, 0); break; case EXPR_CALL: { size_t i; size_t nargs = element->expr.ops.call.nargs; print_generic_expr (stream, element->expr.ops.call.fn, 0); fprintf (stream, " ("); for (i = 0; i < nargs; i++) { print_generic_expr (stream, element->expr.ops.call.args[i], 0); if (i + 1 < nargs) fprintf (stream, ", "); } fprintf (stream, ")"); } break; } fprintf (stream, "\n"); if (element->stmt) { fprintf (stream, " "); print_gimple_stmt (stream, element->stmt, 0, 0); } } /* Delete an expr_hash_elt and reclaim its storage. */ static void free_expr_hash_elt (void *elt) { struct expr_hash_elt *element = ((struct expr_hash_elt *)elt); if (element->expr.kind == EXPR_CALL) free (element->expr.ops.call.args); free (element); } /* Allocate an EDGE_INFO for edge E and attach it to E. Return the new EDGE_INFO structure. */ static struct edge_info * allocate_edge_info (edge e) { struct edge_info *edge_info; edge_info = XCNEW (struct edge_info); e->aux = edge_info; return edge_info; } /* Free all EDGE_INFO structures associated with edges in the CFG. If a particular edge can be threaded, copy the redirection target from the EDGE_INFO structure into the edge's AUX field as required by code to update the CFG and SSA graph for jump threading. */ static void free_all_edge_infos (void) { basic_block bb; edge_iterator ei; edge e; FOR_EACH_BB (bb) { FOR_EACH_EDGE (e, ei, bb->preds) { struct edge_info *edge_info = (struct edge_info *) e->aux; if (edge_info) { if (edge_info->cond_equivalences) free (edge_info->cond_equivalences); free (edge_info); e->aux = NULL; } } } } /* Jump threading, redundancy elimination and const/copy propagation. This pass may expose new symbols that need to be renamed into SSA. For every new symbol exposed, its corresponding bit will be set in VARS_TO_RENAME. */ static unsigned int tree_ssa_dominator_optimize (void) { struct dom_walk_data walk_data; memset (&opt_stats, 0, sizeof (opt_stats)); /* Create our hash tables. */ avail_exprs = htab_create (1024, real_avail_expr_hash, avail_expr_eq, free_expr_hash_elt); avail_exprs_stack = VEC_alloc (expr_hash_elt_t, heap, 20); const_and_copies_stack = VEC_alloc (tree, heap, 20); need_eh_cleanup = BITMAP_ALLOC (NULL); /* Setup callbacks for the generic dominator tree walker. */ walk_data.dom_direction = CDI_DOMINATORS; walk_data.initialize_block_local_data = NULL; walk_data.before_dom_children = dom_opt_enter_block; walk_data.after_dom_children = dom_opt_leave_block; /* Right now we only attach a dummy COND_EXPR to the global data pointer. When we attach more stuff we'll need to fill this out with a real structure. */ walk_data.global_data = NULL; walk_data.block_local_data_size = 0; /* Now initialize the dominator walker. */ init_walk_dominator_tree (&walk_data); calculate_dominance_info (CDI_DOMINATORS); cfg_altered = false; /* We need to know loop structures in order to avoid destroying them in jump threading. Note that we still can e.g. thread through loop headers to an exit edge, or through loop header to the loop body, assuming that we update the loop info. */ loop_optimizer_init (LOOPS_HAVE_SIMPLE_LATCHES); /* Initialize the value-handle array. */ threadedge_initialize_values (); /* We need accurate information regarding back edges in the CFG for jump threading; this may include back edges that are not part of a single loop. */ mark_dfs_back_edges (); /* Recursively walk the dominator tree optimizing statements. */ walk_dominator_tree (&walk_data, ENTRY_BLOCK_PTR); { gimple_stmt_iterator gsi; basic_block bb; FOR_EACH_BB (bb) {for (gsi = gsi_start_bb (bb); !gsi_end_p (gsi); gsi_next (&gsi)) update_stmt_if_modified (gsi_stmt (gsi)); } } /* If we exposed any new variables, go ahead and put them into SSA form now, before we handle jump threading. This simplifies interactions between rewriting of _DECL nodes into SSA form and rewriting SSA_NAME nodes into SSA form after block duplication and CFG manipulation. */ update_ssa (TODO_update_ssa); free_all_edge_infos (); /* Thread jumps, creating duplicate blocks as needed. */ cfg_altered |= thread_through_all_blocks (first_pass_instance); if (cfg_altered) free_dominance_info (CDI_DOMINATORS); /* Removal of statements may make some EH edges dead. Purge such edges from the CFG as needed. */ if (!bitmap_empty_p (need_eh_cleanup)) { unsigned i; bitmap_iterator bi; /* Jump threading may have created forwarder blocks from blocks needing EH cleanup; the new successor of these blocks, which has inherited from the original block, needs the cleanup. */ EXECUTE_IF_SET_IN_BITMAP (need_eh_cleanup, 0, i, bi) { basic_block bb = BASIC_BLOCK (i); if (single_succ_p (bb) == 1 && (single_succ_edge (bb)->flags & EDGE_EH) == 0) { bitmap_clear_bit (need_eh_cleanup, i); bitmap_set_bit (need_eh_cleanup, single_succ (bb)->index); } } gimple_purge_all_dead_eh_edges (need_eh_cleanup); bitmap_zero (need_eh_cleanup); } statistics_counter_event (cfun, "Redundant expressions eliminated", opt_stats.num_re); statistics_counter_event (cfun, "Constants propagated", opt_stats.num_const_prop); statistics_counter_event (cfun, "Copies propagated", opt_stats.num_copy_prop); /* Debugging dumps. */ if (dump_file && (dump_flags & TDF_STATS)) dump_dominator_optimization_stats (dump_file); loop_optimizer_finalize (); /* Delete our main hashtable. */ htab_delete (avail_exprs); /* And finalize the dominator walker. */ fini_walk_dominator_tree (&walk_data); /* Free asserted bitmaps and stacks. */ BITMAP_FREE (need_eh_cleanup); VEC_free (expr_hash_elt_t, heap, avail_exprs_stack); VEC_free (tree, heap, const_and_copies_stack); /* Free the value-handle array. */ threadedge_finalize_values (); ssa_name_values = NULL; return 0; } static bool gate_dominator (void) { return flag_tree_dom != 0; } struct gimple_opt_pass pass_dominator = { { GIMPLE_PASS, "dom", /* name */ gate_dominator, /* gate */ tree_ssa_dominator_optimize, /* execute */ NULL, /* sub */ NULL, /* next */ 0, /* static_pass_number */ TV_TREE_SSA_DOMINATOR_OPTS, /* tv_id */ PROP_cfg | PROP_ssa, /* properties_required */ 0, /* properties_provided */ 0, /* properties_destroyed */ 0, /* todo_flags_start */ TODO_dump_func | TODO_update_ssa | TODO_cleanup_cfg | TODO_verify_ssa /* todo_flags_finish */ } }; /* Given a conditional statement CONDSTMT, convert the condition to a canonical form. */ static void canonicalize_comparison (gimple condstmt) { tree op0; tree op1; enum tree_code code; gcc_assert (gimple_code (condstmt) == GIMPLE_COND); op0 = gimple_cond_lhs (condstmt); op1 = gimple_cond_rhs (condstmt); code = gimple_cond_code (condstmt); /* If it would be profitable to swap the operands, then do so to canonicalize the statement, enabling better optimization. By placing canonicalization of such expressions here we transparently keep statements in canonical form, even when the statement is modified. */ if (tree_swap_operands_p (op0, op1, false)) { /* For relationals we need to swap the operands and change the code. */ if (code == LT_EXPR || code == GT_EXPR || code == LE_EXPR || code == GE_EXPR) { code = swap_tree_comparison (code); gimple_cond_set_code (condstmt, code); gimple_cond_set_lhs (condstmt, op1); gimple_cond_set_rhs (condstmt, op0); update_stmt (condstmt); } } } /* Initialize local stacks for this optimizer and record equivalences upon entry to BB. Equivalences can come from the edge traversed to reach BB or they may come from PHI nodes at the start of BB. */ /* Remove all the expressions in LOCALS from TABLE, stopping when there are LIMIT entries left in LOCALs. */ static void remove_local_expressions_from_table (void) { /* Remove all the expressions made available in this block. */ while (VEC_length (expr_hash_elt_t, avail_exprs_stack) > 0) { expr_hash_elt_t victim = VEC_pop (expr_hash_elt_t, avail_exprs_stack); void **slot; if (victim == NULL) break; /* This must precede the actual removal from the hash table, as ELEMENT and the table entry may share a call argument vector which will be freed during removal. */ if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, "<<<< "); print_expr_hash_elt (dump_file, victim); } slot = htab_find_slot_with_hash (avail_exprs, victim, victim->hash, NO_INSERT); gcc_assert (slot && *slot == (void *) victim); htab_clear_slot (avail_exprs, slot); } } /* Use the source/dest pairs in CONST_AND_COPIES_STACK to restore CONST_AND_COPIES to its original state, stopping when we hit a NULL marker. */ static void restore_vars_to_original_value (void) { while (VEC_length (tree, const_and_copies_stack) > 0) { tree prev_value, dest; dest = VEC_pop (tree, const_and_copies_stack); if (dest == NULL) break; if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, "<<<< COPY "); print_generic_expr (dump_file, dest, 0); fprintf (dump_file, " = "); print_generic_expr (dump_file, SSA_NAME_VALUE (dest), 0); fprintf (dump_file, "\n"); } prev_value = VEC_pop (tree, const_and_copies_stack); set_ssa_name_value (dest, prev_value); } } /* A trivial wrapper so that we can present the generic jump threading code with a simple API for simplifying statements. */ static tree simplify_stmt_for_jump_threading (gimple stmt, gimple within_stmt ATTRIBUTE_UNUSED) { return lookup_avail_expr (stmt, false); } /* Wrapper for common code to attempt to thread an edge. For example, it handles lazily building the dummy condition and the bookkeeping when jump threading is successful. */ static void dom_thread_across_edge (struct dom_walk_data *walk_data, edge e) { if (! walk_data->global_data) { gimple dummy_cond = gimple_build_cond (NE_EXPR, integer_zero_node, integer_zero_node, NULL, NULL); walk_data->global_data = dummy_cond; } thread_across_edge ((gimple) walk_data->global_data, e, false, &const_and_copies_stack, simplify_stmt_for_jump_threading); } /* PHI nodes can create equivalences too. Ignoring any alternatives which are the same as the result, if all the alternatives are equal, then the PHI node creates an equivalence. */ static void record_equivalences_from_phis (basic_block bb) { gimple_stmt_iterator gsi; for (gsi = gsi_start_phis (bb); !gsi_end_p (gsi); gsi_next (&gsi)) { gimple phi = gsi_stmt (gsi); tree lhs = gimple_phi_result (phi); tree rhs = NULL; size_t i; for (i = 0; i < gimple_phi_num_args (phi); i++) { tree t = gimple_phi_arg_def (phi, i); /* Ignore alternatives which are the same as our LHS. Since LHS is a PHI_RESULT, it is known to be a SSA_NAME, so we can simply compare pointers. */ if (lhs == t) continue; /* If we have not processed an alternative yet, then set RHS to this alternative. */ if (rhs == NULL) rhs = t; /* If we have processed an alternative (stored in RHS), then see if it is equal to this one. If it isn't, then stop the search. */ else if (! operand_equal_for_phi_arg_p (rhs, t)) break; } /* If we had no interesting alternatives, then all the RHS alternatives must have been the same as LHS. */ if (!rhs) rhs = lhs; /* If we managed to iterate through each PHI alternative without breaking out of the loop, then we have a PHI which may create a useful equivalence. We do not need to record unwind data for this, since this is a true assignment and not an equivalence inferred from a comparison. All uses of this ssa name are dominated by this assignment, so unwinding just costs time and space. */ if (i == gimple_phi_num_args (phi) && may_propagate_copy (lhs, rhs)) set_ssa_name_value (lhs, rhs); } } /* Ignoring loop backedges, if BB has precisely one incoming edge then return that edge. Otherwise return NULL. */ static edge single_incoming_edge_ignoring_loop_edges (basic_block bb) { edge retval = NULL; edge e; edge_iterator ei; FOR_EACH_EDGE (e, ei, bb->preds) { /* A loop back edge can be identified by the destination of the edge dominating the source of the edge. */ if (dominated_by_p (CDI_DOMINATORS, e->src, e->dest)) continue; /* If we have already seen a non-loop edge, then we must have multiple incoming non-loop edges and thus we return NULL. */ if (retval) return NULL; /* This is the first non-loop incoming edge we have found. Record it. */ retval = e; } return retval; } /* Record any equivalences created by the incoming edge to BB. If BB has more than one incoming edge, then no equivalence is created. */ static void record_equivalences_from_incoming_edge (basic_block bb) { edge e; basic_block parent; struct edge_info *edge_info; /* If our parent block ended with a control statement, then we may be able to record some equivalences based on which outgoing edge from the parent was followed. */ parent = get_immediate_dominator (CDI_DOMINATORS, bb); e = single_incoming_edge_ignoring_loop_edges (bb); /* If we had a single incoming edge from our parent block, then enter any data associated with the edge into our tables. */ if (e && e->src == parent) { unsigned int i; edge_info = (struct edge_info *) e->aux; if (edge_info) { tree lhs = edge_info->lhs; tree rhs = edge_info->rhs; struct cond_equivalence *cond_equivalences = edge_info->cond_equivalences; if (lhs) record_equality (lhs, rhs); if (cond_equivalences) for (i = 0; i < edge_info->max_cond_equivalences; i++) record_cond (&cond_equivalences[i]); } } } /* Dump SSA statistics on FILE. */ void dump_dominator_optimization_stats (FILE *file) { fprintf (file, "Total number of statements: %6ld\n\n", opt_stats.num_stmts); fprintf (file, "Exprs considered for dominator optimizations: %6ld\n", opt_stats.num_exprs_considered); fprintf (file, "\nHash table statistics:\n"); fprintf (file, " avail_exprs: "); htab_statistics (file, avail_exprs); } /* Dump SSA statistics on stderr. */ void debug_dominator_optimization_stats (void) { dump_dominator_optimization_stats (stderr); } /* Dump statistics for the hash table HTAB. */ static void htab_statistics (FILE *file, htab_t htab) { fprintf (file, "size %ld, %ld elements, %f collision/search ratio\n", (long) htab_size (htab), (long) htab_elements (htab), htab_collisions (htab)); } /* Enter condition equivalence into the expression hash table. This indicates that a conditional expression has a known boolean value. */ static void record_cond (struct cond_equivalence *p) { struct expr_hash_elt *element = XCNEW (struct expr_hash_elt); void **slot; initialize_hash_element_from_expr (&p->cond, p->value, element); slot = htab_find_slot_with_hash (avail_exprs, (void *)element, element->hash, INSERT); if (*slot == NULL) { *slot = (void *) element; if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, "1>>> "); print_expr_hash_elt (dump_file, element); } VEC_safe_push (expr_hash_elt_t, heap, avail_exprs_stack, element); } else free (element); } /* Build a cond_equivalence record indicating that the comparison CODE holds between operands OP0 and OP1. */ static void build_and_record_new_cond (enum tree_code code, tree op0, tree op1, struct cond_equivalence *p) { struct hashable_expr *cond = &p->cond; gcc_assert (TREE_CODE_CLASS (code) == tcc_comparison); cond->type = boolean_type_node; cond->kind = EXPR_BINARY; cond->ops.binary.op = code; cond->ops.binary.opnd0 = op0; cond->ops.binary.opnd1 = op1; p->value = boolean_true_node; } /* Record that COND is true and INVERTED is false into the edge information structure. Also record that any conditions dominated by COND are true as well. For example, if a < b is true, then a <= b must also be true. */ static void record_conditions (struct edge_info *edge_info, tree cond, tree inverted) { tree op0, op1; if (!COMPARISON_CLASS_P (cond)) return; op0 = TREE_OPERAND (cond, 0); op1 = TREE_OPERAND (cond, 1); switch (TREE_CODE (cond)) { case LT_EXPR: case GT_EXPR: if (FLOAT_TYPE_P (TREE_TYPE (op0))) { edge_info->max_cond_equivalences = 6; edge_info->cond_equivalences = XNEWVEC (struct cond_equivalence, 6); build_and_record_new_cond (ORDERED_EXPR, op0, op1, &edge_info->cond_equivalences[4]); build_and_record_new_cond (LTGT_EXPR, op0, op1, &edge_info->cond_equivalences[5]); } else { edge_info->max_cond_equivalences = 4; edge_info->cond_equivalences = XNEWVEC (struct cond_equivalence, 4); } build_and_record_new_cond ((TREE_CODE (cond) == LT_EXPR ? LE_EXPR : GE_EXPR), op0, op1, &edge_info->cond_equivalences[2]); build_and_record_new_cond (NE_EXPR, op0, op1, &edge_info->cond_equivalences[3]); break; case GE_EXPR: case LE_EXPR: if (FLOAT_TYPE_P (TREE_TYPE (op0))) { edge_info->max_cond_equivalences = 3; edge_info->cond_equivalences = XNEWVEC (struct cond_equivalence, 3); build_and_record_new_cond (ORDERED_EXPR, op0, op1, &edge_info->cond_equivalences[2]); } else { edge_info->max_cond_equivalences = 2; edge_info->cond_equivalences = XNEWVEC (struct cond_equivalence, 2); } break; case EQ_EXPR: if (FLOAT_TYPE_P (TREE_TYPE (op0))) { edge_info->max_cond_equivalences = 5; edge_info->cond_equivalences = XNEWVEC (struct cond_equivalence, 5); build_and_record_new_cond (ORDERED_EXPR, op0, op1, &edge_info->cond_equivalences[4]); } else { edge_info->max_cond_equivalences = 4; edge_info->cond_equivalences = XNEWVEC (struct cond_equivalence, 4); } build_and_record_new_cond (LE_EXPR, op0, op1, &edge_info->cond_equivalences[2]); build_and_record_new_cond (GE_EXPR, op0, op1, &edge_info->cond_equivalences[3]); break; case UNORDERED_EXPR: edge_info->max_cond_equivalences = 8; edge_info->cond_equivalences = XNEWVEC (struct cond_equivalence, 8); build_and_record_new_cond (NE_EXPR, op0, op1, &edge_info->cond_equivalences[2]); build_and_record_new_cond (UNLE_EXPR, op0, op1, &edge_info->cond_equivalences[3]); build_and_record_new_cond (UNGE_EXPR, op0, op1, &edge_info->cond_equivalences[4]); build_and_record_new_cond (UNEQ_EXPR, op0, op1, &edge_info->cond_equivalences[5]); build_and_record_new_cond (UNLT_EXPR, op0, op1, &edge_info->cond_equivalences[6]); build_and_record_new_cond (UNGT_EXPR, op0, op1, &edge_info->cond_equivalences[7]); break; case UNLT_EXPR: case UNGT_EXPR: edge_info->max_cond_equivalences = 4; edge_info->cond_equivalences = XNEWVEC (struct cond_equivalence, 4); build_and_record_new_cond ((TREE_CODE (cond) == UNLT_EXPR ? UNLE_EXPR : UNGE_EXPR), op0, op1, &edge_info->cond_equivalences[2]); build_and_record_new_cond (NE_EXPR, op0, op1, &edge_info->cond_equivalences[3]); break; case UNEQ_EXPR: edge_info->max_cond_equivalences = 4; edge_info->cond_equivalences = XNEWVEC (struct cond_equivalence, 4); build_and_record_new_cond (UNLE_EXPR, op0, op1, &edge_info->cond_equivalences[2]); build_and_record_new_cond (UNGE_EXPR, op0, op1, &edge_info->cond_equivalences[3]); break; case LTGT_EXPR: edge_info->max_cond_equivalences = 4; edge_info->cond_equivalences = XNEWVEC (struct cond_equivalence, 4); build_and_record_new_cond (NE_EXPR, op0, op1, &edge_info->cond_equivalences[2]); build_and_record_new_cond (ORDERED_EXPR, op0, op1, &edge_info->cond_equivalences[3]); break; default: edge_info->max_cond_equivalences = 2; edge_info->cond_equivalences = XNEWVEC (struct cond_equivalence, 2); break; } /* Now store the original true and false conditions into the first two slots. */ initialize_expr_from_cond (cond, &edge_info->cond_equivalences[0].cond); edge_info->cond_equivalences[0].value = boolean_true_node; /* It is possible for INVERTED to be the negation of a comparison, and not a valid RHS or GIMPLE_COND condition. This happens because invert_truthvalue may return such an expression when asked to invert a floating-point comparison. These comparisons are not assumed to obey the trichotomy law. */ initialize_expr_from_cond (inverted, &edge_info->cond_equivalences[1].cond); edge_info->cond_equivalences[1].value = boolean_false_node; } /* A helper function for record_const_or_copy and record_equality. Do the work of recording the value and undo info. */ static void record_const_or_copy_1 (tree x, tree y, tree prev_x) { set_ssa_name_value (x, y); if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, "0>>> COPY "); print_generic_expr (dump_file, x, 0); fprintf (dump_file, " = "); print_generic_expr (dump_file, y, 0); fprintf (dump_file, "\n"); } VEC_reserve (tree, heap, const_and_copies_stack, 2); VEC_quick_push (tree, const_and_copies_stack, prev_x); VEC_quick_push (tree, const_and_copies_stack, x); } /* Return the loop depth of the basic block of the defining statement of X. This number should not be treated as absolutely correct because the loop information may not be completely up-to-date when dom runs. However, it will be relatively correct, and as more passes are taught to keep loop info up to date, the result will become more and more accurate. */ int loop_depth_of_name (tree x) { gimple defstmt; basic_block defbb; /* If it's not an SSA_NAME, we have no clue where the definition is. */ if (TREE_CODE (x) != SSA_NAME) return 0; /* Otherwise return the loop depth of the defining statement's bb. Note that there may not actually be a bb for this statement, if the ssa_name is live on entry. */ defstmt = SSA_NAME_DEF_STMT (x); defbb = gimple_bb (defstmt); if (!defbb) return 0; return defbb->loop_depth; } /* Record that X is equal to Y in const_and_copies. Record undo information in the block-local vector. */ static void record_const_or_copy (tree x, tree y) { tree prev_x = SSA_NAME_VALUE (x); gcc_assert (TREE_CODE (x) == SSA_NAME); if (TREE_CODE (y) == SSA_NAME) { tree tmp = SSA_NAME_VALUE (y); if (tmp) y = tmp; } record_const_or_copy_1 (x, y, prev_x); } /* Similarly, but assume that X and Y are the two operands of an EQ_EXPR. This constrains the cases in which we may treat this as assignment. */ static void record_equality (tree x, tree y) { tree prev_x = NULL, prev_y = NULL; if (TREE_CODE (x) == SSA_NAME) prev_x = SSA_NAME_VALUE (x); if (TREE_CODE (y) == SSA_NAME) prev_y = SSA_NAME_VALUE (y); /* If one of the previous values is invariant, or invariant in more loops (by depth), then use that. Otherwise it doesn't matter which value we choose, just so long as we canonicalize on one value. */ if (is_gimple_min_invariant (y)) ; else if (is_gimple_min_invariant (x) || (loop_depth_of_name (x) <= loop_depth_of_name (y))) prev_x = x, x = y, y = prev_x, prev_x = prev_y; else if (prev_x && is_gimple_min_invariant (prev_x)) x = y, y = prev_x, prev_x = prev_y; else if (prev_y) y = prev_y; /* After the swapping, we must have one SSA_NAME. */ if (TREE_CODE (x) != SSA_NAME) return; /* For IEEE, -0.0 == 0.0, so we don't necessarily know the sign of a variable compared against zero. If we're honoring signed zeros, then we cannot record this value unless we know that the value is nonzero. */ if (HONOR_SIGNED_ZEROS (TYPE_MODE (TREE_TYPE (x))) && (TREE_CODE (y) != REAL_CST || REAL_VALUES_EQUAL (dconst0, TREE_REAL_CST (y)))) return; record_const_or_copy_1 (x, y, prev_x); } /* Returns true when STMT is a simple iv increment. It detects the following situation: i_1 = phi (..., i_2) i_2 = i_1 +/- ... */ static bool simple_iv_increment_p (gimple stmt) { tree lhs, preinc; gimple phi; size_t i; if (gimple_code (stmt) != GIMPLE_ASSIGN) return false; lhs = gimple_assign_lhs (stmt); if (TREE_CODE (lhs) != SSA_NAME) return false; if (gimple_assign_rhs_code (stmt) != PLUS_EXPR && gimple_assign_rhs_code (stmt) != MINUS_EXPR) return false; preinc = gimple_assign_rhs1 (stmt); if (TREE_CODE (preinc) != SSA_NAME) return false; phi = SSA_NAME_DEF_STMT (preinc); if (gimple_code (phi) != GIMPLE_PHI) return false; for (i = 0; i < gimple_phi_num_args (phi); i++) if (gimple_phi_arg_def (phi, i) == lhs) return true; return false; } /* CONST_AND_COPIES is a table which maps an SSA_NAME to the current known value for that SSA_NAME (or NULL if no value is known). Propagate values from CONST_AND_COPIES into the PHI nodes of the successors of BB. */ static void cprop_into_successor_phis (basic_block bb) { edge e; edge_iterator ei; FOR_EACH_EDGE (e, ei, bb->succs) { int indx; gimple_stmt_iterator gsi; /* If this is an abnormal edge, then we do not want to copy propagate into the PHI alternative associated with this edge. */ if (e->flags & EDGE_ABNORMAL) continue; gsi = gsi_start_phis (e->dest); if (gsi_end_p (gsi)) continue; indx = e->dest_idx; for ( ; !gsi_end_p (gsi); gsi_next (&gsi)) { tree new_val; use_operand_p orig_p; tree orig_val; gimple phi = gsi_stmt (gsi); /* The alternative may be associated with a constant, so verify it is an SSA_NAME before doing anything with it. */ orig_p = gimple_phi_arg_imm_use_ptr (phi, indx); orig_val = get_use_from_ptr (orig_p); if (TREE_CODE (orig_val) != SSA_NAME) continue; /* If we have *ORIG_P in our constant/copy table, then replace ORIG_P with its value in our constant/copy table. */ new_val = SSA_NAME_VALUE (orig_val); if (new_val && new_val != orig_val && (TREE_CODE (new_val) == SSA_NAME || is_gimple_min_invariant (new_val)) && may_propagate_copy (orig_val, new_val)) propagate_value (orig_p, new_val); } } } /* We have finished optimizing BB, record any information implied by taking a specific outgoing edge from BB. */ static void record_edge_info (basic_block bb) { gimple_stmt_iterator gsi = gsi_last_bb (bb); struct edge_info *edge_info; if (! gsi_end_p (gsi)) { gimple stmt = gsi_stmt (gsi); location_t loc = gimple_location (stmt); if (gimple_code (stmt) == GIMPLE_SWITCH) { tree index = gimple_switch_index (stmt); if (TREE_CODE (index) == SSA_NAME) { int i; int n_labels = gimple_switch_num_labels (stmt); tree *info = XCNEWVEC (tree, last_basic_block); edge e; edge_iterator ei; for (i = 0; i < n_labels; i++) { tree label = gimple_switch_label (stmt, i); basic_block target_bb = label_to_block (CASE_LABEL (label)); if (CASE_HIGH (label) || !CASE_LOW (label) || info[target_bb->index]) info[target_bb->index] = error_mark_node; else info[target_bb->index] = label; } FOR_EACH_EDGE (e, ei, bb->succs) { basic_block target_bb = e->dest; tree label = info[target_bb->index]; if (label != NULL && label != error_mark_node) { tree x = fold_convert_loc (loc, TREE_TYPE (index), CASE_LOW (label)); edge_info = allocate_edge_info (e); edge_info->lhs = index; edge_info->rhs = x; } } free (info); } } /* A COND_EXPR may create equivalences too. */ if (gimple_code (stmt) == GIMPLE_COND) { edge true_edge; edge false_edge; tree op0 = gimple_cond_lhs (stmt); tree op1 = gimple_cond_rhs (stmt); enum tree_code code = gimple_cond_code (stmt); extract_true_false_edges_from_block (bb, &true_edge, &false_edge); /* Special case comparing booleans against a constant as we know the value of OP0 on both arms of the branch. i.e., we can record an equivalence for OP0 rather than COND. */ if ((code == EQ_EXPR || code == NE_EXPR) && TREE_CODE (op0) == SSA_NAME && TREE_CODE (TREE_TYPE (op0)) == BOOLEAN_TYPE && is_gimple_min_invariant (op1)) { if (code == EQ_EXPR) { edge_info = allocate_edge_info (true_edge); edge_info->lhs = op0; edge_info->rhs = (integer_zerop (op1) ? boolean_false_node : boolean_true_node); edge_info = allocate_edge_info (false_edge); edge_info->lhs = op0; edge_info->rhs = (integer_zerop (op1) ? boolean_true_node : boolean_false_node); } else { edge_info = allocate_edge_info (true_edge); edge_info->lhs = op0; edge_info->rhs = (integer_zerop (op1) ? boolean_true_node : boolean_false_node); edge_info = allocate_edge_info (false_edge); edge_info->lhs = op0; edge_info->rhs = (integer_zerop (op1) ? boolean_false_node : boolean_true_node); } } else if (is_gimple_min_invariant (op0) && (TREE_CODE (op1) == SSA_NAME || is_gimple_min_invariant (op1))) { tree cond = build2 (code, boolean_type_node, op0, op1); tree inverted = invert_truthvalue_loc (loc, cond); struct edge_info *edge_info; edge_info = allocate_edge_info (true_edge); record_conditions (edge_info, cond, inverted); if (code == EQ_EXPR) { edge_info->lhs = op1; edge_info->rhs = op0; } edge_info = allocate_edge_info (false_edge); record_conditions (edge_info, inverted, cond); if (TREE_CODE (inverted) == EQ_EXPR) { edge_info->lhs = op1; edge_info->rhs = op0; } } else if (TREE_CODE (op0) == SSA_NAME && (is_gimple_min_invariant (op1) || TREE_CODE (op1) == SSA_NAME)) { tree cond = build2 (code, boolean_type_node, op0, op1); tree inverted = invert_truthvalue_loc (loc, cond); struct edge_info *edge_info; edge_info = allocate_edge_info (true_edge); record_conditions (edge_info, cond, inverted); if (code == EQ_EXPR) { edge_info->lhs = op0; edge_info->rhs = op1; } edge_info = allocate_edge_info (false_edge); record_conditions (edge_info, inverted, cond); if (TREE_CODE (inverted) == EQ_EXPR) { edge_info->lhs = op0; edge_info->rhs = op1; } } } /* ??? TRUTH_NOT_EXPR can create an equivalence too. */ } } static void dom_opt_enter_block (struct dom_walk_data *walk_data ATTRIBUTE_UNUSED, basic_block bb) { gimple_stmt_iterator gsi; if (dump_file && (dump_flags & TDF_DETAILS)) fprintf (dump_file, "\n\nOptimizing block #%d\n\n", bb->index); /* Push a marker on the stacks of local information so that we know how far to unwind when we finalize this block. */ VEC_safe_push (expr_hash_elt_t, heap, avail_exprs_stack, NULL); VEC_safe_push (tree, heap, const_and_copies_stack, NULL_TREE); record_equivalences_from_incoming_edge (bb); /* PHI nodes can create equivalences too. */ record_equivalences_from_phis (bb); for (gsi = gsi_start_bb (bb); !gsi_end_p (gsi); gsi_next (&gsi)) optimize_stmt (bb, gsi); /* Now prepare to process dominated blocks. */ record_edge_info (bb); cprop_into_successor_phis (bb); } /* We have finished processing the dominator children of BB, perform any finalization actions in preparation for leaving this node in the dominator tree. */ static void dom_opt_leave_block (struct dom_walk_data *walk_data, basic_block bb) { gimple last; /* If we have an outgoing edge to a block with multiple incoming and outgoing edges, then we may be able to thread the edge, i.e., we may be able to statically determine which of the outgoing edges will be traversed when the incoming edge from BB is traversed. */ if (single_succ_p (bb) && (single_succ_edge (bb)->flags & EDGE_ABNORMAL) == 0 && potentially_threadable_block (single_succ (bb))) { dom_thread_across_edge (walk_data, single_succ_edge (bb)); } else if ((last = last_stmt (bb)) && gimple_code (last) == GIMPLE_COND && EDGE_COUNT (bb->succs) == 2 && (EDGE_SUCC (bb, 0)->flags & EDGE_ABNORMAL) == 0 && (EDGE_SUCC (bb, 1)->flags & EDGE_ABNORMAL) == 0) { edge true_edge, false_edge; extract_true_false_edges_from_block (bb, &true_edge, &false_edge); /* Only try to thread the edge if it reaches a target block with more than one predecessor and more than one successor. */ if (potentially_threadable_block (true_edge->dest)) { struct edge_info *edge_info; unsigned int i; /* Push a marker onto the available expression stack so that we unwind any expressions related to the TRUE arm before processing the false arm below. */ VEC_safe_push (expr_hash_elt_t, heap, avail_exprs_stack, NULL); VEC_safe_push (tree, heap, const_and_copies_stack, NULL_TREE); edge_info = (struct edge_info *) true_edge->aux; /* If we have info associated with this edge, record it into our equivalence tables. */ if (edge_info) { struct cond_equivalence *cond_equivalences = edge_info->cond_equivalences; tree lhs = edge_info->lhs; tree rhs = edge_info->rhs; /* If we have a simple NAME = VALUE equivalence, record it. */ if (lhs && TREE_CODE (lhs) == SSA_NAME) record_const_or_copy (lhs, rhs); /* If we have 0 = COND or 1 = COND equivalences, record them into our expression hash tables. */ if (cond_equivalences) for (i = 0; i < edge_info->max_cond_equivalences; i++) record_cond (&cond_equivalences[i]); } dom_thread_across_edge (walk_data, true_edge); /* And restore the various tables to their state before we threaded this edge. */ remove_local_expressions_from_table (); } /* Similarly for the ELSE arm. */ if (potentially_threadable_block (false_edge->dest)) { struct edge_info *edge_info; unsigned int i; VEC_safe_push (tree, heap, const_and_copies_stack, NULL_TREE); edge_info = (struct edge_info *) false_edge->aux; /* If we have info associated with this edge, record it into our equivalence tables. */ if (edge_info) { struct cond_equivalence *cond_equivalences = edge_info->cond_equivalences; tree lhs = edge_info->lhs; tree rhs = edge_info->rhs; /* If we have a simple NAME = VALUE equivalence, record it. */ if (lhs && TREE_CODE (lhs) == SSA_NAME) record_const_or_copy (lhs, rhs); /* If we have 0 = COND or 1 = COND equivalences, record them into our expression hash tables. */ if (cond_equivalences) for (i = 0; i < edge_info->max_cond_equivalences; i++) record_cond (&cond_equivalences[i]); } /* Now thread the edge. */ dom_thread_across_edge (walk_data, false_edge); /* No need to remove local expressions from our tables or restore vars to their original value as that will be done immediately below. */ } } remove_local_expressions_from_table (); restore_vars_to_original_value (); } /* Search for redundant computations in STMT. If any are found, then replace them with the variable holding the result of the computation. If safe, record this expression into the available expression hash table. */ static void eliminate_redundant_computations (gimple_stmt_iterator* gsi) { tree expr_type; tree cached_lhs; bool insert = true; bool assigns_var_p = false; gimple stmt = gsi_stmt (*gsi); tree def = gimple_get_lhs (stmt); /* Certain expressions on the RHS can be optimized away, but can not themselves be entered into the hash tables. */ if (! def || TREE_CODE (def) != SSA_NAME || SSA_NAME_OCCURS_IN_ABNORMAL_PHI (def) || gimple_vdef (stmt) /* Do not record equivalences for increments of ivs. This would create overlapping live ranges for a very questionable gain. */ || simple_iv_increment_p (stmt)) insert = false; /* Check if the expression has been computed before. */ cached_lhs = lookup_avail_expr (stmt, insert); opt_stats.num_exprs_considered++; /* Get the type of the expression we are trying to optimize. */ if (is_gimple_assign (stmt)) { expr_type = TREE_TYPE (gimple_assign_lhs (stmt)); assigns_var_p = true; } else if (gimple_code (stmt) == GIMPLE_COND) expr_type = boolean_type_node; else if (is_gimple_call (stmt)) { gcc_assert (gimple_call_lhs (stmt)); expr_type = TREE_TYPE (gimple_call_lhs (stmt)); assigns_var_p = true; } else if (gimple_code (stmt) == GIMPLE_SWITCH) expr_type = TREE_TYPE (gimple_switch_index (stmt)); else gcc_unreachable (); if (!cached_lhs) return; /* It is safe to ignore types here since we have already done type checking in the hashing and equality routines. In fact type checking here merely gets in the way of constant propagation. Also, make sure that it is safe to propagate CACHED_LHS into the expression in STMT. */ if ((TREE_CODE (cached_lhs) != SSA_NAME && (assigns_var_p || useless_type_conversion_p (expr_type, TREE_TYPE (cached_lhs)))) || may_propagate_copy_into_stmt (stmt, cached_lhs)) { #if defined ENABLE_CHECKING gcc_assert (TREE_CODE (cached_lhs) == SSA_NAME || is_gimple_min_invariant (cached_lhs)); #endif if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, " Replaced redundant expr '"); print_gimple_expr (dump_file, stmt, 0, dump_flags); fprintf (dump_file, "' with '"); print_generic_expr (dump_file, cached_lhs, dump_flags); fprintf (dump_file, "'\n"); } opt_stats.num_re++; if (assigns_var_p && !useless_type_conversion_p (expr_type, TREE_TYPE (cached_lhs))) cached_lhs = fold_convert (expr_type, cached_lhs); propagate_tree_value_into_stmt (gsi, cached_lhs); /* Since it is always necessary to mark the result as modified, perhaps we should move this into propagate_tree_value_into_stmt itself. */ gimple_set_modified (gsi_stmt (*gsi), true); } } /* STMT, a GIMPLE_ASSIGN, may create certain equivalences, in either the available expressions table or the const_and_copies table. Detect and record those equivalences. */ /* We handle only very simple copy equivalences here. The heavy lifing is done by eliminate_redundant_computations. */ static void record_equivalences_from_stmt (gimple stmt, int may_optimize_p) { tree lhs; enum tree_code lhs_code; gcc_assert (is_gimple_assign (stmt)); lhs = gimple_assign_lhs (stmt); lhs_code = TREE_CODE (lhs); if (lhs_code == SSA_NAME && gimple_assign_single_p (stmt)) { tree rhs = gimple_assign_rhs1 (stmt); /* If the RHS of the assignment is a constant or another variable that may be propagated, register it in the CONST_AND_COPIES table. We do not need to record unwind data for this, since this is a true assignment and not an equivalence inferred from a comparison. All uses of this ssa name are dominated by this assignment, so unwinding just costs time and space. */ if (may_optimize_p && (TREE_CODE (rhs) == SSA_NAME || is_gimple_min_invariant (rhs))) { if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, "==== ASGN "); print_generic_expr (dump_file, lhs, 0); fprintf (dump_file, " = "); print_generic_expr (dump_file, rhs, 0); fprintf (dump_file, "\n"); } set_ssa_name_value (lhs, rhs); } } /* A memory store, even an aliased store, creates a useful equivalence. By exchanging the LHS and RHS, creating suitable vops and recording the result in the available expression table, we may be able to expose more redundant loads. */ if (!gimple_has_volatile_ops (stmt) && gimple_references_memory_p (stmt) && gimple_assign_single_p (stmt) && (TREE_CODE (gimple_assign_rhs1 (stmt)) == SSA_NAME || is_gimple_min_invariant (gimple_assign_rhs1 (stmt))) && !is_gimple_reg (lhs)) { tree rhs = gimple_assign_rhs1 (stmt); gimple new_stmt; /* Build a new statement with the RHS and LHS exchanged. */ if (TREE_CODE (rhs) == SSA_NAME) { /* NOTE tuples. The call to gimple_build_assign below replaced a call to build_gimple_modify_stmt, which did not set the SSA_NAME_DEF_STMT on the LHS of the assignment. Doing so may cause an SSA validation failure, as the LHS may be a default-initialized name and should have no definition. I'm a bit dubious of this, as the artificial statement that we generate here may in fact be ill-formed, but it is simply used as an internal device in this pass, and never becomes part of the CFG. */ gimple defstmt = SSA_NAME_DEF_STMT (rhs); new_stmt = gimple_build_assign (rhs, lhs); SSA_NAME_DEF_STMT (rhs) = defstmt; } else new_stmt = gimple_build_assign (rhs, lhs); gimple_set_vuse (new_stmt, gimple_vdef (stmt)); /* Finally enter the statement into the available expression table. */ lookup_avail_expr (new_stmt, true); } } /* Replace *OP_P in STMT with any known equivalent value for *OP_P from CONST_AND_COPIES. */ static void cprop_operand (gimple stmt, use_operand_p op_p) { tree val; tree op = USE_FROM_PTR (op_p); /* If the operand has a known constant value or it is known to be a copy of some other variable, use the value or copy stored in CONST_AND_COPIES. */ val = SSA_NAME_VALUE (op); if (val && val != op) { /* Do not change the base variable in the virtual operand tables. That would make it impossible to reconstruct the renamed virtual operand if we later modify this statement. Also only allow the new value to be an SSA_NAME for propagation into virtual operands. */ if (!is_gimple_reg (op) && (TREE_CODE (val) != SSA_NAME || is_gimple_reg (val) || get_virtual_var (val) != get_virtual_var (op))) return; /* Do not replace hard register operands in asm statements. */ if (gimple_code (stmt) == GIMPLE_ASM && !may_propagate_copy_into_asm (op)) return; /* Certain operands are not allowed to be copy propagated due to their interaction with exception handling and some GCC extensions. */ if (!may_propagate_copy (op, val)) return; /* Do not propagate addresses that point to volatiles into memory stmts without volatile operands. */ if (POINTER_TYPE_P (TREE_TYPE (val)) && TYPE_VOLATILE (TREE_TYPE (TREE_TYPE (val))) && gimple_has_mem_ops (stmt) && !gimple_has_volatile_ops (stmt)) return; /* Do not propagate copies if the propagated value is at a deeper loop depth than the propagatee. Otherwise, this may move loop variant variables outside of their loops and prevent coalescing opportunities. If the value was loop invariant, it will be hoisted by LICM and exposed for copy propagation. */ if (loop_depth_of_name (val) > loop_depth_of_name (op)) return; /* Do not propagate copies into simple IV increment statements. See PR23821 for how this can disturb IV analysis. */ if (TREE_CODE (val) != INTEGER_CST && simple_iv_increment_p (stmt)) return; /* Dump details. */ if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, " Replaced '"); print_generic_expr (dump_file, op, dump_flags); fprintf (dump_file, "' with %s '", (TREE_CODE (val) != SSA_NAME ? "constant" : "variable")); print_generic_expr (dump_file, val, dump_flags); fprintf (dump_file, "'\n"); } if (TREE_CODE (val) != SSA_NAME) opt_stats.num_const_prop++; else opt_stats.num_copy_prop++; propagate_value (op_p, val); /* And note that we modified this statement. This is now safe, even if we changed virtual operands since we will rescan the statement and rewrite its operands again. */ gimple_set_modified (stmt, true); } } /* CONST_AND_COPIES is a table which maps an SSA_NAME to the current known value for that SSA_NAME (or NULL if no value is known). Propagate values from CONST_AND_COPIES into the uses, vuses and vdef_ops of STMT. */ static void cprop_into_stmt (gimple stmt) { use_operand_p op_p; ssa_op_iter iter; FOR_EACH_SSA_USE_OPERAND (op_p, stmt, iter, SSA_OP_ALL_USES) { if (TREE_CODE (USE_FROM_PTR (op_p)) == SSA_NAME) cprop_operand (stmt, op_p); } } /* Optimize the statement pointed to by iterator SI. We try to perform some simplistic global redundancy elimination and constant propagation: 1- To detect global redundancy, we keep track of expressions that have been computed in this block and its dominators. If we find that the same expression is computed more than once, we eliminate repeated computations by using the target of the first one. 2- Constant values and copy assignments. This is used to do very simplistic constant and copy propagation. When a constant or copy assignment is found, we map the value on the RHS of the assignment to the variable in the LHS in the CONST_AND_COPIES table. */ static void optimize_stmt (basic_block bb, gimple_stmt_iterator si) { gimple stmt, old_stmt; bool may_optimize_p; bool modified_p = false; old_stmt = stmt = gsi_stmt (si); if (gimple_code (stmt) == GIMPLE_COND) canonicalize_comparison (stmt); update_stmt_if_modified (stmt); opt_stats.num_stmts++; if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, "Optimizing statement "); print_gimple_stmt (dump_file, stmt, 0, TDF_SLIM); } /* Const/copy propagate into USES, VUSES and the RHS of VDEFs. */ cprop_into_stmt (stmt); /* If the statement has been modified with constant replacements, fold its RHS before checking for redundant computations. */ if (gimple_modified_p (stmt)) { tree rhs = NULL; /* Try to fold the statement making sure that STMT is kept up to date. */ if (fold_stmt (&si)) { stmt = gsi_stmt (si); gimple_set_modified (stmt, true); if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, " Folded to: "); print_gimple_stmt (dump_file, stmt, 0, TDF_SLIM); } } /* We only need to consider cases that can yield a gimple operand. */ if (gimple_assign_single_p (stmt)) rhs = gimple_assign_rhs1 (stmt); else if (gimple_code (stmt) == GIMPLE_GOTO) rhs = gimple_goto_dest (stmt); else if (gimple_code (stmt) == GIMPLE_SWITCH) /* This should never be an ADDR_EXPR. */ rhs = gimple_switch_index (stmt); if (rhs && TREE_CODE (rhs) == ADDR_EXPR) recompute_tree_invariant_for_addr_expr (rhs); /* Indicate that maybe_clean_or_replace_eh_stmt needs to be called, even if fold_stmt updated the stmt already and thus cleared gimple_modified_p flag on it. */ modified_p = true; } /* Check for redundant computations. Do this optimization only for assignments that have no volatile ops and conditionals. */ may_optimize_p = (!gimple_has_volatile_ops (stmt) && ((is_gimple_assign (stmt) && !gimple_rhs_has_side_effects (stmt)) || (is_gimple_call (stmt) && gimple_call_lhs (stmt) != NULL_TREE && !gimple_rhs_has_side_effects (stmt)) || gimple_code (stmt) == GIMPLE_COND || gimple_code (stmt) == GIMPLE_SWITCH)); if (may_optimize_p) { if (gimple_code (stmt) == GIMPLE_CALL) { /* Resolve __builtin_constant_p. If it hasn't been folded to integer_one_node by now, it's fairly certain that the value simply isn't constant. */ tree callee = gimple_call_fndecl (stmt); if (callee && DECL_BUILT_IN_CLASS (callee) == BUILT_IN_NORMAL && DECL_FUNCTION_CODE (callee) == BUILT_IN_CONSTANT_P) { propagate_tree_value_into_stmt (&si, integer_zero_node); stmt = gsi_stmt (si); } } update_stmt_if_modified (stmt); eliminate_redundant_computations (&si); stmt = gsi_stmt (si); } /* Record any additional equivalences created by this statement. */ if (is_gimple_assign (stmt)) record_equivalences_from_stmt (stmt, may_optimize_p); /* If STMT is a COND_EXPR and it was modified, then we may know where it goes. If that is the case, then mark the CFG as altered. This will cause us to later call remove_unreachable_blocks and cleanup_tree_cfg when it is safe to do so. It is not safe to clean things up here since removal of edges and such can trigger the removal of PHI nodes, which in turn can release SSA_NAMEs to the manager. That's all fine and good, except that once SSA_NAMEs are released to the manager, we must not call create_ssa_name until all references to released SSA_NAMEs have been eliminated. All references to the deleted SSA_NAMEs can not be eliminated until we remove unreachable blocks. We can not remove unreachable blocks until after we have completed any queued jump threading. We can not complete any queued jump threads until we have taken appropriate variables out of SSA form. Taking variables out of SSA form can call create_ssa_name and thus we lose. Ultimately I suspect we're going to need to change the interface into the SSA_NAME manager. */ if (gimple_modified_p (stmt) || modified_p) { tree val = NULL; update_stmt_if_modified (stmt); if (gimple_code (stmt) == GIMPLE_COND) val = fold_binary_loc (gimple_location (stmt), gimple_cond_code (stmt), boolean_type_node, gimple_cond_lhs (stmt), gimple_cond_rhs (stmt)); else if (gimple_code (stmt) == GIMPLE_SWITCH) val = gimple_switch_index (stmt); if (val && TREE_CODE (val) == INTEGER_CST && find_taken_edge (bb, val)) cfg_altered = true; /* If we simplified a statement in such a way as to be shown that it cannot trap, update the eh information and the cfg to match. */ if (maybe_clean_or_replace_eh_stmt (old_stmt, stmt)) { bitmap_set_bit (need_eh_cleanup, bb->index); if (dump_file && (dump_flags & TDF_DETAILS)) fprintf (dump_file, " Flagged to clear EH edges.\n"); } } } /* Search for an existing instance of STMT in the AVAIL_EXPRS table. If found, return its LHS. Otherwise insert STMT in the table and return NULL_TREE. Also, when an expression is first inserted in the table, it is also is also added to AVAIL_EXPRS_STACK, so that it can be removed when we finish processing this block and its children. */ static tree lookup_avail_expr (gimple stmt, bool insert) { void **slot; tree lhs; tree temp; struct expr_hash_elt element; /* Get LHS of assignment or call, else NULL_TREE. */ lhs = gimple_get_lhs (stmt); initialize_hash_element (stmt, lhs, &element); if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, "LKUP "); print_expr_hash_elt (dump_file, &element); } /* Don't bother remembering constant assignments and copy operations. Constants and copy operations are handled by the constant/copy propagator in optimize_stmt. */ if (element.expr.kind == EXPR_SINGLE && (TREE_CODE (element.expr.ops.single.rhs) == SSA_NAME || is_gimple_min_invariant (element.expr.ops.single.rhs))) return NULL_TREE; /* Finally try to find the expression in the main expression hash table. */ slot = htab_find_slot_with_hash (avail_exprs, &element, element.hash, (insert ? INSERT : NO_INSERT)); if (slot == NULL) return NULL_TREE; if (*slot == NULL) { struct expr_hash_elt *element2 = XNEW (struct expr_hash_elt); *element2 = element; element2->stamp = element2; *slot = (void *) element2; if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, "2>>> "); print_expr_hash_elt (dump_file, element2); } VEC_safe_push (expr_hash_elt_t, heap, avail_exprs_stack, element2); return NULL_TREE; } /* Extract the LHS of the assignment so that it can be used as the current definition of another variable. */ lhs = ((struct expr_hash_elt *)*slot)->lhs; /* See if the LHS appears in the CONST_AND_COPIES table. If it does, then use the value from the const_and_copies table. */ if (TREE_CODE (lhs) == SSA_NAME) { temp = SSA_NAME_VALUE (lhs); if (temp) lhs = temp; } if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, "FIND: "); print_generic_expr (dump_file, lhs, 0); fprintf (dump_file, "\n"); } return lhs; } /* Hashing and equality functions for AVAIL_EXPRS. We compute a value number for expressions using the code of the expression and the SSA numbers of its operands. */ static hashval_t avail_expr_hash (const void *p) { gimple stmt = ((const struct expr_hash_elt *)p)->stmt; const struct hashable_expr *expr = &((const struct expr_hash_elt *)p)->expr; tree vuse; hashval_t val = 0; val = iterative_hash_hashable_expr (expr, val); /* If the hash table entry is not associated with a statement, then we can just hash the expression and not worry about virtual operands and such. */ if (!stmt) return val; /* Add the SSA version numbers of the vuse operand. This is important because compound variables like arrays are not renamed in the operands. Rather, the rename is done on the virtual variable representing all the elements of the array. */ if ((vuse = gimple_vuse (stmt))) val = iterative_hash_expr (vuse, val); return val; } static hashval_t real_avail_expr_hash (const void *p) { return ((const struct expr_hash_elt *)p)->hash; } static int avail_expr_eq (const void *p1, const void *p2) { gimple stmt1 = ((const struct expr_hash_elt *)p1)->stmt; const struct hashable_expr *expr1 = &((const struct expr_hash_elt *)p1)->expr; const struct expr_hash_elt *stamp1 = ((const struct expr_hash_elt *)p1)->stamp; gimple stmt2 = ((const struct expr_hash_elt *)p2)->stmt; const struct hashable_expr *expr2 = &((const struct expr_hash_elt *)p2)->expr; const struct expr_hash_elt *stamp2 = ((const struct expr_hash_elt *)p2)->stamp; /* This case should apply only when removing entries from the table. */ if (stamp1 == stamp2) return true; /* FIXME tuples: We add stmts to a hash table and them modify them. To detect the case that we modify a stmt and then search for it, we assume that the hash is always modified by that change. We have to fully check why this doesn't happen on trunk or rewrite this in a more reliable (and easier to understand) way. */ if (((const struct expr_hash_elt *)p1)->hash != ((const struct expr_hash_elt *)p2)->hash) return false; /* In case of a collision, both RHS have to be identical and have the same VUSE operands. */ if (hashable_expr_equal_p (expr1, expr2) && types_compatible_p (expr1->type, expr2->type)) { /* Note that STMT1 and/or STMT2 may be NULL. */ return ((stmt1 ? gimple_vuse (stmt1) : NULL_TREE) == (stmt2 ? gimple_vuse (stmt2) : NULL_TREE)); } return false; } /* PHI-ONLY copy and constant propagation. This pass is meant to clean up degenerate PHIs created by or exposed by jump threading. */ /* Given PHI, return its RHS if the PHI is a degenerate, otherwise return NULL. */ tree degenerate_phi_result (gimple phi) { tree lhs = gimple_phi_result (phi); tree val = NULL; size_t i; /* Ignoring arguments which are the same as LHS, if all the remaining arguments are the same, then the PHI is a degenerate and has the value of that common argument. */ for (i = 0; i < gimple_phi_num_args (phi); i++) { tree arg = gimple_phi_arg_def (phi, i); if (arg == lhs) continue; else if (!arg) break; else if (!val) val = arg; else if (arg == val) continue; /* We bring in some of operand_equal_p not only to speed things up, but also to avoid crashing when dereferencing the type of a released SSA name. */ else if (TREE_CODE (val) != TREE_CODE (arg) || TREE_CODE (val) == SSA_NAME || !operand_equal_p (arg, val, 0)) break; } return (i == gimple_phi_num_args (phi) ? val : NULL); } /* Given a statement STMT, which is either a PHI node or an assignment, remove it from the IL. */ static void remove_stmt_or_phi (gimple stmt) { gimple_stmt_iterator gsi = gsi_for_stmt (stmt); if (gimple_code (stmt) == GIMPLE_PHI) remove_phi_node (&gsi, true); else { gsi_remove (&gsi, true); release_defs (stmt); } } /* Given a statement STMT, which is either a PHI node or an assignment, return the "rhs" of the node, in the case of a non-degenerate phi, NULL is returned. */ static tree get_rhs_or_phi_arg (gimple stmt) { if (gimple_code (stmt) == GIMPLE_PHI) return degenerate_phi_result (stmt); else if (gimple_assign_single_p (stmt)) return gimple_assign_rhs1 (stmt); else gcc_unreachable (); } /* Given a statement STMT, which is either a PHI node or an assignment, return the "lhs" of the node. */ static tree get_lhs_or_phi_result (gimple stmt) { if (gimple_code (stmt) == GIMPLE_PHI) return gimple_phi_result (stmt); else if (is_gimple_assign (stmt)) return gimple_assign_lhs (stmt); else gcc_unreachable (); } /* Propagate RHS into all uses of LHS (when possible). RHS and LHS are derived from STMT, which is passed in solely so that we can remove it if propagation is successful. When propagating into a PHI node or into a statement which turns into a trivial copy or constant initialization, set the appropriate bit in INTERESTING_NAMEs so that we will visit those nodes as well in an effort to pick up secondary optimization opportunities. */ static void propagate_rhs_into_lhs (gimple stmt, tree lhs, tree rhs, bitmap interesting_names) { /* First verify that propagation is valid and isn't going to move a loop variant variable outside its loop. */ if (! SSA_NAME_OCCURS_IN_ABNORMAL_PHI (lhs) && (TREE_CODE (rhs) != SSA_NAME || ! SSA_NAME_OCCURS_IN_ABNORMAL_PHI (rhs)) && may_propagate_copy (lhs, rhs) && loop_depth_of_name (lhs) >= loop_depth_of_name (rhs)) { use_operand_p use_p; imm_use_iterator iter; gimple use_stmt; bool all = true; /* Dump details. */ if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, " Replacing '"); print_generic_expr (dump_file, lhs, dump_flags); fprintf (dump_file, "' with %s '", (TREE_CODE (rhs) != SSA_NAME ? "constant" : "variable")); print_generic_expr (dump_file, rhs, dump_flags); fprintf (dump_file, "'\n"); } /* Walk over every use of LHS and try to replace the use with RHS. At this point the only reason why such a propagation would not be successful would be if the use occurs in an ASM_EXPR. */ FOR_EACH_IMM_USE_STMT (use_stmt, iter, lhs) { /* Leave debug stmts alone. If we succeed in propagating all non-debug uses, we'll drop the DEF, and propagation into debug stmts will occur then. */ if (gimple_debug_bind_p (use_stmt)) continue; /* It's not always safe to propagate into an ASM_EXPR. */ if (gimple_code (use_stmt) == GIMPLE_ASM && ! may_propagate_copy_into_asm (lhs)) { all = false; continue; } /* It's not ok to propagate into the definition stmt of RHS. <bb 9>: # prephitmp.12_36 = PHI <g_67.1_6(9)> g_67.1_6 = prephitmp.12_36; goto <bb 9>; While this is strictly all dead code we do not want to deal with this here. */ if (TREE_CODE (rhs) == SSA_NAME && SSA_NAME_DEF_STMT (rhs) == use_stmt) { all = false; continue; } /* Dump details. */ if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, " Original statement:"); print_gimple_stmt (dump_file, use_stmt, 0, dump_flags); } /* Propagate the RHS into this use of the LHS. */ FOR_EACH_IMM_USE_ON_STMT (use_p, iter) propagate_value (use_p, rhs); /* Special cases to avoid useless calls into the folding routines, operand scanning, etc. First, propagation into a PHI may cause the PHI to become a degenerate, so mark the PHI as interesting. No other actions are necessary. Second, if we're propagating a virtual operand and the propagation does not change the underlying _DECL node for the virtual operand, then no further actions are necessary. */ if (gimple_code (use_stmt) == GIMPLE_PHI || (! is_gimple_reg (lhs) && TREE_CODE (rhs) == SSA_NAME && SSA_NAME_VAR (lhs) == SSA_NAME_VAR (rhs))) { /* Dump details. */ if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, " Updated statement:"); print_gimple_stmt (dump_file, use_stmt, 0, dump_flags); } /* Propagation into a PHI may expose new degenerate PHIs, so mark the result of the PHI as interesting. */ if (gimple_code (use_stmt) == GIMPLE_PHI) { tree result = get_lhs_or_phi_result (use_stmt); bitmap_set_bit (interesting_names, SSA_NAME_VERSION (result)); } continue; } /* From this point onward we are propagating into a real statement. Folding may (or may not) be possible, we may expose new operands, expose dead EH edges, etc. */ /* NOTE tuples. In the tuples world, fold_stmt_inplace cannot fold a call that simplifies to a constant, because the GIMPLE_CALL must be replaced by a GIMPLE_ASSIGN, and there is no way to effect such a transformation in-place. We might want to consider using the more general fold_stmt here. */ fold_stmt_inplace (use_stmt); /* Sometimes propagation can expose new operands to the renamer. */ update_stmt (use_stmt); /* Dump details. */ if (dump_file && (dump_flags & TDF_DETAILS)) { fprintf (dump_file, " Updated statement:"); print_gimple_stmt (dump_file, use_stmt, 0, dump_flags); } /* If we replaced a variable index with a constant, then we would need to update the invariant flag for ADDR_EXPRs. */ if (gimple_assign_single_p (use_stmt) && TREE_CODE (gimple_assign_rhs1 (use_stmt)) == ADDR_EXPR) recompute_tree_invariant_for_addr_expr (gimple_assign_rhs1 (use_stmt)); /* If we cleaned up EH information from the statement, mark its containing block as needing EH cleanups. */ if (maybe_clean_or_replace_eh_stmt (use_stmt, use_stmt)) { bitmap_set_bit (need_eh_cleanup, gimple_bb (use_stmt)->index); if (dump_file && (dump_flags & TDF_DETAILS)) fprintf (dump_file, " Flagged to clear EH edges.\n"); } /* Propagation may expose new trivial copy/constant propagation opportunities. */ if (gimple_assign_single_p (use_stmt) && TREE_CODE (gimple_assign_lhs (use_stmt)) == SSA_NAME && (TREE_CODE (gimple_assign_rhs1 (use_stmt)) == SSA_NAME || is_gimple_min_invariant (gimple_assign_rhs1 (use_stmt)))) { tree result = get_lhs_or_phi_result (use_stmt); bitmap_set_bit (interesting_names, SSA_NAME_VERSION (result)); } /* Propagation into these nodes may make certain edges in the CFG unexecutable. We want to identify them as PHI nodes at the destination of those unexecutable edges may become degenerates. */ else if (gimple_code (use_stmt) == GIMPLE_COND || gimple_code (use_stmt) == GIMPLE_SWITCH || gimple_code (use_stmt) == GIMPLE_GOTO) { tree val; if (gimple_code (use_stmt) == GIMPLE_COND) val = fold_binary_loc (gimple_location (use_stmt), gimple_cond_code (use_stmt), boolean_type_node, gimple_cond_lhs (use_stmt), gimple_cond_rhs (use_stmt)); else if (gimple_code (use_stmt) == GIMPLE_SWITCH) val = gimple_switch_index (use_stmt); else val = gimple_goto_dest (use_stmt); if (val && is_gimple_min_invariant (val)) { basic_block bb = gimple_bb (use_stmt); edge te = find_taken_edge (bb, val); edge_iterator ei; edge e; gimple_stmt_iterator gsi, psi; /* Remove all outgoing edges except TE. */ for (ei = ei_start (bb->succs); (e = ei_safe_edge (ei));) { if (e != te) { /* Mark all the PHI nodes at the destination of the unexecutable edge as interesting. */ for (psi = gsi_start_phis (e->dest); !gsi_end_p (psi); gsi_next (&psi)) { gimple phi = gsi_stmt (psi); tree result = gimple_phi_result (phi); int version = SSA_NAME_VERSION (result); bitmap_set_bit (interesting_names, version); } te->probability += e->probability; te->count += e->count; remove_edge (e); cfg_altered = true; } else ei_next (&ei); } gsi = gsi_last_bb (gimple_bb (use_stmt)); gsi_remove (&gsi, true); /* And fixup the flags on the single remaining edge. */ te->flags &= ~(EDGE_TRUE_VALUE | EDGE_FALSE_VALUE); te->flags &= ~EDGE_ABNORMAL; te->flags |= EDGE_FALLTHRU; if (te->probability > REG_BR_PROB_BASE) te->probability = REG_BR_PROB_BASE; } } } /* Ensure there is nothing else to do. */ gcc_assert (!all || has_zero_uses (lhs)); /* If we were able to propagate away all uses of LHS, then we can remove STMT. */ if (all) remove_stmt_or_phi (stmt); } } /* STMT is either a PHI node (potentially a degenerate PHI node) or a statement that is a trivial copy or constant initialization. Attempt to eliminate T by propagating its RHS into all uses of its LHS. This may in turn set new bits in INTERESTING_NAMES for nodes we want to revisit later. All exit paths should clear INTERESTING_NAMES for the result of STMT. */ static void eliminate_const_or_copy (gimple stmt, bitmap interesting_names) { tree lhs = get_lhs_or_phi_result (stmt); tree rhs; int version = SSA_NAME_VERSION (lhs); /* If the LHS of this statement or PHI has no uses, then we can just eliminate it. This can occur if, for example, the PHI was created by block duplication due to threading and its only use was in the conditional at the end of the block which was deleted. */ if (has_zero_uses (lhs)) { bitmap_clear_bit (interesting_names, version); remove_stmt_or_phi (stmt); return; } /* Get the RHS of the assignment or PHI node if the PHI is a degenerate. */ rhs = get_rhs_or_phi_arg (stmt); if (!rhs) { bitmap_clear_bit (interesting_names, version); return; } propagate_rhs_into_lhs (stmt, lhs, rhs, interesting_names); /* Note that STMT may well have been deleted by now, so do not access it, instead use the saved version # to clear T's entry in the worklist. */ bitmap_clear_bit (interesting_names, version); } /* The first phase in degenerate PHI elimination. Eliminate the degenerate PHIs in BB, then recurse on the dominator children of BB. */ static void eliminate_degenerate_phis_1 (basic_block bb, bitmap interesting_names) { gimple_stmt_iterator gsi; basic_block son; for (gsi = gsi_start_phis (bb); !gsi_end_p (gsi); gsi_next (&gsi)) { gimple phi = gsi_stmt (gsi); eliminate_const_or_copy (phi, interesting_names); } /* Recurse into the dominator children of BB. */ for (son = first_dom_son (CDI_DOMINATORS, bb); son; son = next_dom_son (CDI_DOMINATORS, son)) eliminate_degenerate_phis_1 (son, interesting_names); } /* A very simple pass to eliminate degenerate PHI nodes from the IL. This is meant to be fast enough to be able to be run several times in the optimization pipeline. Certain optimizations, particularly those which duplicate blocks or remove edges from the CFG can create or expose PHIs which are trivial copies or constant initializations. While we could pick up these optimizations in DOM or with the combination of copy-prop and CCP, those solutions are far too heavy-weight for our needs. This implementation has two phases so that we can efficiently eliminate the first order degenerate PHIs and second order degenerate PHIs. The first phase performs a dominator walk to identify and eliminate the vast majority of the degenerate PHIs. When a degenerate PHI is identified and eliminated any affected statements or PHIs are put on a worklist. The second phase eliminates degenerate PHIs and trivial copies or constant initializations using the worklist. This is how we pick up the secondary optimization opportunities with minimal cost. */ static unsigned int eliminate_degenerate_phis (void) { bitmap interesting_names; bitmap interesting_names1; /* Bitmap of blocks which need EH information updated. We can not update it on-the-fly as doing so invalidates the dominator tree. */ need_eh_cleanup = BITMAP_ALLOC (NULL); /* INTERESTING_NAMES is effectively our worklist, indexed by SSA_NAME_VERSION. A set bit indicates that the statement or PHI node which defines the SSA_NAME should be (re)examined to determine if it has become a degenerate PHI or trivial const/copy propagation opportunity. Experiments have show we generally get better compilation time behavior with bitmaps rather than sbitmaps. */ interesting_names = BITMAP_ALLOC (NULL); interesting_names1 = BITMAP_ALLOC (NULL); calculate_dominance_info (CDI_DOMINATORS); cfg_altered = false; /* First phase. Eliminate degenerate PHIs via a dominator walk of the CFG. Experiments have indicated that we generally get better compile-time behavior by visiting blocks in the first phase in dominator order. Presumably this is because walking in dominator order leaves fewer PHIs for later examination by the worklist phase. */ eliminate_degenerate_phis_1 (ENTRY_BLOCK_PTR, interesting_names); /* Second phase. Eliminate second order degenerate PHIs as well as trivial copies or constant initializations identified by the first phase or this phase. Basically we keep iterating until our set of INTERESTING_NAMEs is empty. */ while (!bitmap_empty_p (interesting_names)) { unsigned int i; bitmap_iterator bi; /* EXECUTE_IF_SET_IN_BITMAP does not like its bitmap changed during the loop. Copy it to another bitmap and use that. */ bitmap_copy (interesting_names1, interesting_names); EXECUTE_IF_SET_IN_BITMAP (interesting_names1, 0, i, bi) { tree name = ssa_name (i); /* Ignore SSA_NAMEs that have been released because their defining statement was deleted (unreachable). */ if (name) eliminate_const_or_copy (SSA_NAME_DEF_STMT (ssa_name (i)), interesting_names); } } if (cfg_altered) free_dominance_info (CDI_DOMINATORS); /* Propagation of const and copies may make some EH edges dead. Purge such edges from the CFG as needed. */ if (!bitmap_empty_p (need_eh_cleanup)) { gimple_purge_all_dead_eh_edges (need_eh_cleanup); BITMAP_FREE (need_eh_cleanup); } BITMAP_FREE (interesting_names); BITMAP_FREE (interesting_names1); return 0; } struct gimple_opt_pass pass_phi_only_cprop = { { GIMPLE_PASS, "phicprop", /* name */ gate_dominator, /* gate */ eliminate_degenerate_phis, /* execute */ NULL, /* sub */ NULL, /* next */ 0, /* static_pass_number */ TV_TREE_PHI_CPROP, /* tv_id */ PROP_cfg | PROP_ssa, /* properties_required */ 0, /* properties_provided */ 0, /* properties_destroyed */ 0, /* todo_flags_start */ TODO_cleanup_cfg | TODO_dump_func | TODO_ggc_collect | TODO_verify_ssa | TODO_verify_stmts | TODO_update_ssa /* todo_flags_finish */ } };
Go to most recent revision | Compare with Previous | Blame | View Log