/* SSA Jump Threading
- Copyright (C) 2005-2013 Free Software Foundation, Inc.
+ Copyright (C) 2005-2015 Free Software Foundation, Inc.
Contributed by Jeff Law <law@redhat.com>
This file is part of GCC.
#include "system.h"
#include "coretypes.h"
#include "tm.h"
+#include "alias.h"
+#include "symtab.h"
#include "tree.h"
+#include "fold-const.h"
#include "flags.h"
#include "tm_p.h"
+#include "predict.h"
+#include "hard-reg-set.h"
+#include "function.h"
+#include "dominance.h"
#include "basic-block.h"
#include "cfgloop.h"
-#include "function.h"
#include "timevar.h"
#include "dumpfile.h"
-#include "tree-ssa.h"
+#include "tree-ssa-alias.h"
+#include "internal-fn.h"
+#include "gimple-expr.h"
+#include "gimple.h"
+#include "gimple-iterator.h"
+#include "gimple-ssa.h"
+#include "tree-cfg.h"
+#include "tree-phinodes.h"
+#include "ssa-iterators.h"
+#include "stringpool.h"
+#include "tree-ssanames.h"
#include "tree-ssa-propagate.h"
#include "tree-ssa-threadupdate.h"
#include "langhooks.h"
#include "params.h"
+#include "tree-ssa-scopedtables.h"
#include "tree-ssa-threadedge.h"
+#include "tree-ssa-loop.h"
+#include "builtins.h"
+#include "cfg.h"
+#include "cfganal.h"
/* To avoid code explosion due to jump threading, we limit the
number of statements we are going to copy. This variable
{
if (SSA_NAME_VERSION (name) >= ssa_name_values.length ())
ssa_name_values.safe_grow_cleared (SSA_NAME_VERSION (name) + 1);
+ if (value && TREE_OVERFLOW_P (value))
+ value = drop_tree_overflow (value);
ssa_name_values[SSA_NAME_VERSION (name)] = value;
}
{
gimple_stmt_iterator gsi;
+ /* Special case. We can get blocks that are forwarders, but are
+ not optimized away because they forward from outside a loop
+ to the loop header. We want to thread through them as we can
+ sometimes thread to the loop exit, which is obviously profitable.
+ the interesting case here is when the block has PHIs. */
+ if (gsi_end_p (gsi_start_nondebug_bb (bb))
+ && !gsi_end_p (gsi_start_phis (bb)))
+ return true;
+
/* If BB has a single successor or a single predecessor, then
there is no threading opportunity. */
if (single_succ_p (bb) || single_pred_p (bb))
return op;
}
-/* We record temporary equivalences created by PHI nodes or
- statements within the target block. Doing so allows us to
- identify more jump threading opportunities, even in blocks
- with side effects.
-
- We keep track of those temporary equivalences in a stack
- structure so that we can unwind them when we're done processing
- a particular edge. This routine handles unwinding the data
- structures. */
-
-static void
-remove_temporary_equivalences (vec<tree> *stack)
-{
- while (stack->length () > 0)
- {
- tree prev_value, dest;
-
- dest = stack->pop ();
-
- /* A NULL value indicates we should stop unwinding, otherwise
- pop off the next entry as they're recorded in pairs. */
- if (dest == NULL)
- break;
-
- prev_value = stack->pop ();
- set_ssa_name_value (dest, prev_value);
- }
-}
-
-/* Record a temporary equivalence, saving enough information so that
- we can restore the state of recorded equivalences when we're
- done processing the current edge. */
-
-static void
-record_temporary_equivalence (tree x, tree y, vec<tree> *stack)
-{
- tree prev_x = SSA_NAME_VALUE (x);
-
- if (TREE_CODE (y) == SSA_NAME)
- {
- tree tmp = SSA_NAME_VALUE (y);
- y = tmp ? tmp : y;
- }
-
- set_ssa_name_value (x, y);
- stack->reserve (2);
- stack->quick_push (prev_x);
- stack->quick_push (x);
-}
-
/* Record temporary equivalences created by PHIs at the target of the
edge E. Record unwind information for the equivalences onto STACK.
If a PHI which prevents threading is encountered, then return FALSE
- indicating we should not thread this edge, else return TRUE. */
+ indicating we should not thread this edge, else return TRUE.
+
+ If SRC_MAP/DST_MAP exist, then mark the source and destination SSA_NAMEs
+ of any equivalences recorded. We use this to make invalidation after
+ traversing back edges less painful. */
static bool
-record_temporary_equivalences_from_phis (edge e, vec<tree> *stack)
+record_temporary_equivalences_from_phis (edge e, const_and_copies *const_and_copies)
{
- gimple_stmt_iterator gsi;
+ gphi_iterator gsi;
/* Each PHI creates a temporary equivalence, record them.
These are context sensitive equivalences and will be removed
later. */
for (gsi = gsi_start_phis (e->dest); !gsi_end_p (gsi); gsi_next (&gsi))
{
- gimple phi = gsi_stmt (gsi);
+ gphi *phi = gsi.phi ();
tree src = PHI_ARG_DEF_FROM_EDGE (phi, e);
tree dst = gimple_phi_result (phi);
if (!virtual_operand_p (dst))
stmt_count++;
- record_temporary_equivalence (dst, src, stack);
+ const_and_copies->record_const_or_copy (dst, src);
}
return true;
}
static gimple
record_temporary_equivalences_from_stmts_at_dest (edge e,
- vec<tree> *stack,
+ const_and_copies *const_and_copies,
tree (*simplify) (gimple,
- gimple))
+ gimple),
+ bool backedge_seen)
{
gimple stmt = NULL;
gimple_stmt_iterator gsi;
/* If the statement has volatile operands, then we assume we
can not thread through this block. This is overly
conservative in some ways. */
- if (gimple_code (stmt) == GIMPLE_ASM && gimple_asm_volatile_p (stmt))
+ if (gimple_code (stmt) == GIMPLE_ASM
+ && gimple_asm_volatile_p (as_a <gasm *> (stmt)))
return NULL;
/* If duplicating this block is going to cause too much code
&& (gimple_code (stmt) != GIMPLE_CALL
|| gimple_call_lhs (stmt) == NULL_TREE
|| TREE_CODE (gimple_call_lhs (stmt)) != SSA_NAME))
- continue;
+ {
+ /* STMT might still have DEFS and we need to invalidate any known
+ equivalences for them.
+
+ Consider if STMT is a GIMPLE_ASM with one or more outputs that
+ feeds a conditional inside a loop. We might derive an equivalence
+ due to the conditional. */
+ tree op;
+ ssa_op_iter iter;
+
+ if (backedge_seen)
+ FOR_EACH_SSA_TREE_OPERAND (op, stmt, iter, SSA_OP_DEF)
+ const_and_copies->invalidate (op);
+
+ continue;
+ }
/* The result of __builtin_object_size depends on all the arguments
of a phi node. Temporarily using only one edge produces invalid
if (fndecl
&& (DECL_FUNCTION_CODE (fndecl) == BUILT_IN_OBJECT_SIZE
|| DECL_FUNCTION_CODE (fndecl) == BUILT_IN_CONSTANT_P))
- continue;
+ {
+ if (backedge_seen)
+ {
+ tree lhs = gimple_get_lhs (stmt);
+ const_and_copies->invalidate (lhs);
+ }
+ continue;
+ }
}
/* At this point we have a statement which assigns an RHS to an
/* Try to fold/lookup the new expression. Inserting the
expression into the hash table is unlikely to help. */
if (is_gimple_call (stmt))
- cached_lhs = fold_call_stmt (stmt, false);
+ cached_lhs = fold_call_stmt (as_a <gcall *> (stmt), false);
else
cached_lhs = fold_assignment_stmt (stmt);
}
/* Record the context sensitive equivalence if we were able
- to simplify this statement. */
+ to simplify this statement.
+
+ If we have traversed a backedge at some point during threading,
+ then always enter something here. Either a real equivalence,
+ or a NULL_TREE equivalence which is effectively invalidation of
+ prior equivalences. */
if (cached_lhs
&& (TREE_CODE (cached_lhs) == SSA_NAME
|| is_gimple_min_invariant (cached_lhs)))
- record_temporary_equivalence (gimple_get_lhs (stmt), cached_lhs, stack);
+ const_and_copies->record_const_or_copy (gimple_get_lhs (stmt), cached_lhs);
+ else if (backedge_seen)
+ const_and_copies->invalidate (gimple_get_lhs (stmt));
}
return stmt;
}
+/* Once we have passed a backedge in the CFG when threading, we do not want to
+ utilize edge equivalences for simplification purpose. They are no longer
+ necessarily valid. We use this callback rather than the ones provided by
+ DOM/VRP to achieve that effect. */
+static tree
+dummy_simplify (gimple stmt1 ATTRIBUTE_UNUSED, gimple stmt2 ATTRIBUTE_UNUSED)
+{
+ return NULL_TREE;
+}
+
/* Simplify the control statement at the end of the block E->dest.
To avoid allocating memory unnecessarily, a scratch GIMPLE_COND
static tree
simplify_control_stmt_condition (edge e,
gimple stmt,
- gimple dummy_cond,
+ gcond *dummy_cond,
tree (*simplify) (gimple, gimple),
bool handle_dominating_asserts)
{
/* Get the current value of both operands. */
if (TREE_CODE (op0) == SSA_NAME)
{
- tree tmp = SSA_NAME_VALUE (op0);
- if (tmp)
- op0 = tmp;
+ for (int i = 0; i < 2; i++)
+ {
+ if (TREE_CODE (op0) == SSA_NAME
+ && SSA_NAME_VALUE (op0))
+ op0 = SSA_NAME_VALUE (op0);
+ else
+ break;
+ }
}
if (TREE_CODE (op1) == SSA_NAME)
{
- tree tmp = SSA_NAME_VALUE (op1);
- if (tmp)
- op1 = tmp;
+ for (int i = 0; i < 2; i++)
+ {
+ if (TREE_CODE (op1) == SSA_NAME
+ && SSA_NAME_VALUE (op1))
+ op1 = SSA_NAME_VALUE (op1);
+ else
+ break;
+ }
}
if (handle_dominating_asserts)
miss threading opportunities. */
if (tree_swap_operands_p (op0, op1, false))
{
- tree tmp;
cond_code = swap_tree_comparison (cond_code);
- tmp = op0;
- op0 = op1;
- op1 = tmp;
+ std::swap (op0, op1);
}
/* Stuff the operator and operands into our dummy conditional
}
if (code == GIMPLE_SWITCH)
- cond = gimple_switch_index (stmt);
+ cond = gimple_switch_index (as_a <gswitch *> (stmt));
else if (code == GIMPLE_GOTO)
cond = gimple_goto_dest (stmt);
else
rather than use a relational operator. These are simpler to handle. */
if (TREE_CODE (cond) == SSA_NAME)
{
+ tree original_lhs = cond;
cached_lhs = cond;
/* Get the variable's current value from the equivalence chains.
It is possible to get loops in the SSA_NAME_VALUE chains
(consider threading the backedge of a loop where we have
a loop invariant SSA_NAME used in the condition. */
- if (cached_lhs
- && TREE_CODE (cached_lhs) == SSA_NAME
- && SSA_NAME_VALUE (cached_lhs))
- cached_lhs = SSA_NAME_VALUE (cached_lhs);
+ if (cached_lhs)
+ {
+ for (int i = 0; i < 2; i++)
+ {
+ if (TREE_CODE (cached_lhs) == SSA_NAME
+ && SSA_NAME_VALUE (cached_lhs))
+ cached_lhs = SSA_NAME_VALUE (cached_lhs);
+ else
+ break;
+ }
+ }
/* If we're dominated by a suitable ASSERT_EXPR, then
update CACHED_LHS appropriately. */
pass specific callback to try and simplify it further. */
if (cached_lhs && ! is_gimple_min_invariant (cached_lhs))
cached_lhs = (*simplify) (stmt, stmt);
+
+ /* We couldn't find an invariant. But, callers of this
+ function may be able to do something useful with the
+ unmodified destination. */
+ if (!cached_lhs)
+ cached_lhs = original_lhs;
}
else
cached_lhs = NULL;
return cached_lhs;
}
-/* Return TRUE if the statement at the end of e->dest depends on
- the output of any statement in BB. Otherwise return FALSE.
-
- This is used when we are threading a backedge and need to ensure
- that temporary equivalences from BB do not affect the condition
- in e->dest. */
-
-static bool
-cond_arg_set_in_bb (edge e, basic_block bb)
-{
- ssa_op_iter iter;
- use_operand_p use_p;
- gimple last = last_stmt (e->dest);
-
- /* E->dest does not have to end with a control transferring
- instruction. This can occur when we try to extend a jump
- threading opportunity deeper into the CFG. In that case
- it is safe for this check to return false. */
- if (!last)
- return false;
-
- if (gimple_code (last) != GIMPLE_COND
- && gimple_code (last) != GIMPLE_GOTO
- && gimple_code (last) != GIMPLE_SWITCH)
- return false;
-
- FOR_EACH_SSA_USE_OPERAND (use_p, last, iter, SSA_OP_USE | SSA_OP_VUSE)
- {
- tree use = USE_FROM_PTR (use_p);
-
- if (TREE_CODE (use) == SSA_NAME
- && gimple_code (SSA_NAME_DEF_STMT (use)) != GIMPLE_PHI
- && gimple_bb (SSA_NAME_DEF_STMT (use)) == bb)
- return true;
- }
- return false;
-}
-
/* Copy debug stmts from DEST's chain of single predecessors up to
SRC, so that we don't lose the bindings as PHI nodes are introduced
when DEST gains new predecessors. */
i++;
}
- vec<tree, va_stack> fewvars = vNULL;
- pointer_set_t *vars = NULL;
+ auto_vec<tree, alloc_count> fewvars;
+ hash_set<tree> *vars = NULL;
/* If we're already starting with 3/4 of alloc_count, go for a
- pointer_set, otherwise start with an unordered stack-allocated
+ hash_set, otherwise start with an unordered stack-allocated
VEC. */
if (i * 4 > alloc_count * 3)
- vars = pointer_set_create ();
- else if (alloc_count)
- vec_stack_alloc (tree, fewvars, alloc_count);
+ vars = new hash_set<tree>;
/* Now go through the initial debug stmts in DEST again, this time
actually inserting in VARS or FEWVARS. Don't bother checking for
gcc_unreachable ();
if (vars)
- pointer_set_insert (vars, var);
+ vars->add (var);
else
fewvars.quick_push (var);
}
or somesuch. Adding `&& bb == src' to the condition
below will preserve all potentially relevant debug
notes. */
- if (vars && pointer_set_insert (vars, var))
+ if (vars && vars->add (var))
continue;
else if (!vars)
{
fewvars.quick_push (var);
else
{
- vars = pointer_set_create ();
+ vars = new hash_set<tree>;
for (i = 0; i < alloc_count; i++)
- pointer_set_insert (vars, fewvars[i]);
+ vars->add (fewvars[i]);
fewvars.release ();
- pointer_set_insert (vars, var);
+ vars->add (var);
}
}
while (bb != src && single_pred_p (bb));
if (vars)
- pointer_set_destroy (vars);
+ delete vars;
else if (fewvars.exists ())
fewvars.release ();
}
try and simplify the condition at the end of TAKEN_EDGE->dest. */
static bool
thread_around_empty_blocks (edge taken_edge,
- gimple dummy_cond,
+ gcond *dummy_cond,
bool handle_dominating_asserts,
tree (*simplify) (gimple, gimple),
bitmap visited,
- vec<jump_thread_edge *> *path)
+ vec<jump_thread_edge *> *path,
+ bool *backedge_seen_p)
{
basic_block bb = taken_edge->dest;
gimple_stmt_iterator gsi;
if (single_succ_p (bb))
{
taken_edge = single_succ_edge (bb);
- if ((taken_edge->flags & EDGE_DFS_BACK) == 0
- && !bitmap_bit_p (visited, taken_edge->dest->index))
+ if (!bitmap_bit_p (visited, taken_edge->dest->index))
{
jump_thread_edge *x
= new jump_thread_edge (taken_edge, EDGE_NO_COPY_SRC_BLOCK);
path->safe_push (x);
bitmap_set_bit (visited, taken_edge->dest->index);
+ *backedge_seen_p |= ((taken_edge->flags & EDGE_DFS_BACK) != 0);
+ if (*backedge_seen_p)
+ simplify = dummy_simplify;
return thread_around_empty_blocks (taken_edge,
dummy_cond,
handle_dominating_asserts,
simplify,
visited,
- path);
+ path,
+ backedge_seen_p);
}
}
&& gimple_code (stmt) != GIMPLE_SWITCH)
return false;
+ /* If we have traversed a backedge, then we do not want to look
+ at certain expressions in the table that can not be relied upon.
+ Luckily the only code that looked at those expressions is the
+ SIMPLIFY callback, which we replace if we can no longer use it. */
+ if (*backedge_seen_p)
+ simplify = dummy_simplify;
+
/* Extract and simplify the condition. */
cond = simplify_control_stmt_condition (taken_edge, stmt, dummy_cond,
simplify, handle_dominating_asserts);
jump_thread_edge *x
= new jump_thread_edge (taken_edge, EDGE_NO_COPY_SRC_BLOCK);
path->safe_push (x);
+ *backedge_seen_p |= ((taken_edge->flags & EDGE_DFS_BACK) != 0);
+ if (*backedge_seen_p)
+ simplify = dummy_simplify;
thread_around_empty_blocks (taken_edge,
dummy_cond,
handle_dominating_asserts,
simplify,
visited,
- path);
+ path,
+ backedge_seen_p);
+ return true;
+ }
+
+ return false;
+}
+
+/* Return true if the CFG contains at least one path from START_BB to END_BB.
+ When a path is found, record in PATH the blocks from END_BB to START_BB.
+ VISITED_BBS is used to make sure we don't fall into an infinite loop. Bound
+ the recursion to basic blocks belonging to LOOP. */
+
+static bool
+fsm_find_thread_path (basic_block start_bb, basic_block end_bb,
+ vec<basic_block, va_gc> *&path,
+ hash_set<basic_block> *visited_bbs, loop_p loop)
+{
+ if (loop != start_bb->loop_father)
+ return false;
+
+ if (start_bb == end_bb)
+ {
+ vec_safe_push (path, start_bb);
return true;
}
+ if (!visited_bbs->add (start_bb))
+ {
+ edge e;
+ edge_iterator ei;
+ FOR_EACH_EDGE (e, ei, start_bb->succs)
+ if (fsm_find_thread_path (e->dest, end_bb, path, visited_bbs, loop))
+ {
+ vec_safe_push (path, start_bb);
+ return true;
+ }
+ }
+
return false;
}
+static int max_threaded_paths;
+
+/* We trace the value of the variable EXPR back through any phi nodes looking
+ for places where it gets a constant value and save the path. Stop after
+ having recorded MAX_PATHS jump threading paths. */
+
+static void
+fsm_find_control_statement_thread_paths (tree expr,
+ hash_set<basic_block> *visited_bbs,
+ vec<basic_block, va_gc> *&path,
+ bool seen_loop_phi)
+{
+ tree var = SSA_NAME_VAR (expr);
+ gimple def_stmt = SSA_NAME_DEF_STMT (expr);
+ basic_block var_bb = gimple_bb (def_stmt);
+
+ if (var == NULL || var_bb == NULL)
+ return;
+
+ /* For the moment we assume that an SSA chain only contains phi nodes, and
+ eventually one of the phi arguments will be an integer constant. In the
+ future, this could be extended to also handle simple assignments of
+ arithmetic operations. */
+ if (gimple_code (def_stmt) != GIMPLE_PHI)
+ return;
+
+ /* Avoid infinite recursion. */
+ if (visited_bbs->add (var_bb))
+ return;
+
+ gphi *phi = as_a <gphi *> (def_stmt);
+ int next_path_length = 0;
+ basic_block last_bb_in_path = path->last ();
+
+ if (loop_containing_stmt (phi)->header == gimple_bb (phi))
+ {
+ /* Do not walk through more than one loop PHI node. */
+ if (seen_loop_phi)
+ return;
+ seen_loop_phi = true;
+ }
+
+ /* Following the chain of SSA_NAME definitions, we jumped from a definition in
+ LAST_BB_IN_PATH to a definition in VAR_BB. When these basic blocks are
+ different, append to PATH the blocks from LAST_BB_IN_PATH to VAR_BB. */
+ if (var_bb != last_bb_in_path)
+ {
+ edge e;
+ int e_count = 0;
+ edge_iterator ei;
+ vec<basic_block, va_gc> *next_path;
+ vec_alloc (next_path, n_basic_blocks_for_fn (cfun));
+
+ FOR_EACH_EDGE (e, ei, last_bb_in_path->preds)
+ {
+ hash_set<basic_block> *visited_bbs = new hash_set<basic_block>;
+
+ if (fsm_find_thread_path (var_bb, e->src, next_path, visited_bbs,
+ e->src->loop_father))
+ ++e_count;
+
+ delete visited_bbs;
+
+ /* If there is more than one path, stop. */
+ if (e_count > 1)
+ {
+ vec_free (next_path);
+ return;
+ }
+ }
+
+ /* Stop if we have not found a path: this could occur when the recursion
+ is stopped by one of the bounds. */
+ if (e_count == 0)
+ {
+ vec_free (next_path);
+ return;
+ }
+
+ /* Append all the nodes from NEXT_PATH to PATH. */
+ vec_safe_splice (path, next_path);
+ next_path_length = next_path->length ();
+ vec_free (next_path);
+ }
+
+ gcc_assert (path->last () == var_bb);
+
+ /* Iterate over the arguments of PHI. */
+ unsigned int i;
+ for (i = 0; i < gimple_phi_num_args (phi); i++)
+ {
+ tree arg = gimple_phi_arg_def (phi, i);
+ basic_block bbi = gimple_phi_arg_edge (phi, i)->src;
+
+ /* Skip edges pointing outside the current loop. */
+ if (!arg || var_bb->loop_father != bbi->loop_father)
+ continue;
+
+ if (TREE_CODE (arg) == SSA_NAME)
+ {
+ vec_safe_push (path, bbi);
+ /* Recursively follow SSA_NAMEs looking for a constant definition. */
+ fsm_find_control_statement_thread_paths (arg, visited_bbs, path,
+ seen_loop_phi);
+
+ path->pop ();
+ continue;
+ }
+
+ if (TREE_CODE (arg) != INTEGER_CST)
+ continue;
+
+ int path_length = path->length ();
+ /* A path with less than 2 basic blocks should not be jump-threaded. */
+ if (path_length < 2)
+ continue;
+
+ if (path_length > PARAM_VALUE (PARAM_MAX_FSM_THREAD_LENGTH))
+ {
+ if (dump_file && (dump_flags & TDF_DETAILS))
+ fprintf (dump_file, "FSM jump-thread path not considered: "
+ "the number of basic blocks on the path "
+ "exceeds PARAM_MAX_FSM_THREAD_LENGTH.\n");
+ continue;
+ }
+
+ if (max_threaded_paths <= 0)
+ {
+ if (dump_file && (dump_flags & TDF_DETAILS))
+ fprintf (dump_file, "FSM jump-thread path not considered: "
+ "the number of previously recorded FSM paths to thread "
+ "exceeds PARAM_MAX_FSM_THREAD_PATHS.\n");
+ continue;
+ }
+
+ /* Add BBI to the path. */
+ vec_safe_push (path, bbi);
+ ++path_length;
+
+ int n_insns = 0;
+ gimple_stmt_iterator gsi;
+ int j;
+ loop_p loop = (*path)[0]->loop_father;
+ bool path_crosses_loops = false;
+
+ /* Count the number of instructions on the path: as these instructions
+ will have to be duplicated, we will not record the path if there are
+ too many instructions on the path. Also check that all the blocks in
+ the path belong to a single loop. */
+ for (j = 1; j < path_length - 1; j++)
+ {
+ basic_block bb = (*path)[j];
+
+ if (bb->loop_father != loop)
+ {
+ path_crosses_loops = true;
+ break;
+ }
+
+ for (gsi = gsi_start_bb (bb); !gsi_end_p (gsi); gsi_next (&gsi))
+ {
+ gimple stmt = gsi_stmt (gsi);
+ /* Do not count empty statements and labels. */
+ if (gimple_code (stmt) != GIMPLE_NOP
+ && gimple_code (stmt) != GIMPLE_LABEL
+ && !is_gimple_debug (stmt))
+ ++n_insns;
+ }
+ }
+
+ if (path_crosses_loops)
+ {
+ if (dump_file && (dump_flags & TDF_DETAILS))
+ fprintf (dump_file, "FSM jump-thread path not considered: "
+ "the path crosses loops.\n");
+ path->pop ();
+ continue;
+ }
+
+ if (n_insns >= PARAM_VALUE (PARAM_MAX_FSM_THREAD_PATH_INSNS))
+ {
+ if (dump_file && (dump_flags & TDF_DETAILS))
+ fprintf (dump_file, "FSM jump-thread path not considered: "
+ "the number of instructions on the path "
+ "exceeds PARAM_MAX_FSM_THREAD_PATH_INSNS.\n");
+ path->pop ();
+ continue;
+ }
+
+ vec<jump_thread_edge *> *jump_thread_path
+ = new vec<jump_thread_edge *> ();
+
+ /* Record the edges between the blocks in PATH. */
+ for (j = 0; j < path_length - 1; j++)
+ {
+ edge e = find_edge ((*path)[path_length - j - 1],
+ (*path)[path_length - j - 2]);
+ gcc_assert (e);
+ jump_thread_edge *x = new jump_thread_edge (e, EDGE_FSM_THREAD);
+ jump_thread_path->safe_push (x);
+ }
+
+ /* Add the edge taken when the control variable has value ARG. */
+ edge taken_edge = find_taken_edge ((*path)[0], arg);
+ jump_thread_edge *x
+ = new jump_thread_edge (taken_edge, EDGE_NO_COPY_SRC_BLOCK);
+ jump_thread_path->safe_push (x);
+
+ register_jump_thread (jump_thread_path);
+ --max_threaded_paths;
+
+ /* Remove BBI from the path. */
+ path->pop ();
+ }
+
+ /* Remove all the nodes that we added from NEXT_PATH. */
+ if (next_path_length)
+ vec_safe_truncate (path, (path->length () - next_path_length));
+}
+
/* We are exiting E->src, see if E->dest ends with a conditional
jump which has a known value when reached via E.
SIMPLIFY is a pass-specific function used to simplify statements.
Our caller is responsible for restoring the state of the expression
- and const_and_copies stacks. */
+ and const_and_copies stacks.
-static bool
+ Positive return value is success. Zero return value is failure, but
+ the block can still be duplicated as a joiner in a jump thread path,
+ negative indicates the block should not be duplicated and thus is not
+ suitable for a joiner in a jump threading path. */
+
+static int
thread_through_normal_block (edge e,
- gimple dummy_cond,
+ gcond *dummy_cond,
bool handle_dominating_asserts,
- vec<tree> *stack,
+ const_and_copies *const_and_copies,
tree (*simplify) (gimple, gimple),
vec<jump_thread_edge *> *path,
- bitmap visited)
+ bitmap visited,
+ bool *backedge_seen_p)
{
- /* If E is a backedge, then we want to verify that the COND_EXPR,
- SWITCH_EXPR or GOTO_EXPR at the end of e->dest is not affected
- by any statements in e->dest. If it is affected, then it is not
- safe to thread this edge. */
- if (e->flags & EDGE_DFS_BACK)
- {
- if (cond_arg_set_in_bb (e, e->dest))
- return false;
- }
-
- /* PHIs create temporary equivalences. */
- if (!record_temporary_equivalences_from_phis (e, stack))
- return false;
+ /* If we have traversed a backedge, then we do not want to look
+ at certain expressions in the table that can not be relied upon.
+ Luckily the only code that looked at those expressions is the
+ SIMPLIFY callback, which we replace if we can no longer use it. */
+ if (*backedge_seen_p)
+ simplify = dummy_simplify;
+
+ /* PHIs create temporary equivalences.
+ Note that if we found a PHI that made the block non-threadable, then
+ we need to bubble that up to our caller in the same manner we do
+ when we prematurely stop processing statements below. */
+ if (!record_temporary_equivalences_from_phis (e, const_and_copies))
+ return -1;
/* Now walk each statement recording any context sensitive
temporary equivalences we can detect. */
gimple stmt
- = record_temporary_equivalences_from_stmts_at_dest (e, stack, simplify);
+ = record_temporary_equivalences_from_stmts_at_dest (e, const_and_copies, simplify,
+ *backedge_seen_p);
+
+ /* There's two reasons STMT might be null, and distinguishing
+ between them is important.
+
+ First the block may not have had any statements. For example, it
+ might have some PHIs and unconditionally transfer control elsewhere.
+ Such blocks are suitable for jump threading, particularly as a
+ joiner block.
+
+ The second reason would be if we did not process all the statements
+ in the block (because there were too many to make duplicating the
+ block profitable. If we did not look at all the statements, then
+ we may not have invalidated everything needing invalidation. Thus
+ we must signal to our caller that this block is not suitable for
+ use as a joiner in a threading path. */
if (!stmt)
- return false;
-
+ {
+ /* First case. The statement simply doesn't have any instructions, but
+ does have PHIs. */
+ if (gsi_end_p (gsi_start_nondebug_bb (e->dest))
+ && !gsi_end_p (gsi_start_phis (e->dest)))
+ return 0;
+
+ /* Second case. */
+ return -1;
+ }
+
/* If we stopped at a COND_EXPR or SWITCH_EXPR, see if we know which arm
will be taken. */
if (gimple_code (stmt) == GIMPLE_COND
cond = simplify_control_stmt_condition (e, stmt, dummy_cond, simplify,
handle_dominating_asserts);
- if (cond && is_gimple_min_invariant (cond))
+ if (!cond)
+ return 0;
+
+ if (is_gimple_min_invariant (cond))
{
edge taken_edge = find_taken_edge (e->dest, cond);
basic_block dest = (taken_edge ? taken_edge->dest : NULL);
/* DEST could be NULL for a computed jump to an absolute
address. */
- if (dest == NULL || dest == e->dest || bitmap_bit_p (visited, dest->index))
- return false;
-
- jump_thread_edge *x
- = new jump_thread_edge (e, EDGE_START_JUMP_THREAD);
- path->safe_push (x);
+ if (dest == NULL
+ || dest == e->dest
+ || bitmap_bit_p (visited, dest->index))
+ return 0;
+
+ /* Only push the EDGE_START_JUMP_THREAD marker if this is
+ first edge on the path. */
+ if (path->length () == 0)
+ {
+ jump_thread_edge *x
+ = new jump_thread_edge (e, EDGE_START_JUMP_THREAD);
+ path->safe_push (x);
+ *backedge_seen_p |= ((e->flags & EDGE_DFS_BACK) != 0);
+ }
- x = new jump_thread_edge (taken_edge, EDGE_COPY_SRC_BLOCK);
+ jump_thread_edge *x
+ = new jump_thread_edge (taken_edge, EDGE_COPY_SRC_BLOCK);
path->safe_push (x);
+ *backedge_seen_p |= ((taken_edge->flags & EDGE_DFS_BACK) != 0);
+ if (*backedge_seen_p)
+ simplify = dummy_simplify;
/* See if we can thread through DEST as well, this helps capture
secondary effects of threading without having to re-run DOM or
- VRP. */
- if ((e->flags & EDGE_DFS_BACK) == 0
- || ! cond_arg_set_in_bb (taken_edge, e->dest))
- {
- /* We don't want to thread back to a block we have already
- visited. This may be overly conservative. */
- bitmap_set_bit (visited, dest->index);
- bitmap_set_bit (visited, e->dest->index);
- thread_around_empty_blocks (taken_edge,
- dummy_cond,
- handle_dominating_asserts,
- simplify,
- visited,
- path);
- }
- return true;
+ VRP.
+
+ We don't want to thread back to a block we have already
+ visited. This may be overly conservative. */
+ bitmap_set_bit (visited, dest->index);
+ bitmap_set_bit (visited, e->dest->index);
+ thread_around_empty_blocks (taken_edge,
+ dummy_cond,
+ handle_dominating_asserts,
+ simplify,
+ visited,
+ path,
+ backedge_seen_p);
+ return 1;
}
+
+ if (!flag_expensive_optimizations
+ || optimize_function_for_size_p (cfun)
+ || TREE_CODE (cond) != SSA_NAME
+ || e->dest->loop_father != e->src->loop_father
+ || loop_depth (e->dest->loop_father) == 0)
+ return 0;
+
+ /* When COND cannot be simplified, try to find paths from a control
+ statement back through the PHI nodes which would affect that control
+ statement. */
+ vec<basic_block, va_gc> *bb_path;
+ vec_alloc (bb_path, n_basic_blocks_for_fn (cfun));
+ vec_safe_push (bb_path, e->dest);
+ hash_set<basic_block> *visited_bbs = new hash_set<basic_block>;
+
+ max_threaded_paths = PARAM_VALUE (PARAM_MAX_FSM_THREAD_PATHS);
+ fsm_find_control_statement_thread_paths (cond, visited_bbs, bb_path,
+ false);
+
+ delete visited_bbs;
+ vec_free (bb_path);
}
- return false;
+ return 0;
}
/* We are exiting E->src, see if E->dest ends with a conditional
SIMPLIFY is a pass-specific function used to simplify statements. */
void
-thread_across_edge (gimple dummy_cond,
+thread_across_edge (gcond *dummy_cond,
edge e,
bool handle_dominating_asserts,
- vec<tree> *stack,
+ const_and_copies *const_and_copies,
tree (*simplify) (gimple, gimple))
{
bitmap visited = BITMAP_ALLOC (NULL);
+ bool backedge_seen;
stmt_count = 0;
bitmap_clear (visited);
bitmap_set_bit (visited, e->src->index);
bitmap_set_bit (visited, e->dest->index);
- if (thread_through_normal_block (e, dummy_cond, handle_dominating_asserts,
- stack, simplify, path, visited))
+ backedge_seen = ((e->flags & EDGE_DFS_BACK) != 0);
+ if (backedge_seen)
+ simplify = dummy_simplify;
+
+ int threaded = thread_through_normal_block (e, dummy_cond,
+ handle_dominating_asserts,
+ const_and_copies, simplify, path,
+ visited, &backedge_seen);
+ if (threaded > 0)
{
propagate_threaded_block_debug_into (path->last ()->e->dest,
e->dest);
- remove_temporary_equivalences (stack);
+ const_and_copies->pop_to_marker ();
BITMAP_FREE (visited);
register_jump_thread (path);
return;
}
else
{
- /* There should be no edges on the path, so no need to walk through
- the vector entries. */
+ /* Negative and zero return values indicate no threading was possible,
+ thus there should be no edges on the thread path and no need to walk
+ through the vector entries. */
gcc_assert (path->length () == 0);
path->release ();
+ delete path;
+
+ /* A negative status indicates the target block was deemed too big to
+ duplicate. Just quit now rather than trying to use the block as
+ a joiner in a jump threading path.
+
+ This prevents unnecessary code growth, but more importantly if we
+ do not look at all the statements in the block, then we may have
+ missed some invalidations if we had traversed a backedge! */
+ if (threaded < 0)
+ {
+ BITMAP_FREE (visited);
+ const_and_copies->pop_to_marker ();
+ return;
+ }
}
/* We were unable to determine what out edge from E->dest is taken. However,
edge_iterator ei;
bool found;
+ /* If E->dest has abnormal outgoing edges, then there's no guarantee
+ we can safely redirect any of the edges. Just punt those cases. */
+ FOR_EACH_EDGE (taken_edge, ei, e->dest->succs)
+ if (taken_edge->flags & EDGE_ABNORMAL)
+ {
+ const_and_copies->pop_to_marker ();
+ BITMAP_FREE (visited);
+ return;
+ }
+
/* Look at each successor of E->dest to see if we can thread through it. */
FOR_EACH_EDGE (taken_edge, ei, e->dest->succs)
{
+ /* Push a fresh marker so we can unwind the equivalences created
+ for each of E->dest's successors. */
+ const_and_copies->push_marker ();
+
/* Avoid threading to any block we have already visited. */
bitmap_clear (visited);
- bitmap_set_bit (visited, taken_edge->dest->index);
+ bitmap_set_bit (visited, e->src->index);
bitmap_set_bit (visited, e->dest->index);
+ bitmap_set_bit (visited, taken_edge->dest->index);
vec<jump_thread_edge *> *path = new vec<jump_thread_edge *> ();
/* Record whether or not we were able to thread through a successor
x = new jump_thread_edge (taken_edge, EDGE_COPY_SRC_JOINER_BLOCK);
path->safe_push (x);
found = false;
- if ((e->flags & EDGE_DFS_BACK) == 0
- || ! cond_arg_set_in_bb (path->last ()->e, e->dest))
- found = thread_around_empty_blocks (taken_edge,
- dummy_cond,
- handle_dominating_asserts,
- simplify,
- visited,
- path);
+ backedge_seen = ((e->flags & EDGE_DFS_BACK) != 0);
+ backedge_seen |= ((taken_edge->flags & EDGE_DFS_BACK) != 0);
+ if (backedge_seen)
+ simplify = dummy_simplify;
+ found = thread_around_empty_blocks (taken_edge,
+ dummy_cond,
+ handle_dominating_asserts,
+ simplify,
+ visited,
+ path,
+ &backedge_seen);
+
+ if (backedge_seen)
+ simplify = dummy_simplify;
+
+ if (!found)
+ found = thread_through_normal_block (path->last ()->e, dummy_cond,
+ handle_dominating_asserts,
+ const_and_copies, simplify, path, visited,
+ &backedge_seen) > 0;
/* If we were able to thread through a successor of E->dest, then
record the jump threading opportunity. */
}
else
{
- for (unsigned int i = 0; i < path->length (); i++)
- delete (*path)[i];
- path->release();
+ delete_jump_thread_path (path);
}
+
+ /* And unwind the equivalence table. */
+ const_and_copies->pop_to_marker ();
}
BITMAP_FREE (visited);
}
- remove_temporary_equivalences (stack);
+ const_and_copies->pop_to_marker ();
}