/**
* Scheduling algorithms.
* Just a simple list scheduling algorithm is here.
- * @date 20.10.2004
+ * @date 20.10.2004
* @author Sebastian Hack
+ * @cvs-id $Id$
*/
#ifdef HAVE_CONFIG_H
#include <string.h>
#include <limits.h>
+#include "benode_t.h"
+#include "be_t.h"
+
#include "obst.h"
#include "list.h"
#include "iterator.h"
#include "irmode_t.h"
#include "irdump.h"
#include "irprintf_t.h"
+#include "array.h"
#include "debug.h"
+#include "irtools.h"
#include "besched_t.h"
#include "beutil.h"
#include "belive_t.h"
#include "belistsched.h"
+#include "beschedmris.h"
#include "bearch.h"
+#include "bestat.h"
-#define MAX(x,y) ((x) > (y) ? (x) : (y))
-#define MIN(x,y) ((x) < (y) ? (x) : (y))
+/**
+ * All scheduling info needed per node.
+ */
+typedef struct _sched_irn_t {
+ sched_timestep_t delay; /**< The delay for this node if already calculated, else 0. */
+ sched_timestep_t etime; /**< The earliest time of this node. */
+ unsigned num_user; /**< The number real users (mode datab) of this node */
+ unsigned num_not_sched_user; /**< The number of not yet scheduled users of this node */
+ int reg_diff; /**< The difference of num(out registers) - num(in registers) */
+ int preorder; /**< The pre-order position */
+ unsigned critical_path_len; /**< The weighted length of the longest critical path */
+ unsigned already_sched : 1; /**< Set if this node is already scheduled */
+ unsigned is_root : 1; /**< is a root node of a block */
+} sched_irn_t;
/**
* Scheduling environment for the whole graph.
*/
typedef struct _sched_env_t {
- const ir_graph *irg; /**< The graph to schedule. */
- const list_sched_selector_t *selector; /**< The node selector. */
- void *selector_env; /**< A pointer to give to the selector. */
+ sched_irn_t *sched_info; /**< scheduling info per node */
+ const list_sched_selector_t *selector; /**< The node selector. */
+ const arch_env_t *arch_env; /**< The architecture environment. */
+ const ir_graph *irg; /**< The graph to schedule. */
+ void *selector_env; /**< A pointer to give to the selector. */
+ int sel_strategy; /**< Node selection strategy (muchnik, mueller, isa) */
} sched_env_t;
#if 0
}
#endif
-static ir_node *trivial_select(void *block_env, pset *ready_set)
+/**
+ * The trivial selector:
+ * Just assure that branches are executed last, otherwise select
+ * the first node ready.
+ */
+static ir_node *trivial_select(void *block_env, nodeset *ready_set)
{
- ir_node *res;
+ const arch_env_t *arch_env = block_env;
+ ir_node *irn = NULL;
+
+ /* assure that branches and constants are executed last */
+ for (irn = nodeset_first(ready_set); irn; irn = nodeset_next(ready_set)) {
+ if (! arch_irn_class_is(arch_env, irn, branch)) {
+ nodeset_break(ready_set);
+ return irn;
+ }
+ }
- res = pset_first(ready_set);
- pset_break(ready_set);
- return res;
+
+ /* at last: schedule branches */
+ irn = nodeset_first(ready_set);
+ nodeset_break(ready_set);
+
+ return irn;
}
-static int default_to_appear_in_schedule(void *env, const ir_node *irn)
+static void *trivial_init_graph(const list_sched_selector_t *vtab, const arch_env_t *arch_env, ir_graph *irg)
{
- return to_appear_in_schedule(irn);
+ return (void *) arch_env;
+}
+
+static void *trivial_init_block(void *graph_env, ir_node *bl)
+{
+ return graph_env;
+}
+
+static INLINE int must_appear_in_schedule(const list_sched_selector_t *sel, void *block_env, const ir_node *irn)
+{
+ int res = -1;
+
+ if(sel->to_appear_in_schedule)
+ res = sel->to_appear_in_schedule(block_env, irn);
+
+ return res >= 0 ? res : (to_appear_in_schedule(irn) || be_is_Keep(irn) || be_is_RegParams(irn));
}
static const list_sched_selector_t trivial_selector_struct = {
- NULL,
- NULL,
+ trivial_init_graph,
+ trivial_init_block,
trivial_select,
- default_to_appear_in_schedule,
- NULL,
- NULL
+ NULL, /* to_appear_in_schedule */
+ NULL, /* exectime */
+ NULL, /* latency */
+ NULL, /* finish_block */
+ NULL /* finish_graph */
};
const list_sched_selector_t *trivial_selector = &trivial_selector_struct;
scheduled. */
} usage_stats_t;
+typedef struct {
+ const list_sched_selector_t *vtab;
+ const arch_env_t *arch_env;
+} reg_pressure_main_env_t;
+
typedef struct {
struct obstack obst;
+ const reg_pressure_main_env_t *main_env;
usage_stats_t *root;
- pset *already_scheduled;
- const list_sched_selector_t *vtab;
+ nodeset *already_scheduled;
} reg_pressure_selector_env_t;
static INLINE usage_stats_t *get_or_set_usage_stats(reg_pressure_selector_env_t *env, ir_node *irn)
return us;
}
-static int max_hops_walker(ir_node *irn, ir_node *tgt, int depth, unsigned visited_nr)
+static int max_hops_walker(reg_pressure_selector_env_t *env, ir_node *irn, ir_node *curr_bl, int depth, unsigned visited_nr)
{
- int i, n;
- int res = 0;
-
- if(irn != tgt) {
- res = INT_MAX;
+ ir_node *bl = get_nodes_block(irn);
+ /*
+ * If the reached node is not in the block desired,
+ * return the value passed for this situation.
+ */
+ if(get_nodes_block(irn) != bl)
+ return block_dominates(bl, curr_bl) ? 0 : INT_MAX;
+ /*
+ * If the node is in the current block but not
+ * yet scheduled, we keep on searching from that node.
+ */
+ if(!nodeset_find(env->already_scheduled, irn)) {
+ int i, n;
+ int res = 0;
for(i = 0, n = get_irn_arity(irn); i < n; ++i) {
ir_node *operand = get_irn_n(irn, i);
int tmp;
set_irn_visited(operand, visited_nr);
- tmp = max_hops_walker(operand, tgt, depth + 1, visited_nr);
+ tmp = max_hops_walker(env, operand, bl, depth + 1, visited_nr);
res = MAX(tmp, res);
}
}
+
+ return res;
}
- return res;
+ /*
+ * If the node is in the current block and scheduled, return
+ * the depth which indicates the number of steps to the
+ * region of scheduled nodes.
+ */
+ return depth;
}
static int compute_max_hops(reg_pressure_selector_env_t *env, ir_node *irn)
{
ir_node *bl = get_nodes_block(irn);
ir_graph *irg = get_irn_irg(bl);
- int res = INT_MAX;
+ int res = 0;
const ir_edge_t *edge;
foreach_out_edge(irn, edge) {
- ir_node *user = get_edge_src_irn(edge);
-
- if(get_nodes_block(user) == bl && !pset_find_ptr(env->already_scheduled, user)) {
- unsigned visited_nr = get_irg_visited(irg) + 1;
- int max_hops;
+ ir_node *user = get_edge_src_irn(edge);
+ unsigned visited_nr = get_irg_visited(irg) + 1;
+ int max_hops;
- set_irg_visited(irg, visited_nr);
- max_hops = max_hops_walker(user, irn, 0, visited_nr);
- res = MAX(res, max_hops);
- }
+ set_irg_visited(irg, visited_nr);
+ max_hops = max_hops_walker(env, user, irn, 0, visited_nr);
+ res = MAX(res, max_hops);
}
return res;
}
-static void *reg_pressure_graph_init(const list_sched_selector_t *vtab, const arch_isa_t *isa, ir_graph *irg)
+static void *reg_pressure_graph_init(const list_sched_selector_t *vtab, const arch_env_t *arch_env, ir_graph *irg)
{
+ reg_pressure_main_env_t *main_env = xmalloc(sizeof(main_env[0]));
+
+ main_env->arch_env = arch_env;
+ main_env->vtab = vtab;
irg_walk_graph(irg, firm_clear_link, NULL, NULL);
- return (void *) vtab;
+
+ return main_env;
}
static void *reg_pressure_block_init(void *graph_env, ir_node *bl)
reg_pressure_selector_env_t *env = xmalloc(sizeof(env[0]));
obstack_init(&env->obst);
- env->already_scheduled = pset_new_ptr(32);
+ env->already_scheduled = new_nodeset(32);
env->root = NULL;
- env->vtab = graph_env;
+ env->main_env = graph_env;
/*
* Collect usage statistics.
*/
sched_foreach(bl, irn) {
- if(env->vtab->to_appear_in_schedule(env, irn)) {
+ if(must_appear_in_schedule(env->main_env->vtab, env, irn)) {
int i, n;
for(i = 0, n = get_irn_arity(irn); i < n; ++i) {
- ir_node *op = get_irn_n(irn, i);
- if(env->vtab->to_appear_in_schedule(env, irn)) {
+ //ir_node *op = get_irn_n(irn, i);
+ if(must_appear_in_schedule(env->main_env->vtab, env, irn)) {
usage_stats_t *us = get_or_set_usage_stats(env, irn);
+#if 0 /* Liveness is not computed here! */
if(is_live_end(bl, op))
us->uses_in_block = 99999;
else
+#endif
us->uses_in_block++;
}
}
set_irn_link(us->irn, NULL);
obstack_free(&env->obst, NULL);
- del_pset(env->already_scheduled);
+ del_nodeset(env->already_scheduled);
free(env);
}
+static int get_result_hops_sum(reg_pressure_selector_env_t *env, ir_node *irn)
+{
+ int res = 0;
+ if(get_irn_mode(irn) == mode_T) {
+ const ir_edge_t *edge;
+
+ foreach_out_edge(irn, edge)
+ res += get_result_hops_sum(env, get_edge_src_irn(edge));
+ }
+
+ else if(mode_is_data(get_irn_mode(irn)))
+ res = compute_max_hops(env, irn);
+
+
+ return res;
+}
+
static INLINE int reg_pr_costs(reg_pressure_selector_env_t *env, ir_node *irn)
{
int i, n;
for(i = 0, n = get_irn_arity(irn); i < n; ++i) {
ir_node *op = get_irn_n(irn, i);
- if(env->vtab->to_appear_in_schedule(env, op))
+ if(must_appear_in_schedule(env->main_env->vtab, env, op))
sum += compute_max_hops(env, op);
}
+ sum += get_result_hops_sum(env, irn);
+
return sum;
}
-ir_node *reg_pressure_select(void *block_env, pset *ready_set)
+static ir_node *reg_pressure_select(void *block_env, nodeset *ready_set)
{
reg_pressure_selector_env_t *env = block_env;
ir_node *irn, *res = NULL;
int curr_cost = INT_MAX;
- assert(pset_count(ready_set) > 0);
-
- for(irn = pset_first(ready_set); irn; irn = pset_next(ready_set)) {
- int costs = reg_pr_costs(env, irn);
- if(costs <= curr_cost) {
- res = irn;
- curr_cost = costs;
+ assert(nodeset_count(ready_set) > 0);
+
+ for (irn = nodeset_first(ready_set); irn; irn = nodeset_next(ready_set)) {
+ /*
+ Ignore branch instructions for the time being.
+ They should only be scheduled if there is nothing else.
+ */
+ if (! arch_irn_class_is(env->main_env->arch_env, irn, branch)) {
+ int costs = reg_pr_costs(env, irn);
+ if (costs <= curr_cost) {
+ res = irn;
+ curr_cost = costs;
+ }
}
}
- pset_insert_ptr(env->already_scheduled, res);
+ /*
+ There was no result so we only saw a branch.
+ Take it and finish.
+ */
+
+ if(!res) {
+ res = nodeset_first(ready_set);
+ nodeset_break(ready_set);
+
+ assert(res && "There must be a node scheduled.");
+ }
+
+ nodeset_insert(env->already_scheduled, res);
return res;
}
-static const list_sched_selector_t reg_pressure_selector_struct = {
- reg_pressure_graph_init,
- reg_pressure_block_init,
- reg_pressure_select,
- default_to_appear_in_schedule,
- reg_pressure_block_free,
- NULL
-};
+/**
+ * Environment for a block scheduler.
+ */
+typedef struct _block_sched_env_t {
+ sched_irn_t *sched_info; /**< scheduling info per node, copied from the global scheduler object */
+ sched_timestep_t curr_time; /**< current time of the scheduler */
+ nodeset *cands; /**< the set of candidates */
+ ir_node *block; /**< the current block */
+ sched_env_t *sched_env; /**< the scheduler environment */
+ nodeset *live; /**< simple liveness during scheduling */
+ const list_sched_selector_t *selector;
+ void *selector_block_env;
+ DEBUG_ONLY(firm_dbg_module_t *dbg;)
+} block_sched_env_t;
-const list_sched_selector_t *reg_pressure_selector = ®_pressure_selector_struct;
+/**
+ * Returns non-zero if the node is already scheduled
+ */
+static INLINE int is_already_scheduled(block_sched_env_t *env, ir_node *n)
+{
+ int idx = get_irn_idx(n);
-static void list_sched_block(ir_node *block, void *env_ptr);
+ assert(idx < ARR_LEN(env->sched_info));
+ return env->sched_info[idx].already_sched;
+}
-void list_sched(const struct _arch_isa_t *isa, ir_graph *irg)
+/**
+ * Mark a node as already scheduled
+ */
+static INLINE void mark_already_scheduled(block_sched_env_t *env, ir_node *n)
{
- sched_env_t env;
- const list_sched_selector_t *selector;
+ int idx = get_irn_idx(n);
- memset(&env, 0, sizeof(env));
- selector = env.selector = isa->impl->get_list_sched_selector(isa);
- env.selector_env = selector->init_graph ? selector->init_graph(selector, isa, irg) : NULL;
- env.irg = irg;
+ assert(idx < ARR_LEN(env->sched_info));
+ env->sched_info[idx].already_sched = 1;
+}
- /* Assure, that the out edges are computed */
- edges_assure(irg);
+/**
+ * Returns non-zero if the node is a root node
+ */
+static INLINE unsigned is_root_node(block_sched_env_t *env, ir_node *n)
+{
+ int idx = get_irn_idx(n);
- /* Schedule each single block. */
- irg_block_walk_graph(irg, list_sched_block, NULL, &env);
+ assert(idx < ARR_LEN(env->sched_info));
+ return env->sched_info[idx].is_root;
+}
- if(selector->finish_graph)
- selector->finish_graph(env.selector_env);
+/**
+ * Mark a node as roto node
+ */
+static INLINE void mark_root_node(block_sched_env_t *env, ir_node *n)
+{
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ env->sched_info[idx].is_root = 1;
}
+/**
+ * Get the current delay.
+ */
+static INLINE sched_timestep_t get_irn_delay(block_sched_env_t *env, ir_node *n) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ return env->sched_info[idx].delay;
+}
/**
- * Environment for a block scheduler.
+ * Set the current delay.
*/
-typedef struct _block_sched_env_t {
- int curr_time;
- pset *ready_set;
- pset *already_scheduled;
- ir_node *block;
- firm_dbg_module_t *dbg;
- const list_sched_selector_t *selector;
- void *selector_block_env;
-} block_sched_env_t;
+static INLINE void set_irn_delay(block_sched_env_t *env, ir_node *n, sched_timestep_t delay) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ env->sched_info[idx].delay = delay;
+}
+
+/**
+ * Get the current etime.
+ */
+static INLINE sched_timestep_t get_irn_etime(block_sched_env_t *env, ir_node *n) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ return env->sched_info[idx].etime;
+}
+
+/**
+ * Set the current etime.
+ */
+static INLINE void set_irn_etime(block_sched_env_t *env, ir_node *n, sched_timestep_t etime) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ env->sched_info[idx].etime = etime;
+}
+
+/**
+ * Get the number of users.
+ */
+static INLINE unsigned get_irn_num_user(block_sched_env_t *env, ir_node *n) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ return env->sched_info[idx].num_user;
+}
+
+/**
+ * Set the number of users.
+ */
+static INLINE void set_irn_num_user(block_sched_env_t *env, ir_node *n, unsigned num_user) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ env->sched_info[idx].num_user = num_user;
+}
+
+/**
+ * Get the register difference.
+ */
+static INLINE int get_irn_reg_diff(block_sched_env_t *env, ir_node *n) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ return env->sched_info[idx].reg_diff;
+}
+
+/**
+ * Set the register difference.
+ */
+static INLINE void set_irn_reg_diff(block_sched_env_t *env, ir_node *n, int reg_diff) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ env->sched_info[idx].reg_diff = reg_diff;
+}
+
+/**
+ * Get the pre-order position.
+ */
+static INLINE int get_irn_preorder(block_sched_env_t *env, ir_node *n) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ return env->sched_info[idx].preorder;
+}
+
+/**
+ * Set the pre-order position.
+ */
+static INLINE void set_irn_preorder(block_sched_env_t *env, ir_node *n, int pos) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ env->sched_info[idx].preorder = pos;
+}
+
+/**
+ * Get the pre-order position.
+ */
+static INLINE unsigned get_irn_critical_path_len(block_sched_env_t *env, ir_node *n) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ return env->sched_info[idx].critical_path_len;
+}
+
+/**
+ * Set the pre-order position.
+ */
+static INLINE void set_irn_critical_path_len(block_sched_env_t *env, ir_node *n, unsigned len) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ env->sched_info[idx].critical_path_len = len;
+}
+
+/**
+ * returns the exec-time for node n.
+ */
+static sched_timestep_t exectime(sched_env_t *env, ir_node *n) {
+ if (be_is_Keep(n) || is_Proj(n))
+ return 0;
+ if (env->selector->exectime)
+ return env->selector->exectime(env->selector_env, n);
+ return 1;
+}
+
+/**
+ * Calculates the latency for between two ops
+ */
+static sched_timestep_t latency(sched_env_t *env, ir_node *pred, int pred_cycle, ir_node *curr, int curr_cycle) {
+ /* a Keep hides a root */
+ if (be_is_Keep(curr))
+ return exectime(env, pred);
+
+ /* Proj's are executed immediately */
+ if (is_Proj(curr))
+ return 0;
+
+ /* predecessors Proj's must be skipped */
+ if (is_Proj(pred))
+ pred = get_Proj_pred(pred);
+
+ if (env->selector->latency)
+ return env->selector->latency(env->selector_env, pred, pred_cycle, curr, curr_cycle);
+ return 1;
+}
/**
* Try to put a node in the ready set.
- * @param env The block scheduler environment.
- * @param irn The node to make ready.
+ * @param env The block scheduler environment.
+ * @param pred The previous scheduled node.
+ * @param irn The node to make ready.
* @return 1, if the node could be made ready, 0 else.
*/
-static INLINE int make_ready(block_sched_env_t *env, ir_node *irn)
+static INLINE int make_ready(block_sched_env_t *env, ir_node *pred, ir_node *irn)
{
- int i, n;
+ int i, n;
+ sched_timestep_t etime_p, etime;
/* Blocks cannot be scheduled. */
- if(is_Block(irn))
+ if (is_Block(irn))
return 0;
/*
* Check, if the given ir node is in a different block as the
* currently scheduled one. If that is so, don't make the node ready.
*/
- if(env->block != get_nodes_block(irn))
+ if (env->block != get_nodes_block(irn))
return 0;
- for(i = 0, n = get_irn_arity(irn); i < n; ++i) {
+ for (i = 0, n = get_irn_arity(irn); i < n; ++i) {
ir_node *op = get_irn_n(irn, i);
+ /* if irn is an End we have keep-alives and op might be a block, skip that */
+ if (is_Block(op)) {
+ assert(get_irn_op(irn) == op_End);
+ continue;
+ }
+
/* If the operand is local to the scheduled block and not yet
* scheduled, this nodes cannot be made ready, so exit. */
- if(!pset_find_ptr(env->already_scheduled, op) && get_nodes_block(op) == env->block)
+ if (!is_already_scheduled(env, op) && get_nodes_block(op) == env->block)
return 0;
}
- DBG((env->dbg, LEVEL_2, "\tmaking ready: %+F\n", irn));
- pset_insert_ptr(env->ready_set, irn);
+ nodeset_insert(env->cands, irn);
- return 1;
-}
+ /* calculate the etime of this node */
+ etime = env->curr_time;
+ if (pred) {
+ etime_p = get_irn_etime(env, pred);
+ etime += latency(env->sched_env, pred, 1, irn, 0);
-/**
- * Check, if a node is ready in a block schedule.
- * @param env The block schedule environment.
- * @param irn The node to check for.
- * @return 1 if the node was ready, 0 if not.
- */
-#define is_ready(env,irn) \
- (pset_find_ptr((env)->ready_set, irn) != NULL)
+ etime = etime_p > etime ? etime_p : etime;
+ }
-/**
- * Check, if a node has already been schedules.
- * @param env The block schedule environment.
- * @param irn The node to check for.
- * @return 1 if the node was already scheduled, 0 if not.
- */
-#define is_scheduled(env,irn) \
- (pset_find_ptr((env)->already_scheduled, irn) != NULL)
+ set_irn_etime(env, irn, etime);
+
+ DB((env->dbg, LEVEL_2, "\tmaking ready: %+F etime %u\n", irn, etime));
+
+ return 1;
+}
/**
* Try, to make all users of a node ready.
foreach_out_edge(irn, edge) {
ir_node *user = edge->src;
if(!is_Phi(user))
- make_ready(env, user);
+ make_ready(env, irn, user);
}
}
/**
- * Compare to nodes using pointer equality.
- * @param p1 Node one.
- * @param p2 Node two.
- * @return 0 if they are identical.
+ * Returns the number of not yet schedules users.
*/
-static int node_cmp_func(const void *p1, const void *p2)
-{
- return p1 != p2;
+static INLINE int get_irn_not_sched_user(block_sched_env_t *env, ir_node *n) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ return env->sched_info[idx].num_not_sched_user;
+}
+
+/**
+ * Sets the number of not yet schedules users.
+ */
+static INLINE void set_irn_not_sched_user(block_sched_env_t *env, ir_node *n, int num) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ env->sched_info[idx].num_not_sched_user = num;
+}
+
+/**
+ * Add @p num to the number of not yet schedules users and returns the result.
+ */
+static INLINE int add_irn_not_sched_user(block_sched_env_t *env, ir_node *n, int num) {
+ int idx = get_irn_idx(n);
+
+ assert(idx < ARR_LEN(env->sched_info));
+ env->sched_info[idx].num_not_sched_user += num;
+ return env->sched_info[idx].num_not_sched_user;
+}
+
+/**
+ * Returns the number of users of a node having mode datab.
+ */
+static int get_num_successors(ir_node *irn) {
+ int sum = 0;
+ const ir_edge_t *edge;
+
+ if (get_irn_mode(irn) == mode_T) {
+ /* for mode_T nodes: count the users of all Projs */
+ foreach_out_edge(irn, edge) {
+ ir_node *proj = get_edge_src_irn(edge);
+ ir_mode *mode = get_irn_mode(proj);
+
+ if (mode == mode_T)
+ sum += get_num_successors(proj);
+ else if (mode_is_datab(mode))
+ sum += get_irn_n_edges(proj);
+ }
+ }
+ else {
+ /* do not count keep-alive edges */
+ foreach_out_edge(irn, edge) {
+ if (get_irn_opcode(get_edge_src_irn(edge)) != iro_End)
+ sum++;
+ }
+ }
+
+ return sum;
+}
+
+/**
+ * Adds irn to @p live, updates all inputs that this user is scheduled
+ * and counts all of it's non scheduled users.
+ */
+static void update_sched_liveness(block_sched_env_t *env, ir_node *irn) {
+ int i;
+
+ /* ignore Projs */
+ if (is_Proj(irn))
+ return;
+
+ for (i = get_irn_arity(irn) - 1; i >= 0; i--) {
+ ir_node *in = get_irn_n(irn, i);
+
+ /* if in is a proj: update predecessor */
+ while (is_Proj(in))
+ in = get_Proj_pred(in);
+
+ /* if in is still in the live set: reduce number of users by one */
+ if (nodeset_find(env->live, in)) {
+ if (add_irn_not_sched_user(env, in, -1) <= 0)
+ nodeset_remove(env->live, in);
+ }
+ }
+
+ /*
+ get_num_successors returns the number of all users. This includes
+ users in different blocks as well. As the each block is scheduled separately
+ the liveness info of those users will not be updated and so these
+ users will keep up the register pressure as it is desired.
+ */
+ i = get_num_successors(irn);
+ if (i > 0) {
+ set_irn_not_sched_user(env, irn, i);
+ nodeset_insert(env->live, irn);
+ }
+}
+
+/**
+ * Returns the current register pressure for the current block
+ * while scheduling.
+ * This is the number of already scheduled nodes having not yet
+ * scheduled users.
+ */
+static INLINE int get_cur_reg_pressure(block_sched_env_t *env) {
+ /*
+ Nodes with all users scheduled are removed from live set,
+ so the number of nodes in this set represent the current
+ register pressure in this block.
+ */
+ return nodeset_count(env->live);
}
/**
* Append an instruction to a schedule.
- * @param env The block scheduleing environment.
+ * @param env The block scheduling environment.
* @param irn The node to add to the schedule.
- * @return The given node.
+ * @return The given node.
*/
static ir_node *add_to_sched(block_sched_env_t *env, ir_node *irn)
{
/* If the node consumes/produces data, it is appended to the schedule
* list, otherwise, it is not put into the list */
- if(to_appear_in_schedule(irn)) {
+ if(must_appear_in_schedule(env->selector, env->selector_block_env, irn)) {
sched_info_t *info = get_irn_sched_info(irn);
INIT_LIST_HEAD(&info->list);
info->scheduled = 1;
+ update_sched_liveness(env, irn);
sched_add_before(env->block, irn);
DBG((env->dbg, LEVEL_2, "\tadding %+F\n", irn));
}
/* Insert the node in the set of all already scheduled nodes. */
- pset_insert_ptr(env->already_scheduled, irn);
+ mark_already_scheduled(env, irn);
/* Remove the node from the ready set */
- if(pset_find_ptr(env->ready_set, irn))
- pset_remove_ptr(env->ready_set, irn);
+ if(nodeset_find(env->cands, irn))
+ nodeset_remove(env->cands, irn);
return irn;
}
-
/**
* Add the proj nodes of a tuple-mode irn to the schedule immediately
* after the tuple-moded irn. By pinning the projs after the irn, no
* values.
*
* @param irn The tuple-moded irn.
- * @param list The schedule list to append all the projs.
- * @param time The time step to which the irn and all its projs are
- * related to.
- * @param obst The obstack the scheduling data structures shall be
- * created upon.
- * @param ready_set The ready set of the list scheduler.
- * @param already_scheduled A set containing all nodes already
- * scheduled.
*/
static void add_tuple_projs(block_sched_env_t *env, ir_node *irn)
{
assert(get_irn_mode(irn) == mode_T && "Mode of node must be tuple");
+ if(is_Bad(irn))
+ return;
+
foreach_out_edge(irn, edge) {
ir_node *out = edge->src;
assert(is_Proj(out) && "successor of a modeT node must be a proj");
- if(get_irn_mode(out) == mode_T)
+ if (get_irn_mode(out) == mode_T)
add_tuple_projs(env, out);
else {
add_to_sched(env, out);
}
}
+/**
+ * Returns the difference of regs_output - regs_input;
+ */
+static int get_reg_difference(block_sched_env_t *be, ir_node *irn) {
+ int num_out = 0;
+ int num_in = 0;
+ int i;
+
+ if (get_irn_mode(irn) == mode_T) {
+ /* mode_T nodes: num out regs == num Projs with mode datab */
+ const ir_edge_t *edge;
+ foreach_out_edge(irn, edge) {
+ ir_node *proj = get_edge_src_irn(edge);
+ if (mode_is_datab(get_irn_mode(proj)))
+ num_out++;
+ }
+ }
+ else
+ num_out = 1;
+
+ /* num in regs: number of ins with mode datab and not ignore */
+ for (i = get_irn_arity(irn) - 1; i >= 0; i--) {
+ ir_node *in = get_irn_n(irn, i);
+ if (mode_is_datab(get_irn_mode(in)) && ! arch_irn_is(be->sched_env->arch_env, in, ignore))
+ num_in++;
+ }
+
+ return num_out - num_in;
+}
+
+/**
+ * Execute the heuristic function,
+ */
+static ir_node *select_node_heuristic(block_sched_env_t *be, nodeset *ns)
+{
+ ir_node *irn, *cand = NULL;
+
+/* prefer instructions which can be scheduled early */
+#define PRIO_TIME 16
+/* prefer instructions with lots of successors */
+#define PRIO_NUMSUCCS 8
+/* prefer instructions with long critical path */
+#define PRIO_LEVEL 12
+/* prefer instructions coming early in preorder */
+#define PRIO_PREORD 8
+/* weight of current register pressure */
+#define PRIO_CUR_PRESS 20
+/* weight of register pressure difference */
+#define PRIO_CHG_PRESS 8
+
+ /* schedule keeps as early as possible */
+ foreach_nodeset(ns, irn) {
+ if (be_is_Keep(irn)) {
+ nodeset_break(ns);
+ return irn;
+ }
+ }
+
+ if (be->sched_env->sel_strategy & BE_SCHED_SELECT_HEUR) {
+ int max_prio = INT_MIN;
+ int cur_prio = INT_MIN;
+ int cur_pressure = get_cur_reg_pressure(be);
+ int reg_fact, cand_reg_fact;
+
+ /* priority based selection, heuristic inspired by mueller diss */
+ foreach_nodeset(ns, irn) {
+ /* make sure that branches are scheduled last */
+ if (! arch_irn_class_is(be->sched_env->arch_env, irn, branch)) {
+ int rdiff = get_irn_reg_diff(be, irn);
+ int sign = rdiff < 0;
+ int chg = (rdiff < 0 ? -rdiff : rdiff) << PRIO_CHG_PRESS;
+
+ reg_fact = chg << cur_pressure;
+ if (reg_fact < chg)
+ reg_fact = INT_MAX - 2;
+ reg_fact = sign ? -reg_fact : reg_fact;
+
+ cur_prio = (get_irn_critical_path_len(be, irn) << PRIO_LEVEL)
+ //- (get_irn_delay(be, irn) << PRIO_LEVEL)
+ + (get_irn_num_user(be, irn) << PRIO_NUMSUCCS)
+ - (get_irn_etime(be, irn) << PRIO_TIME)
+// - ((get_irn_reg_diff(be, irn) >> PRIO_CHG_PRESS) << ((cur_pressure >> PRIO_CUR_PRESS) - 3))
+ - reg_fact
+ + (get_irn_preorder(be, irn) << PRIO_PREORD); /* high preorder means early schedule */
+ if (cur_prio > max_prio) {
+ cand = irn;
+ max_prio = cur_prio;
+ cand_reg_fact = reg_fact;
+ }
+
+ DBG((be->dbg, LEVEL_4, "checked NODE %+F\n", irn));
+ DBG((be->dbg, LEVEL_4, "\tpriority: %d\n", cur_prio));
+ DBG((be->dbg, LEVEL_4, "\tpath len: %d (%d)\n", get_irn_critical_path_len(be, irn), get_irn_critical_path_len(be, irn) << PRIO_LEVEL));
+ DBG((be->dbg, LEVEL_4, "\tdelay: %d (%d)\n", get_irn_delay(be, irn), get_irn_delay(be, irn) << PRIO_LEVEL));
+ DBG((be->dbg, LEVEL_4, "\t#user: %d (%d)\n", get_irn_num_user(be, irn), get_irn_num_user(be, irn) << PRIO_NUMSUCCS));
+ DBG((be->dbg, LEVEL_4, "\tetime: %d (%d)\n", get_irn_etime(be, irn), -(get_irn_etime(be, irn) << PRIO_TIME)));
+ DBG((be->dbg, LEVEL_4, "\tpreorder: %d (%d)\n", get_irn_preorder(be, irn), get_irn_preorder(be, irn) << PRIO_PREORD));
+ DBG((be->dbg, LEVEL_4, "\treg diff: %d (%d)\n", get_irn_reg_diff(be, irn), -cand_reg_fact));
+ DBG((be->dbg, LEVEL_4, "\tpressure: %d\n", cur_pressure));
+ }
+ }
+
+ if (cand) {
+ DBG((be->dbg, LEVEL_4, "heuristic selected %+F:\n", cand));
+ }
+ else {
+ cand = nodeset_first(ns);
+ }
+ }
+ else {
+ /* use backend selector */
+ cand = be->selector->select(be->selector_block_env, ns);
+ }
+
+ return cand;
+}
+
+/**
+ * Returns non-zero if root is a root in the block block.
+ */
+static int is_root(ir_node *root, ir_node *block) {
+ const ir_edge_t *edge;
+
+ foreach_out_edge(root, edge) {
+ ir_node *succ = get_edge_src_irn(edge);
+
+ if (is_Block(succ))
+ continue;
+ /* Phi nodes are always in "another block */
+ if (is_Phi(succ))
+ continue;
+ if (get_nodes_block(succ) == block)
+ return 0;
+ }
+ return 1;
+}
+
+/* we need a special mark */
+static char _mark;
+#define MARK &_mark
+
+static firm_dbg_module_t *xxxdbg;
+
+/**
+ * descent into a dag and create a pre-order list.
+ */
+static void descent(ir_node *root, ir_node *block, ir_node **list, block_sched_env_t *env, unsigned path_len) {
+ int i;
+
+ if (! is_Phi(root)) {
+ path_len += exectime(env->sched_env, root);
+ if (get_irn_critical_path_len(env, root) < path_len) {
+ set_irn_critical_path_len(env, root, path_len);
+ }
+
+ /* Phi nodes always leave the block */
+ for (i = get_irn_arity(root) - 1; i >= 0; --i) {
+ ir_node *pred = get_irn_n(root, i);
+
+ DBG((xxxdbg, LEVEL_3, " node %+F\n", pred));
+ /* Blocks may happen as predecessors of End nodes */
+ if (is_Block(pred))
+ continue;
+
+ /* already seen nodes are not marked */
+ if (get_irn_link(pred) != MARK)
+ continue;
+
+ /* don't leave our block */
+ if (get_nodes_block(pred) != block)
+ continue;
+
+ set_irn_link(pred, NULL);
+
+ descent(pred, block, list, env, path_len);
+ }
+ }
+ set_irn_link(root, *list);
+ *list = root;
+}
+
/**
* Perform list scheduling on a block.
*
*/
static void list_sched_block(ir_node *block, void *env_ptr)
{
- sched_env_t *env = env_ptr;
- block_sched_env_t be;
+ sched_env_t *env = env_ptr;
const list_sched_selector_t *selector = env->selector;
+ ir_node *start_node = get_irg_start(get_irn_irg(block));
+ sched_info_t *info = get_irn_sched_info(block);
+
+ block_sched_env_t be;
const ir_edge_t *edge;
ir_node *irn;
- ir_node *start_node = get_irg_start(get_irn_irg(block));
- ir_node *final_jmp = NULL;
- int j, m;
- int phi_seen = 0;
- sched_info_t *info = get_irn_sched_info(block);
+ int j, m, cur_pos;
+
+ ir_node *root = NULL, *preord = NULL;
+ ir_node *curr;
/* Initialize the block's list head that will hold the schedule. */
INIT_LIST_HEAD(&info->list);
/* Initialize the block scheduling environment */
- be.dbg = firm_dbg_register("firm.be.sched");
+ be.sched_info = env->sched_info;
be.block = block;
be.curr_time = 0;
- be.ready_set = new_pset(node_cmp_func, get_irn_n_edges(block));
- be.already_scheduled = new_pset(node_cmp_func, get_irn_n_edges(block));
+ be.cands = new_nodeset(get_irn_n_edges(block));
+ be.live = new_nodeset(get_irn_n_edges(block));
be.selector = selector;
+ be.sched_env = env;
+ FIRM_DBG_REGISTER(be.dbg, "firm.be.sched");
+ FIRM_DBG_REGISTER(xxxdbg, "firm.be.sched");
- firm_dbg_set_mask(be.dbg, 0);
+ // firm_dbg_set_mask(be.dbg, SET_LEVEL_3);
- if(selector->init_block)
+ if (selector->init_block)
be.selector_block_env = selector->init_block(env->selector_env, block);
DBG((be.dbg, LEVEL_1, "scheduling %+F\n", block));
+ /* First step: Find the root set. */
+ foreach_out_edge(block, edge) {
+ ir_node *succ = get_edge_src_irn(edge);
+
+ if (is_root(succ, block)) {
+ mark_root_node(&be, succ);
+ set_irn_link(succ, root);
+ root = succ;
+ }
+ else
+ set_irn_link(succ, MARK);
+ }
+
+ /* Second step: calculate the pre-order list. */
+ preord = NULL;
+ for (curr = root; curr; curr = irn) {
+ irn = get_irn_link(curr);
+ DBG((be.dbg, LEVEL_2, " DAG root %+F\n", curr));
+ descent(curr, block, &preord, &be, 0);
+ }
+ root = preord;
+
+ /* Third step: calculate the Delay. Note that our
+ * list is now in pre-order, starting at root
+ */
+ for (cur_pos = 0, curr = root; curr; curr = get_irn_link(curr), cur_pos++) {
+ sched_timestep_t d;
+
+ if (arch_irn_class_is(env->arch_env, curr, branch)) {
+ /* assure, that branches can be executed last */
+ d = 0;
+ }
+ else {
+ if (is_root_node(&be, curr))
+ d = exectime(env, curr);
+ else {
+ d = 0;
+ foreach_out_edge(curr, edge) {
+ ir_node *n = get_edge_src_irn(edge);
+
+ if (get_nodes_block(n) == block) {
+ sched_timestep_t ld;
+
+ ld = latency(env, curr, 1, n, 0) + get_irn_delay(&be, n);
+ d = ld > d ? ld : d;
+ }
+ }
+ }
+ }
+ set_irn_delay(&be, curr, d);
+ DB((be.dbg, LEVEL_2, "\t%+F delay %u\n", curr, d));
+
+ /* set the etime of all nodes to 0 */
+ set_irn_etime(&be, curr, 0);
+
+ set_irn_preorder(&be, curr, cur_pos);
+ }
+
+
/* Then one can add all nodes are ready to the set. */
foreach_out_edge(block, edge) {
ir_node *irn = get_edge_src_irn(edge);
/* Skip the end node because of keepalive edges. */
- if(get_irn_opcode(irn) == iro_End)
+ if (get_irn_opcode(irn) == iro_End)
continue;
- /* Phi functions are scheduled immediately, since they only transfer
- * data flow from the predecessors to this block. */
- if(is_Phi(irn)) {
+ if (is_Phi(irn)) {
+ /* Phi functions are scheduled immediately, since they only transfer
+ * data flow from the predecessors to this block. */
+
+ /* Increase the time step. */
+ be.curr_time += get_irn_etime(&be, irn);
add_to_sched(&be, irn);
make_users_ready(&be, irn);
- phi_seen = 1;
}
+ else if (irn == start_node) {
+ /* The start block will be scheduled as the first node */
+ be.curr_time += get_irn_etime(&be, irn);
- else if(irn == start_node) {
add_to_sched(&be, irn);
add_tuple_projs(&be, irn);
}
-
- else if(get_irn_opcode(irn) == iro_Jmp) {
- final_jmp = irn;
- }
-
- /* Other nodes must have all operands in other blocks to be made
- * ready */
else {
+ /* Other nodes must have all operands in other blocks to be made
+ * ready */
int ready = 1;
/* Check, if the operands of a node are not local to this block */
- for(j = 0, m = get_irn_arity(irn); j < m; ++j) {
+ for (j = 0, m = get_irn_arity(irn); j < m; ++j) {
ir_node *operand = get_irn_n(irn, j);
- if(get_nodes_block(operand) == block) {
+ if (get_nodes_block(operand) == block) {
ready = 0;
break;
}
+ else {
+ /* live in values increase register pressure */
+ update_sched_liveness(&be, operand);
+ }
}
/* Make the node ready, if all operands live in a foreign block */
- if(ready) {
+ if (ready) {
DBG((be.dbg, LEVEL_2, "\timmediately ready: %+F\n", irn));
- make_ready(&be, irn);
+ make_ready(&be, NULL, irn);
}
+
+ /* calculate number of users (needed for heuristic) */
+ set_irn_num_user(&be, irn, get_num_successors(irn));
+
+ /* calculate register difference (needed for heuristic) */
+ set_irn_reg_diff(&be, irn, get_reg_difference(&be, irn));
}
}
- /* Increase the time, if some phi functions have been scheduled */
- be.curr_time += phi_seen;
+ while (nodeset_count(be.cands) > 0) {
+ nodeset *mcands; /**< the set of candidates with maximum delay time */
+ nodeset *ecands; /**< the set of nodes in mcands whose etime <= curr_time */
+ nodeset *local_cands;
+ sched_timestep_t max_delay = 0;
+
+ /* collect statistics about amount of ready nodes */
+ be_do_stat_sched_ready(block, be.cands);
- while(pset_count(be.ready_set) > 0) {
- /* select a node to be scheduled and check if it was ready */
- irn = selector->select(be.selector_block_env, be.ready_set);
+ /* calculate the max delay of all candidates */
+ foreach_nodeset(be.cands, irn) {
+ sched_timestep_t d = get_irn_delay(&be, irn);
- DBG((be.dbg, LEVEL_3, "\tpicked node %+F\n", irn));
+ max_delay = d > max_delay ? d : max_delay;
+ }
+
+ if (be.sched_env->sel_strategy & BE_SCHED_SELECT_MUCHNIK) {
+ mcands = new_nodeset(8);
+ ecands = new_nodeset(8);
+ }
+ else {
+ local_cands = new_nodeset(8);
+ }
+
+ /* calculate mcands and ecands */
+ foreach_nodeset(be.cands, irn) {
+ if (be_is_Keep(irn)) {
+ nodeset_break(be.cands);
+ break;
+ }
+ if (be.sched_env->sel_strategy & BE_SCHED_SELECT_MUCHNIK) {
+ if (get_irn_delay(&be, irn) == max_delay) {
+ nodeset_insert(mcands, irn);
+ if (get_irn_etime(&be, irn) <= be.curr_time)
+ nodeset_insert(ecands, irn);
+ }
+ }
+ else {
+ nodeset_insert(local_cands, irn);
+ }
+ }
+
+ if (irn) {
+ /* Keeps must be immediately scheduled */
+ }
+ else {
+ DB((be.dbg, LEVEL_2, "\tbe.curr_time = %u\n", be.curr_time));
+
+ /* select a node to be scheduled and check if it was ready */
+ if (! (be.sched_env->sel_strategy & BE_SCHED_SELECT_MUCHNIK)) {
+ DB((be.dbg, LEVEL_3, "\tlocal_cands = %d\n", nodeset_count(local_cands)));
+ irn = select_node_heuristic(&be, local_cands);
+ }
+ else if (nodeset_count(mcands) == 1) {
+ irn = nodeset_first(mcands);
+ DB((be.dbg, LEVEL_3, "\tirn = %+F, mcand = 1, max_delay = %u\n", irn, max_delay));
+ }
+ else {
+ int cnt = nodeset_count(ecands);
+ if (cnt == 1) {
+ irn = nodeset_first(ecands);
+
+ if (arch_irn_class_is(env->arch_env, irn, branch)) {
+ /* BEWARE: don't select a JUMP if others are still possible */
+ goto force_mcands;
+ }
+ DB((be.dbg, LEVEL_3, "\tirn = %+F, ecand = 1, max_delay = %u\n", irn, max_delay));
+ }
+ else if (cnt > 1) {
+ DB((be.dbg, LEVEL_3, "\tecand = %d, max_delay = %u\n", cnt, max_delay));
+ irn = select_node_heuristic(&be, ecands);
+ }
+ else {
+force_mcands:
+ DB((be.dbg, LEVEL_3, "\tmcand = %d\n", nodeset_count(mcands)));
+ irn = select_node_heuristic(&be, mcands);
+ }
+ }
+ }
+
+ if (be.sched_env->sel_strategy & BE_SCHED_SELECT_MUCHNIK) {
+ del_nodeset(mcands);
+ del_nodeset(ecands);
+ }
+ else {
+ del_nodeset(local_cands);
+ }
+
+ DB((be.dbg, LEVEL_2, "\tpicked node %+F\n", irn));
+
+ /* Increase the time step. */
+ be.curr_time += exectime(env, irn);
/* Add the node to the schedule. */
add_to_sched(&be, irn);
- if(get_irn_mode(irn) == mode_T)
+ if (get_irn_mode(irn) == mode_T)
add_tuple_projs(&be, irn);
else
make_users_ready(&be, irn);
- /* Increase the time step. */
- be.curr_time += 1;
-
/* remove the scheduled node from the ready list. */
- if(pset_find_ptr(be.ready_set, irn))
- pset_remove_ptr(be.ready_set, irn);
+ if (nodeset_find(be.cands, irn))
+ nodeset_remove(be.cands, irn);
}
- if(selector->finish_block)
+ if (selector->finish_block)
selector->finish_block(be.selector_block_env);
- if(final_jmp)
- add_to_sched(&be, final_jmp);
+ del_nodeset(be.cands);
+ del_nodeset(be.live);
+}
+
+static const list_sched_selector_t reg_pressure_selector_struct = {
+ reg_pressure_graph_init,
+ reg_pressure_block_init,
+ reg_pressure_select,
+ NULL, /* to_appear_in_schedule */
+ NULL, /* exectime */
+ NULL, /* latency */
+ reg_pressure_block_free,
+ free
+};
+
+const list_sched_selector_t *reg_pressure_selector = ®_pressure_selector_struct;
+
+/* List schedule a graph. */
+void list_sched(const be_irg_t *birg, be_options_t *be_opts)
+{
+ const arch_env_t *arch_env = birg->main_env->arch_env;
+ ir_graph *irg = birg->irg;
+
+ int num_nodes;
+ sched_env_t env;
+ mris_env_t *mris;
+
+ /* Assure, that the out edges are computed */
+ edges_assure(irg);
+
+ if(be_opts->mris)
+ mris = be_sched_mris_preprocess(birg);
+
+ num_nodes = get_irg_last_idx(irg);
+
+ memset(&env, 0, sizeof(env));
+ env.selector = arch_env->isa->impl->get_list_sched_selector(arch_env->isa);
+ env.arch_env = arch_env;
+ env.irg = irg;
+ env.sel_strategy = be_opts->sched_select;
+ env.sched_info = NEW_ARR_F(sched_irn_t, num_nodes);
+
+ memset(env.sched_info, 0, num_nodes * sizeof(*env.sched_info));
+
+ if (env.selector->init_graph)
+ env.selector_env = env.selector->init_graph(env.selector, arch_env, irg);
+
+ /* Schedule each single block. */
+ irg_block_walk_graph(irg, list_sched_block, NULL, &env);
+
+ if (env.selector->finish_graph)
+ env.selector->finish_graph(env.selector_env);
+
+ if(be_opts->mris)
+ be_sched_mris_free(mris);
- del_pset(be.ready_set);
- del_pset(be.already_scheduled);
+ DEL_ARR_F(env.sched_info);
}