--- /dev/null
+/*
+ * Author: Matthias Braun
+ * Date: 26.7.06
+ * Copyright: (c) Universitaet Karlsruhe
+ * Licence: This file protected by GPL - GNU GENERAL PUBLIC LICENSE.
+ */
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+
+#include <stdlib.h>
+
+#include "set.h"
+
+#include "irgwalk.h"
+#include "ircons.h"
+#include "irprintf.h"
+#include "execfreq.h"
+#include "unionfind.h"
+#include "type.h"
+#include "irdump_t.h"
+
+#include "benode_t.h"
+#include "bespillslots.h"
+#include "bechordal_t.h"
+#include "bejavacoal.h"
+
+
+#define DBG_COALESCING 1
+#define DBG_INTERFERENCES 2
+
+DEBUG_ONLY(
+static firm_dbg_module_t *dbg = NULL;
+)
+
+typedef struct _spill_t {
+ ir_node *spill;
+ /** regclass of the spilled value */
+ const arch_register_class_t *cls;
+ /** index into spillslot_unionfind unionfind structure */
+ int spillslot;
+} spill_t;
+
+typedef struct _affinity_edge_t {
+ double affinity;
+ int slot1, slot2;
+} affinity_edge_t;
+
+typedef struct _ss_env_t {
+ struct obstack obst;
+ const be_chordal_env_t *chordal_env;
+ set *spills;
+ ir_node **reloads;
+ affinity_edge_t **affinity_edges;
+ set *memperms;
+} ss_env_t;
+
+/** Compare 2 affinity edges (used in quicksort) */
+static int cmp_affinity(const void *d1, const void *d2) {
+ const affinity_edge_t *e1 = d1;
+ const affinity_edge_t *e2 = d2;
+
+ return e1->affinity < e2->affinity ? -1 : 1;
+}
+
+int cmp_spill(const void* d1, const void* d2, size_t size) {
+ const spill_t* s1 = d1;
+ const spill_t* s2 = d2;
+ return s1->spill != s2->spill;
+}
+
+static spill_t *get_spill(ss_env_t *env, ir_node *node) {
+ spill_t spill, *res;
+ int hash = HASH_PTR(node);
+
+ spill.spill = node;
+ res = set_find(env->spills, &spill, sizeof(spill), hash);
+
+ return res;
+}
+
+/*
+ * ____ _ _ _ ____ _ _ _
+ * / ___|___ | | | ___ ___| |_ / ___| _ __ (_) | |___
+ * | | / _ \| | |/ _ \/ __| __| \___ \| '_ \| | | / __|
+ * | |__| (_) | | | __/ (__| |_ ___) | |_) | | | \__ \
+ * \____\___/|_|_|\___|\___|\__| |____/| .__/|_|_|_|___/
+ * |_|
+ */
+
+static spill_t *collect_spill(ss_env_t *env, ir_node *node, const arch_register_class_t *cls) {
+ spill_t spill, *res;
+ int hash = HASH_PTR(node);
+
+ assert(be_is_Spill(node));
+
+ spill.spill = node;
+ res = set_find(env->spills, &spill, sizeof(spill), hash);
+
+ if(res == NULL) {
+ spill.spillslot = set_count(env->spills);
+ spill.cls = cls;
+ res = set_insert(env->spills, &spill, sizeof(spill), hash);
+ } else {
+ assert(cls == res->cls);
+ }
+
+ return res;
+}
+
+static spill_t *collect_memphi(ss_env_t *env, ir_node *node, const arch_register_class_t *cls) {
+ int i, arity;
+ spill_t spill, *res;
+ int hash = HASH_PTR(node);
+
+ assert(is_Phi(node));
+
+ spill.spill = node;
+ res = set_find(env->spills, &spill, sizeof(spill), hash);
+ if(res != NULL) {
+ return res;
+ }
+
+ spill.spillslot = set_count(env->spills);
+ spill.cls = cls;
+ res = set_insert(env->spills, &spill, sizeof(spill), hash);
+
+ // is 1 of the arguments a spill?
+ for(i = 0, arity = get_irn_arity(node); i < arity; ++i) {
+ affinity_edge_t *affinty_edge;
+ ir_node* arg = get_irn_n(node, i);
+ spill_t* arg_spill;
+
+ if(be_is_Spill(arg)) {
+ arg_spill = collect_spill(env, arg, cls);
+ } else {
+ // if it wasn't a spill then it must be a Mem-Phi
+ assert(is_Phi(arg));
+ arg_spill = collect_memphi(env, arg, cls);
+ }
+
+ // add an affinity edge
+ affinty_edge = obstack_alloc(&env->obst, sizeof(affinty_edge[0]));
+ affinty_edge->affinity = get_block_execfreq(env->chordal_env->exec_freq, get_nodes_block(arg));
+ affinty_edge->slot1 = res->spillslot;
+ affinty_edge->slot2 = arg_spill->spillslot;
+ ARR_APP1(affinity_edge_t*, env->affinity_edges, affinty_edge);
+ }
+
+ return res;
+}
+
+/**
+ * This walker function searches for reloads and collects all the spills
+ * and memphis attached to them.
+ */
+static void collect_spills_walker(ir_node *node, void *data) {
+ ss_env_t *env = data;
+
+ if(be_is_Reload(node)) {
+ ir_node *spill = get_irn_n(node, be_pos_Reload_mem);
+ const arch_env_t *arch_env = env->chordal_env->birg->main_env->arch_env;
+ const arch_register_class_t *cls = arch_get_irn_reg_class(arch_env, node, -1);
+
+ if(is_Phi(spill)) {
+ collect_memphi(env, spill, cls);
+ } else {
+ collect_spill(env, spill, cls);
+ }
+ ARR_APP1(ir_node*, env->reloads, node);
+ }
+}
+
+/*
+ * ____ _ ____ _ _
+ * / ___|___ __ _| | ___ ___ ___ ___ / ___|| | ___ | |_ ___
+ * | | / _ \ / _` | |/ _ \/ __|/ __/ _ \ \___ \| |/ _ \| __/ __|
+ * | |__| (_) | (_| | | __/\__ \ (_| __/ ___) | | (_) | |_\__ \
+ * \____\___/ \__,_|_|\___||___/\___\___| |____/|_|\___/ \__|___/
+ */
+
+static int merge_interferences(ss_env_t *env, bitset_t** interferences, int* spillslot_unionfind, int s1, int s2)
+{
+ int res;
+ int i;
+ int spillcount;
+
+ // merge spillslots and interferences
+ res = uf_union(spillslot_unionfind, s1, s2);
+ // we assume that we always merge s2 to s1 so swap s1, s2 if necessary
+ if(res != 0) {
+ int t = s1;
+ s1 = s2;
+ s2 = t;
+ }
+
+ bitset_or(interferences[s1], interferences[s2]);
+
+ // update other interferences
+ spillcount = set_count(env->spills);
+ for(i = 0; i < spillcount; ++i) {
+ bitset_t *intfs = interferences[i];
+ if(bitset_is_set(intfs, s2))
+ bitset_set(intfs, s1);
+ }
+
+ return res;
+}
+
+#if 0
+
+static void dump_interference_graph(ss_env_t *env, bitset_t **interferences, const char* suffix) {
+ char name[256];
+ int i;
+ int spillcount;
+ spill_t *spill;
+ FILE *f;
+ static int cnt = 0;
+
+ snprintf(name, sizeof(name), "%d-%s-spillslots-%s.vcg", cnt++, get_irg_dump_name(env->chordal_env->birg->irg), suffix);
+
+ f = fopen(name, "w");
+ assert(f != NULL);
+
+ fprintf(f, "graph: {\n");
+
+ spillcount = set_count(env->spills);
+ for(spill = set_first(env->spills), i = 0; spill != NULL; spill = set_next(env->spills), ++i) {
+ int slotid = spill->spillslot;
+ fprintf(f, "\tnode: { title: \"n%d\" label: \"%d\" }\n", i, slotid);
+ }
+
+ for(i = 0; i < ARR_LEN(env->affinity_edges); ++i) {
+ affinity_edge_t *edge = env->affinity_edges[i];
+ fprintf(f, "\tedge: { sourcename: \"n%d\" targetname: \"n%d\" color: green }\n", edge->slot1, edge->slot2);
+ }
+
+ for(i = 0; i < spillcount; ++i) {
+ int i2;
+ for(i2 = 0; i2 < spillcount; ++i2) {
+ if(bitset_is_set(interferences[i], i2)) {
+ fprintf(f, "\tedge: { sourcename: \"n%d\" targetname: \"n%d\" color: red }\n", i, i2);
+ }
+ }
+ }
+
+ fprintf(f, "}\n");
+ fclose(f);
+}
+
+static void show_stats(ss_env_t *env) {
+ int spillcount;
+ int slotcount;
+ int *slotused;
+ spill_t *spill;
+
+ spillcount = set_count(env->spills);
+ fprintf(stderr, "%s: Collected %d spills\n", get_irg_dump_name(env->chordal_env->birg->irg), spillcount);
+
+ slotused = alloca(spillcount * sizeof(slotused[0]));
+ memset(slotused, 0, spillcount * sizeof(slotused[0]));
+
+ slotcount = 0;
+ for(spill = set_first(env->spills); spill != NULL; spill = set_next(env->spills)) {
+ int slot = spill->spillslot;
+ if(slotused[slot] == 0) {
+ slotused[slot] = 1;
+ slotcount++;
+ }
+ }
+
+ fprintf(stderr, "%s: Coalesced to %d spillslots\n", get_irg_dump_name(env->chordal_env->birg->irg), slotcount);
+}
+
+#endif
+
+static void assign_spillslots(ss_env_t *env);
+
+/**
+ * A greedy coalescing algorithm for spillslots:
+ * 1. Sort the list of affinity edges
+ * 2. Try to merge slots with affinity edges (most expensive slots first)
+ * 3. Try to merge everything else that is possible
+ */
+static void do_greedy_coalescing(ss_env_t *env)
+{
+ int spillcount;
+ spill_t **spilllist;
+ spill_t *spill;
+ int i, i2;
+ int affinity_edge_count;
+ bitset_t **interferences;
+ int* spillslot_unionfind;
+
+ spillcount = set_count(env->spills);
+ if(spillcount == 0)
+ return;
+
+ DBG((dbg, DBG_COALESCING, "Coalescing %d spillslots\n", spillcount));
+
+ interferences = alloca(spillcount * sizeof(interferences[0]));
+ spillslot_unionfind = alloca(spillcount * sizeof(spillslot_unionfind[0]));
+ spilllist = alloca(spillcount * sizeof(spilllist[0]));
+
+ uf_init(spillslot_unionfind, 0, spillcount);
+
+ DEBUG_ONLY(
+ memset(spilllist, 0, spillcount * sizeof(spilllist[0]));
+ );
+
+ for(spill = set_first(env->spills), i = 0; spill != NULL; spill = set_next(env->spills), ++i) {
+ assert(spill->spillslot < spillcount);
+ spilllist[spill->spillslot] = spill;
+ }
+
+ for(i = 0; i < spillcount; ++i) {
+ interferences[i] = bitset_alloca(spillcount);
+ }
+
+ // construct interferences
+ for(i = 0; i < spillcount; ++i) {
+ for(i2 = i+1; i2 < spillcount; ++i2) {
+ if(values_interfere(env->chordal_env->lv, spilllist[i]->spill, spilllist[i2]->spill)) {
+ DBG((dbg, DBG_INTERFERENCES, "Slot %d and %d interfere\n", i, i2));
+ bitset_set(interferences[i], i2);
+ bitset_set(interferences[i2], i);
+ }
+ }
+ }
+
+ // sort affinity edges
+ affinity_edge_count = ARR_LEN(env->affinity_edges);
+ qsort(env->affinity_edges, affinity_edge_count, sizeof(env->affinity_edges[0]), cmp_affinity);
+
+ //dump_interference_graph(env, interferences, "before");
+
+ // try to merge affine nodes
+ for(i = 0; i < affinity_edge_count; ++i) {
+ const affinity_edge_t *edge = env->affinity_edges[i];
+ int s1 = uf_find(spillslot_unionfind, edge->slot1);
+ int s2 = uf_find(spillslot_unionfind, edge->slot2);
+
+ /* test if values interfere */
+ if(bitset_is_set(interferences[s1], s2)) {
+ assert(bitset_is_set(interferences[s2], s1));
+ continue;
+ }
+
+ DBG((dbg, DBG_COALESCING, "Merging %d and %d because of affinity edge\n", s1, s2));
+
+ merge_interferences(env, interferences, spillslot_unionfind, s1, s2);
+ }
+
+ // try to merge as much remaining spillslots as possible
+ for(i = 0; i < spillcount; ++i) {
+ int s1 = uf_find(spillslot_unionfind, i);
+ if(s1 != i)
+ continue;
+
+ for(i2 = i+1; i2 < spillcount; ++i2) {
+ int s2 = uf_find(spillslot_unionfind, i2);
+ if(s2 != i2)
+ continue;
+
+ /* test if values interfere
+ * we have to test n1-n2 and n2-n1, because only 1 side gets updated
+ * when node merging occurs
+ */
+ if(bitset_is_set(interferences[s1], s2)) {
+ assert(bitset_is_set(interferences[s2], s1));
+ continue;
+ }
+
+ DBG((dbg, DBG_COALESCING, "Merging %d and %d because it is possible\n", s1, s2));
+
+ if(merge_interferences(env, interferences, spillslot_unionfind, s1, s2) != 0) {
+ // we can break the loop here, because s2 is the new supernode now
+ // and we'll test s2 again later anyway
+ break;
+ }
+ }
+ }
+
+ // assign spillslots
+ for(i = 0; i < spillcount; ++i) {
+ spill_t *spill = spilllist[i];
+
+ spill->spillslot = uf_find(spillslot_unionfind, i);
+ }
+
+ //dump_interference_graph(env, interferences, "after");
+}
+
+static void do_java_coalescing(ss_env_t *env)
+{
+ int spillcount;
+ spill_t **spilllist;
+ spill_t *spill;
+ int i, i2;
+ be_java_coal_t *coal;
+
+ spillcount = set_count(env->spills);
+ if(spillcount == 0)
+ return;
+
+ spilllist = alloca(spillcount * sizeof(spilllist[0]));
+
+ DEBUG_ONLY(
+ memset(spilllist, 0, spillcount * sizeof(spilllist[0]));
+ );
+
+ coal = be_java_coal_init("spillslot coalescing", spillcount, spillcount, 1);
+
+ for(spill = set_first(env->spills), i = 0; spill != NULL; spill = set_next(env->spills), ++i) {
+ assert(spill->spillslot < spillcount);
+ DEBUG_ONLY(assert(spilllist[spill->spillslot] == NULL));
+ spilllist[spill->spillslot] = spill;
+
+ be_java_coal_set_color(coal, spill->spillslot, spill->spillslot);
+ }
+
+ // construct interferences
+ for(i = 0; i < spillcount; ++i) {
+ for(i2 = i+1; i2 < spillcount; ++i2) {
+ if(values_interfere(env->chordal_env->lv, spilllist[i]->spill, spilllist[i2]->spill)) {
+ be_java_coal_add_int_edge(coal, i, i2);
+ }
+ }
+ }
+
+ for(i = 0; i < ARR_LEN(env->affinity_edges); ++i) {
+ const affinity_edge_t *edge = env->affinity_edges[i];
+ int n = edge->slot1;
+ int m = edge->slot2;
+ int costs = (int) (edge->affinity * 10000);
+ be_java_coal_add_aff_edge(coal, n, m, costs);
+ }
+
+ be_java_coal_coalesce(coal);
+
+ // construct spillslots
+ for(i = 0; i < spillcount; ++i) {
+ spill_t *spill = spilllist[i];
+ spill->spillslot = be_java_coal_get_color(coal, i);
+ }
+ be_java_coal_destroy(coal);
+}
+
+/*
+ * _ _ _____ _ _ _ _
+ * / \ ___ ___(_) __ _ _ __ | ____|_ __ | |_(_) |_(_) ___ ___
+ * / _ \ / __/ __| |/ _` | '_ \ | _| | '_ \| __| | __| |/ _ \/ __|
+ * / ___ \\__ \__ \ | (_| | | | | | |___| | | | |_| | |_| | __/\__ \
+ * /_/ \_\___/___/_|\__, |_| |_| |_____|_| |_|\__|_|\__|_|\___||___/
+ * |___/
+ */
+
+typedef struct _spill_slot_t {
+ int size;
+ int align;
+ entity *entity;
+} spill_slot_t;
+
+typedef struct _memperm_entry_t {
+ ir_node* node;
+ int pos;
+ entity *in;
+ entity *out;
+ struct _memperm_entry_t *next;
+} memperm_entry_t;
+
+typedef struct _memperm_t {
+ ir_node *block;
+ int entrycount;
+ memperm_entry_t *entries;
+} memperm_t;
+
+int cmp_memperm(const void* d1, const void* d2, size_t size) {
+ const memperm_t* e1 = d1;
+ const memperm_t* e2 = d2;
+ return e1->block != e2->block;
+}
+
+static memperm_t *get_memperm(ss_env_t *env, ir_node *block) {
+ memperm_t entry, *res;
+ int hash;
+
+ entry.block = block;
+ hash = HASH_PTR(block);
+
+ res = set_find(env->memperms, &entry, sizeof(entry), hash);
+
+ if(res == NULL) {
+ entry.entrycount = 0;
+ entry.entries = NULL;
+ res = set_insert(env->memperms, &entry, sizeof(entry), hash);
+ }
+
+ return res;
+}
+
+static entity* create_stack_entity(ss_env_t *env, spill_slot_t *slot) {
+ ir_type* frame = get_irg_frame_type(env->chordal_env->irg);
+ entity* res = frame_alloc_area(frame, slot->size, slot->align, 0);
+
+ slot->entity = res;
+
+ return res;
+}
+
+static int get_spillslotsize_for_spill(ss_env_t *env, spill_t *spill) {
+ const ir_mode *mode = arch_register_class_mode(spill->cls);
+
+ return get_mode_size_bytes(mode);
+}
+
+static int get_spillslotalign_for_spill(ss_env_t *env, spill_t *spill) {
+ const arch_isa_t *isa = env->chordal_env->birg->main_env->arch_env->isa;
+
+ return arch_isa_get_reg_class_alignment(isa, spill->cls);
+}
+
+/**
+ * Enlarges a spillslot (if necessary) so that it can carry a value of size
+ * @p othersize and alignment @p otheralign.
+ */
+static void enlarge_spillslot(spill_slot_t *slot, int otheralign, int othersize) {
+ if(othersize > slot->size) {
+ slot->size = othersize;
+ }
+ if(otheralign > slot->align) {
+ if(otheralign % slot->align != 0)
+ slot->align *= otheralign;
+ else
+ slot->align = otheralign;
+ } else if(slot->align % otheralign != 0) {
+ slot->align *= otheralign;
+ }
+}
+
+/**
+ * Create stack entities for the spillslots and assign them to the spill and
+ * reload nodes.
+ */
+static void assign_spillslots(ss_env_t *env) {
+ int i;
+ int spillcount;
+ spill_t *spill;
+ spill_slot_t* spillslots;
+
+ spillcount = set_count(env->spills);
+ spillslots = alloca(spillcount * sizeof(spillslots[0]));
+
+ memset(spillslots, 0, spillcount * sizeof(spillslots[0]));
+
+ // construct spillslots
+ for(spill = set_first(env->spills); spill != NULL; spill = set_next(env->spills)) {
+ int slotid = spill->spillslot;
+ spill_slot_t *slot = & (spillslots[slotid]);
+ int align = get_spillslotalign_for_spill(env, spill);
+ int size = get_spillslotsize_for_spill(env, spill);
+
+ if(slot->align == 0 && slot->size == 0) {
+ slot->align = align;
+ slot->size = size;
+ } else {
+ enlarge_spillslot(slot, align, size);
+ }
+ }
+
+ for(spill = set_first(env->spills); spill != NULL; spill = set_next(env->spills)) {
+ spill_slot_t *slot;
+ ir_node *node = spill->spill;
+ int slotid = spill->spillslot;
+
+ slot = &spillslots[slotid];
+ if(slot->entity == NULL) {
+ create_stack_entity(env, slot);
+ }
+
+ if(be_is_Spill(node)) {
+ be_set_frame_entity(node, slot->entity);
+ } else {
+ int i, arity;
+
+ // should be a PhiM
+ assert(is_Phi(node));
+
+ for(i = 0, arity = get_irn_arity(node); i < arity; ++i) {
+ ir_node *arg = get_irn_n(node, i);
+ spill_t *argspill;
+ int argslotid;
+
+ argspill = get_spill(env, arg);
+ assert(argspill != NULL);
+
+ argslotid = argspill->spillslot;
+ if(slotid != argslotid) {
+ memperm_t *memperm;
+ memperm_entry_t *entry;
+ spill_slot_t *argslot = &spillslots[argslotid];
+ if(argslot->entity == NULL) {
+ create_stack_entity(env, argslot);
+ }
+
+ memperm = get_memperm(env, get_nodes_block(arg));
+
+ entry = obstack_alloc(&env->obst, sizeof(entry[0]));
+ entry->node = node;
+ entry->pos = i;
+ entry->in = argslot->entity;
+ entry->out = slot->entity;
+ entry->next = memperm->entries;
+ memperm->entrycount++;
+ memperm->entries = entry;
+ }
+ }
+ }
+ }
+
+ for(i = 0; i < ARR_LEN(env->reloads); ++i) {
+ const ir_node* reload = env->reloads[i];
+ ir_node* spillnode = get_irn_n(reload, be_pos_Reload_mem);
+ spill_t *spill = get_spill(env, spillnode);
+ const spill_slot_t *slot = & spillslots[spill->spillslot];
+
+ assert(slot->entity != NULL);
+
+ be_set_frame_entity(reload, slot->entity);
+ }
+}
+
+static void create_memperms(ss_env_t *env) {
+ memperm_t *memperm;
+
+ for(memperm = set_first(env->memperms); memperm != NULL; memperm = set_next(env->memperms)) {
+ int i;
+ memperm_entry_t *entry;
+ ir_node** nodes = alloca(memperm->entrycount * sizeof(nodes[0]));
+ ir_node* mempermnode;
+
+ assert(memperm->entrycount > 0);
+
+ for(entry = memperm->entries, i = 0; entry != NULL; entry = entry->next, ++i) {
+ ir_node* arg = get_irn_n(entry->node, entry->pos);
+ nodes[i] = arg;
+ }
+
+ mempermnode = be_new_MemPerm(env->chordal_env->birg->main_env->arch_env, env->chordal_env->irg, memperm->block,
+ memperm->entrycount, nodes);
+
+ for(entry = memperm->entries, i = 0; entry != NULL; entry = entry->next, ++i) {
+ ir_node *proj;
+ ir_node* arg = get_irn_n(entry->node, entry->pos);
+
+ be_set_MemPerm_in_entity(mempermnode, i, entry->in);
+ be_set_MemPerm_out_entity(mempermnode, i, entry->out);
+ proj = new_Proj(mempermnode, get_irn_mode(arg), i);
+ set_irn_n(arg, entry->pos, proj);
+ }
+ ir_printf("Memperm created in block %+F\n", memperm->block);
+ }
+}
+
+void be_coalesce_spillslots(const be_chordal_env_t *chordal_env) {
+ ss_env_t env;
+
+ obstack_init(&env.obst);
+ env.chordal_env = chordal_env;
+ env.spills = new_set(cmp_spill, 10);
+ env.reloads = NEW_ARR_F(ir_node*, 0);
+ env.affinity_edges = NEW_ARR_F(affinity_edge_t*, 0);
+ env.memperms = new_set(cmp_memperm, 10);
+ FIRM_DBG_REGISTER(dbg, "firm.be.spillslots");
+ //firm_dbg_set_mask(dbg, DBG_COALESCING);
+
+ /* Get initial spill slots */
+ irg_walk_graph(chordal_env->irg, NULL, collect_spills_walker, &env);
+
+ do_greedy_coalescing(&env);
+ //do_java_coalescing(&env);
+
+ assign_spillslots(&env);
+
+ create_memperms(&env);
+
+ //show_stats(&env);
+
+ del_set(env.memperms);
+ DEL_ARR_F(env.reloads);
+ DEL_ARR_F(env.affinity_edges);
+ del_set(env.spills);
+ obstack_free(&env.obst, NULL);
+}
projects / libfirm / commitdiff