X-Git-Url: http://nsz.repo.hu/git/?a=blobdiff_plain;f=ir%2Fbe%2Fbeflags.c;h=40282aee6c3abba5fffb1744fdbb5e6527fe009b;hb=52a59fbe6fbd93a12a265842fd08fbc8b12305f8;hp=090a1b7cf61724cc28108507169b9febb4ff6e97;hpb=897c780100ab59a7968eef621f2fa47bd711b089;p=libfirm

diff --git a/ir/be/beflags.c b/ir/be/beflags.c
index 090a1b7cf..40282aee6 100644
--- a/ir/be/beflags.c
+++ b/ir/be/beflags.c
@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 1995-2007 University of Karlsruhe.  All right reserved.
+ * Copyright (C) 1995-2008 University of Karlsruhe.  All right reserved.
  *
  * This file is part of libFirm.
  *
@@ -22,10 +22,22 @@
  * @brief       modifies schedule so flags dependencies are respected.
  * @author      Matthias Braun, Christoph Mallon
  * @version     $Id: besched.h 14693 2007-06-21 15:35:49Z beck $
+ *
+ * Fixup schedule to respect flag constraints by moving and rematerialisation of
+ * nodes.
+ *
+ * Flags are modeled as register classes with ignore registers. However to avoid
+ * bloating the graph, only flag-consumer -> producer dependencies are
+ * explicitely modeled in the graph. Nodes that just change the flags are only
+ * marked with the arch_irn_flags_modify_flags flag.
+ *
+ * Flags are usually a limited resource that can't (or at least shouldn't) be
+ * spilled. So in some situations (for example 2 adc-nodes that use the flags of
+ * a single add node on x86) operations have to be repeated to work correctly.
  */
-#ifdef HAVE_CONFIG_H
 #include "config.h"
-#endif
+
+#include <stdbool.h>
 
 #include "irgwalk.h"
 #include "irnode_t.h"
@@ -36,31 +48,49 @@
 #include "error.h"
 
 #include "beflags.h"
-#include "bearch_t.h"
-#include "beirg_t.h"
-#include "besched_t.h"
+#include "bearch.h"
+#include "beirg.h"
+#include "beirgmod.h"
+#include "besched.h"
+#include "benode.h"
+#include "belive.h"
+#include "beabihelper.h"
 
-static const arch_env_t            *arch_env   = NULL;
-static const arch_register_class_t *flag_class = NULL;
-static const arch_register_t       *flags_reg  = NULL;
-static func_rematerialize           remat      = NULL;
+static const arch_register_class_t *flag_class;
+static const arch_register_t       *flags_reg;
+static func_rematerialize           remat;
+static check_modifies_flags         check_modify;
+static int                          changed;
 
 static ir_node *default_remat(ir_node *node, ir_node *after)
 {
-	ir_node *copy  = exact_copy(node);
+	ir_node *block, *copy;
+	if (is_Block(after))
+		block = after;
+	else
+		block = get_nodes_block(after);
+
+	copy = exact_copy(node);
+	set_nodes_block(copy, block);
 	sched_add_after(after, copy);
 
 	return copy;
 }
 
+static bool default_check_modifies(const ir_node *node)
+{
+	return arch_irn_is(node, modify_flags);
+}
+
 /**
- * tests wether we can legally move node node after node after
+ * tests whether we can legally move node node after node after
  * (only works for nodes in same block)
  */
-static int can_move(ir_node *node, ir_node *after)
+static bool can_move(ir_node *node, ir_node *after)
 {
 	const ir_edge_t *edge;
-	assert(get_nodes_block(node) == get_nodes_block(after));
+	ir_node *node_block = get_nodes_block(node);
+	assert(node_block == get_nodes_block(after));
 
 	/* TODO respect dep edges */
 	assert(get_irn_n_edges_kind(node, EDGE_KIND_DEP) == 0);
@@ -68,30 +98,45 @@ static int can_move(ir_node *node, ir_node *after)
 	/** all users have to be after the after node */
 	foreach_out_edge(node, edge) {
 		ir_node *out = get_edge_src_irn(edge);
-		if(is_Proj(out)) {
+		if (is_Proj(out)) {
 			const ir_edge_t *edge2;
 			assert(get_irn_n_edges_kind(out, EDGE_KIND_DEP) == 0);
 			foreach_out_edge(out, edge2) {
 				ir_node *out2 = get_edge_src_irn(edge2);
-				/* phi represents a usage at block end */
-				if(is_Phi(out2))
+				if (get_nodes_block(out2) != node_block)
 					continue;
-				if(sched_get_time_step(out2) <= sched_get_time_step(after)) {
-					return 0;
+				/* Phi or End represents a usage at block end. */
+				if (is_Phi(out2) || is_End(out2))
+					continue;
+				if (is_Sync(out2)) {
+					const ir_edge_t *edge3;
+					foreach_out_edge(out2, edge3) {
+						ir_node *out3 = get_edge_src_irn(edge3);
+						/* Phi or End represents a usage at block end. */
+						if (is_Phi(out3) || is_End(out3))
+							continue;
+						assert(!is_Sync(out3));
+						if (sched_get_time_step(out3) <= sched_get_time_step(after)) {
+							return false;
+						}
+					}
+				} else if (sched_get_time_step(out2) <= sched_get_time_step(after)) {
+					return false;
 				}
 			}
 		} else {
+			if (get_nodes_block(out) != node_block)
+				continue;
 			/* phi represents a usage at block end */
-			if(is_Phi(out))
+			if (is_Phi(out))
 				continue;
-			if(sched_get_time_step(out) <= sched_get_time_step(after)) {
-				return 0;
+			if (sched_get_time_step(out) <= sched_get_time_step(after)) {
+				return false;
 			}
 		}
 	}
 
-	ir_fprintf(stderr, "Can move node %+F after node %+F\n", node, after);
-	return 1;
+	return true;
 }
 
 static void rematerialize_or_move(ir_node *flags_needed, ir_node *node,
@@ -101,20 +146,23 @@ static void rematerialize_or_move(ir_node *flags_needed, ir_node *node,
 	ir_node *copy;
 	ir_node *value;
 
-	if(can_move(flags_needed, node)) {
+	if (!is_Block(node) &&
+			get_nodes_block(flags_needed) == get_nodes_block(node) &&
+			can_move(flags_needed, node)) {
 		/* move it */
 		sched_remove(flags_needed);
 		sched_add_after(node, flags_needed);
+		/* No need to update liveness, because the node stays in the same block */
 		return;
 	}
 
-	copy = remat(flags_needed, node);
+	changed = 1;
+	copy    = remat(flags_needed, node);
 
-	if(get_irn_mode(copy) == mode_T) {
-		ir_node *block = get_nodes_block(copy);
-		ir_mode *mode  = flag_class->mode;
-		value = new_rd_Proj(NULL, current_ir_graph, block,
-							copy, mode, pn);
+	if (get_irn_mode(copy) == mode_T) {
+		ir_mode *mode = flag_class->mode;
+		value = new_rd_Proj(NULL, copy, mode, pn);
+		be_add_missing_keeps_node(copy);
 	} else {
 		value = copy;
 	}
@@ -123,106 +171,145 @@ static void rematerialize_or_move(ir_node *flags_needed, ir_node *node,
 	do {
 		int i;
 		int arity = get_irn_arity(n);
-		for(i = 0; i < arity; ++i) {
+		for (i = 0; i < arity; ++i) {
 			ir_node *in = get_irn_n(n, i);
 			in = skip_Proj(in);
-			if(in == flags_needed) {
+			if (in == flags_needed) {
 				set_irn_n(n, i, value);
 				break;
 			}
 		}
-		n = get_irn_link(n);
-	} while(n != NULL);
+		n = (ir_node*)get_irn_link(n);
+	} while (n != NULL);
+
+	/* No need to introduce the copy, because it only lives in this block, but
+	 * we have to update the liveness of all operands */
+	if (is_Block(node) ||
+			get_nodes_block(node) != get_nodes_block(flags_needed)) {
+		ir_graph *irg = get_irn_irg(node);
+		be_lv_t  *lv  = be_get_irg_liveness(irg);
+		int       i;
+
+		if (lv != NULL) {
+			for (i = get_irn_arity(copy) - 1; i >= 0; --i) {
+				be_liveness_update(lv, get_irn_n(copy, i));
+			}
+		}
+	}
 }
 
+/**
+ * walks up the schedule and makes sure there are no flag-destroying nodes
+ * between a flag-consumer -> flag-producer chain. Fixes problematic situations
+ * by moving and/or rematerialisation of the flag-producers.
+ * (This can be extended in the future to do some register allocation on targets
+ *  like ppc32 where we conceptually have 8 flag registers)
+ */
 static void fix_flags_walker(ir_node *block, void *env)
 {
 	ir_node *node;
 	ir_node *flags_needed   = NULL;
 	ir_node *flag_consumers = NULL;
-	int      pn;
+	int      pn = -1;
 	(void) env;
 
 	sched_foreach_reverse(block, node) {
 		int i, arity;
 		ir_node *new_flags_needed = NULL;
+		ir_node *test;
 
-		if(node == flags_needed) {
+		if (is_Phi(node))
+			break;
+
+		if (node == flags_needed) {
 			/* all ok */
 			flags_needed   = NULL;
 			flag_consumers = NULL;
 		}
 
-		/* test wether node destroys the flags */
-		if(flags_needed != NULL && arch_irn_is(arch_env, node, modify_flags)) {
+		/* test whether node destroys the flags */
+		test = node;
+		if (be_is_Keep(test))
+			test = sched_prev(test);
+
+		if (flags_needed != NULL && check_modify(test)) {
 			/* rematerialize */
 			rematerialize_or_move(flags_needed, node, flag_consumers, pn);
 			flags_needed   = NULL;
 			flag_consumers = NULL;
 		}
 
-		/* test wether the current node needs flags */
+		/* test whether the current node needs flags */
 		arity = get_irn_arity(node);
-		for(i = 0; i < arity; ++i) {
-			//ir_node *in = get_irn_n(node, i);
-			const arch_register_class_t *cls
-				= arch_get_irn_reg_class(arch_env, node, i);
-			if(cls == flag_class) {
+		for (i = 0; i < arity; ++i) {
+			const arch_register_req_t *req
+				= arch_get_irn_register_req_in(node, i);
+			if (req->cls == flag_class) {
 				assert(new_flags_needed == NULL);
 				new_flags_needed = get_irn_n(node, i);
 			}
 		}
 
-		if(new_flags_needed == NULL)
+		if (new_flags_needed == NULL)
 			continue;
 
-		if(new_flags_needed != flags_needed) {
-			if(flags_needed != NULL) {
+		/* spiller can't (correctly) remat flag consumers at the moment */
+		assert(!arch_irn_is(node, rematerializable));
+
+		if (skip_Proj(new_flags_needed) != flags_needed) {
+			if (flags_needed != NULL) {
 				/* rematerialize node */
 				rematerialize_or_move(flags_needed, node, flag_consumers, pn);
-				flags_needed    = NULL;
+				flags_needed   = NULL;
 				flag_consumers = NULL;
 			}
 
-			if(get_nodes_block(new_flags_needed) != block) {
-				panic("remat across blocks not implemented yet");
-				flags_needed   = NULL;
-				flag_consumers = NULL;
-			} else {
-				flags_needed = new_flags_needed;
-				arch_set_irn_register(arch_env, flags_needed, flags_reg);
-				if(is_Proj(flags_needed)) {
-					pn           = get_Proj_proj(flags_needed);
-					flags_needed = get_Proj_pred(flags_needed);
-				}
-				flag_consumers = node;
-				set_irn_link(flag_consumers, NULL);
-				assert(arch_irn_is(arch_env, flags_needed, rematerializable));
+			flags_needed = new_flags_needed;
+			arch_set_irn_register(flags_needed, flags_reg);
+			if (is_Proj(flags_needed)) {
+				pn           = get_Proj_proj(flags_needed);
+				flags_needed = get_Proj_pred(flags_needed);
 			}
+			flag_consumers = node;
+			set_irn_link(flag_consumers, NULL);
+			assert(arch_irn_is(flags_needed, rematerializable));
 		} else {
 			/* link all consumers in a list */
-			set_irn_link(flag_consumers, node);
+			set_irn_link(node, flag_consumers);
 			flag_consumers = node;
 		}
 	}
 
+	if (flags_needed != NULL) {
+		assert(get_nodes_block(flags_needed) != block);
+		rematerialize_or_move(flags_needed, node, flag_consumers, pn);
+		flags_needed   = NULL;
+		flag_consumers = NULL;
+	}
+
 	assert(flags_needed   == NULL);
 	assert(flag_consumers == NULL);
 }
 
-void be_sched_fix_flags(be_irg_t *birg, const arch_register_class_t *flag_cls,
-                        func_rematerialize remat_func)
+void be_sched_fix_flags(ir_graph *irg, const arch_register_class_t *flag_cls,
+                        func_rematerialize remat_func,
+                        check_modifies_flags check_modifies_flags_func)
 {
-	ir_graph *irg = be_get_birg_irg(birg);
-
-	arch_env   = be_get_birg_arch_env(birg);
-	flag_class = flag_cls;
-	flags_reg  = & flag_class->regs[0];
-	remat      = remat_func;
-	if(remat == NULL)
+	flag_class   = flag_cls;
+	flags_reg    = & flag_class->regs[0];
+	remat        = remat_func;
+	check_modify = check_modifies_flags_func;
+	changed      = 0;
+	if (remat == NULL)
 		remat = &default_remat;
+	if (check_modify == NULL)
+		check_modify = &default_check_modifies;
 
-	set_using_irn_link(irg);
+	ir_reserve_resources(irg, IR_RESOURCE_IRN_LINK);
 	irg_block_walk_graph(irg, fix_flags_walker, NULL, NULL);
-	clear_using_irn_link(irg);
+	ir_free_resources(irg, IR_RESOURCE_IRN_LINK);
+
+	if (changed) {
+		be_remove_dead_nodes_from_schedule(irg);
+	}
 }