Split OP_SETVAL into OP_SETVAL (fp expressions and labels) and OP_SYMADDR

(symbol addresses).

They are pretty different. Symbol addresses have special meaning during
various phases, from symbol simplification to CSE.

8 files changed, 54 insertions, 45 deletions

diff --git a/cse.c b/cse.c
index 31ceafd..758884c 100644
--- a/cse.c
+++ b/cse.c
@@ -76,6 +76,9 @@ static void clean_up_one_instruction(struct basic_block *bb, struct instruction
 
 	case OP_SETVAL:
 		hash += hashval(insn->val);
+		break;
+
+	case OP_SYMADDR:
 		hash += hashval(insn->symbol);
 		break;
 
@@ -188,11 +191,14 @@ static int insn_compare(const void *_i1, const void *_i2)
 			return i1->src1 < i2->src1 ? -1 : 1;
 		break;
 
+	case OP_SYMADDR:
+		if (i1->symbol != i2->symbol)
+			return i1->symbol < i2->symbol ? -1 : 1;
+		break;
+
 	case OP_SETVAL:
 		if (i1->val != i2->val)
 			return i1->val < i2->val ? -1 : 1;
-		if (i1->symbol != i2->symbol)
-			return i1->symbol < i2->symbol ? -1 : 1;
 		break;
 
 	/* Other */
diff --git a/example.c b/example.c
index 4716b54..5894fc9 100644
--- a/example.c
+++ b/example.c
@@ -502,7 +502,7 @@ static struct hardreg *fill_reg(struct bb_state *state, struct hardreg *hardreg,
 	case PSEUDO_REG:
 		def = pseudo->def;
 		if (def->opcode == OP_SETVAL) {
-			output_insn(state, "movl $<%s>,%s", show_pseudo(def->symbol), hardreg->name);
+			output_insn(state, "movl $<%s>,%s", show_pseudo(def->target), hardreg->name);
 			break;
 		}
 		src = find_pseudo_storage(state, pseudo, hardreg);
@@ -1298,7 +1298,7 @@ static void fill_output(struct bb_state *state, pseudo_t pseudo, struct storage
 	case PSEUDO_REG:
 		def = pseudo->def;
 		if (def->opcode == OP_SETVAL) {
-			write_val_to_storage(state, def->symbol, out);
+			write_val_to_storage(state, pseudo, out);
 			return;
 		}
 	default:
diff --git a/flow.c b/flow.c
index c523efc..9d2f0fb 100644
--- a/flow.c
+++ b/flow.c
@@ -53,15 +53,10 @@ static int pseudo_truth_value(pseudo_t pseudo)
 
 	case PSEUDO_REG: {
 		struct instruction *insn = pseudo->def;
-		if (insn->opcode == OP_SETVAL && insn->target == pseudo) {
-			struct expression *expr = insn->val;
-
-			/* A symbol address is always considered true.. */
-			if (!expr)
-				return 1;
-			if (expr->type == EXPR_VALUE)
-				return !!expr->value;
-		}
+
+		/* A symbol address is always considered true.. */
+		if (insn->opcode == OP_SYMADDR && insn->target == pseudo)
+			return 1;
 	}
 		/* Fall through */
 	default:
@@ -614,7 +609,7 @@ static void simplify_one_symbol(struct entrypoint *ep, struct symbol *sym)
 			break;
 		case OP_LOAD:
 			break;
-		case OP_SETVAL:
+		case OP_SYMADDR:
 			if (!insn->bb)
 				continue;
 			mod |= MOD_ADDRESSABLE;
diff --git a/linearize.c b/linearize.c
index d2cf224..21eccd5 100644
--- a/linearize.c
+++ b/linearize.c
@@ -205,6 +205,7 @@ static const char* opcodes[] = {
 	[OP_LOAD] = "load",
 	[OP_STORE] = "store",
 	[OP_SETVAL] = "set",
+	[OP_SYMADDR] = "symaddr",
 	[OP_GET_ELEMENT_PTR] = "getelem",
 
 	/* Other */
@@ -289,27 +290,25 @@ const char *show_instruction(struct instruction *insn)
 		buf += sprintf(buf, ".L%p", insn->bb_true ? insn->bb_true : insn->bb_false);
 		break;
 
-	case OP_SETVAL: {
-		struct expression *expr = insn->val;
-		pseudo_t pseudo = insn->symbol;
+	case OP_SYMADDR: {
+		struct symbol *sym = insn->symbol->sym;
 		buf += sprintf(buf, "%s <- ", show_pseudo(insn->target));
-		if (pseudo) {
-			struct symbol *sym = pseudo->sym;
-			if (!sym) {
-				buf += sprintf(buf, "%s", show_pseudo(pseudo));
-				break;
-			}
-			if (sym->bb_target) {
-				buf += sprintf(buf, ".L%p", sym->bb_target);
-				break;
-			}
-			if (sym->ident) {
-				buf += sprintf(buf, "%s", show_ident(sym->ident));
-				break;
-			}
-			buf += sprintf(buf, "<anon symbol:%p>", sym);
+
+		if (sym->bb_target) {
+			buf += sprintf(buf, ".L%p", sym->bb_target);
 			break;
 		}
+		if (sym->ident) {
+			buf += sprintf(buf, "%s", show_ident(sym->ident));
+			break;
+		}
+		buf += sprintf(buf, "<anon symbol:%p>", sym);
+		break;
+	}
+		
+	case OP_SETVAL: {
+		struct expression *expr = insn->val;
+		buf += sprintf(buf, "%s <- ", show_pseudo(insn->target));
 
 		if (!expr) {
 			buf += sprintf(buf, "%s", "<none>");
@@ -942,11 +941,17 @@ static pseudo_t add_setval(struct entrypoint *ep, struct symbol *ctype, struct e
 	pseudo_t target = alloc_pseudo(insn);
 	insn->target = target;
 	insn->val = val;
-	if (!val) {
-		pseudo_t addr = symbol_pseudo(ep, ctype);
-		use_pseudo(addr, &insn->symbol);
-		insn->size = bits_in_pointer;
-	}
+	add_one_insn(ep, insn);
+	return target;
+}
+
+static pseudo_t add_symbol_address(struct entrypoint *ep, struct symbol *sym)
+{
+	struct instruction *insn = alloc_instruction(OP_SYMADDR, bits_in_pointer);
+	pseudo_t target = alloc_pseudo(insn);
+
+	insn->target = target;
+	use_pseudo(symbol_pseudo(ep, sym), &insn->symbol);
 	add_one_insn(ep, insn);
 	return target;
 }
@@ -1442,7 +1447,7 @@ pseudo_t linearize_expression(struct entrypoint *ep, struct expression *expr)
 	switch (expr->type) {
 	case EXPR_SYMBOL:
 		linearize_one_symbol(ep, expr->symbol);
-		return add_setval(ep, expr->symbol, NULL);
+		return add_symbol_address(ep, expr->symbol);
 
 	case EXPR_VALUE:
 		return value_pseudo(expr->value);
diff --git a/linearize.h b/linearize.h
index 401e665..4077542 100644
--- a/linearize.h
+++ b/linearize.h
@@ -173,6 +173,7 @@ enum opcode {
 	OP_LOAD,
 	OP_STORE,
 	OP_SETVAL,
+	OP_SYMADDR,
 	OP_GET_ELEMENT_PTR,
 
 	/* Other */
diff --git a/liveness.c b/liveness.c
index ecfc7fd..d21b3f1 100644
--- a/liveness.c
+++ b/liveness.c
@@ -93,6 +93,10 @@ static void track_instruction_usage(struct basic_block *bb, struct instruction *
 		break;
 
 	case OP_SETVAL:
+		DEFINES(target);
+		break;
+
+	case OP_SYMADDR:
 		USES(symbol); DEFINES(target);
 		break;
 
diff --git a/memops.c b/memops.c
index 7246ca5..daae061 100644
--- a/memops.c
+++ b/memops.c
@@ -69,7 +69,7 @@ static inline int address_taken(pseudo_t pseudo)
 	pseudo_t *usep;
 	FOR_EACH_PTR(pseudo->users, usep) {
 		struct instruction *insn = container(usep, struct instruction, src);
-		if (insn->bb && insn->opcode == OP_SETVAL)
+		if (insn->bb && insn->opcode == OP_SYMADDR)
 			return 1;
 	} END_FOR_EACH_PTR(usep);
 	return 0;
diff --git a/simplify.c b/simplify.c
index b2bb8d7..8392241 100644
--- a/simplify.c
+++ b/simplify.c
@@ -188,11 +188,9 @@ void kill_instruction(struct instruction *insn)
 		repeat_phase |= REPEAT_CSE;
 		return;
 
-	 case OP_SETVAL:
+	 case OP_SYMADDR:
 	 	insn->bb = NULL;
-	 	repeat_phase |= REPEAT_CSE;
-	 	if (insn->symbol)
-	 		repeat_phase |= REPEAT_SYMBOL_CLEANUP;
+	 	repeat_phase |= REPEAT_CSE | REPEAT_SYMBOL_CLEANUP;
 	 	return;
 	}
 }
@@ -485,7 +483,7 @@ static int simplify_one_memop(struct instruction *insn, pseudo_t orig)
 
 	if (addr->type == PSEUDO_REG) {
 		struct instruction *def = addr->def;
-		if (def->opcode == OP_SETVAL && def->src) {
+		if (def->opcode == OP_SYMADDR && def->src) {
 			kill_use(&insn->src);
 			use_pseudo(def->src, &insn->src);
 			return REPEAT_CSE | REPEAT_SYMBOL_CLEANUP;
@@ -736,7 +734,7 @@ int simplify_instruction(struct instruction *insn)
 		return simplify_unop(insn);
 	case OP_LOAD: case OP_STORE:
 		return simplify_memop(insn);
-	case OP_SETVAL:
+	case OP_SYMADDR:
 		if (dead_insn(insn, NULL, NULL, NULL))
 			return REPEAT_CSE | REPEAT_SYMBOL_CLEANUP;
 		return replace_with_pseudo(insn, insn->symbol);