* compile.c (compile_array, compile_array_):

Divide big array (or hash) literals into several blocks and
  concatetene them.  There was a problem that a big array (hash)
  literal causes SystemStackError exception (stack overflow)
  because VM push all contents of the literal onto VM stack to
  make an array (or hash).  To solve this issue, we make several
  arrays (hashes) and concatenate them to make a big array (hash)
  object.
  ??
* compile.c (iseq_compile_each, setup_args): use modified
  compile_array.
* vm.c (m_core_hash_from_ary, m_core_hash_merge_ary,
  m_core_hash_merge_ptr): added for above change.
* id.c (Init_id), parse.y: add core method ids.
* bootstraptest/test_literal.rb: add simple tests.
* bootstraptest/test_eval.rb: remove rescue clause to catch
  SystemStackError exception.
* test/ruby/test_literal.rb: add tests to check no stack overflow.



git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@35306 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

ChangeLog

@@ -1,3 +1,30 @@
+Wed Apr 11 17:16:49 2012  Koichi Sasada  <ko1@atdot.net>
+
+	* compile.c (compile_array, compile_array_):
+	  Divide big array (or hash) literals into several blocks and
+	  concatetene them.  There was a problem that a big array (hash)
+          literal causes SystemStackError exception (stack overflow)
+          because VM push all contents of the literal onto VM stack to
+          make an array (or hash).  To solve this issue, we make several
+          arrays (hashes) and concatenate them to make a big array (hash)
+          object.
+	  ??
+
+	* compile.c (iseq_compile_each, setup_args): use modified
+	  compile_array.
+
+	* vm.c (m_core_hash_from_ary, m_core_hash_merge_ary,
+	  m_core_hash_merge_ptr): added for above change.
+
+	* id.c (Init_id), parse.y: add core method ids.
+
+	* bootstraptest/test_literal.rb: add simple tests.
+
+	* bootstraptest/test_eval.rb: remove rescue clause to catch
+	  SystemStackError exception.
+
+	* test/ruby/test_literal.rb: add tests to check no stack overflow.
+
 Thu Apr 12 07:10:37 2012  Eric Hodel  <drbrain@segment7.net>
 
 	* lib/uri/generic.rb (module URI):  URI now downcases the scheme to

bootstraptest/test_eval.rb

@@ -316,6 +316,5 @@ assert_normal_exit %q{
   end
   begin
     eval "class C; @@h = #{hash.inspect}; end"
-  rescue SystemStackError
   end
 }, '[ruby-core:25714]'

bootstraptest/test_literal.rb

@@ -200,3 +200,25 @@ assert_equal 'ok', %q{
 assert_equal 'ok', %q{
   "#{}o""#{}k""#{}"
 }, '[ruby-core:25284]'
+
+assert_equal 'ok', %q{ #  long array literal
+  x = nil
+  eval "a = [#{(1..10_000).map{'x'}.join(", ")}]"
+  :ok
+}
+
+assert_equal 'ok', %q{ #  long array literal (optimized)
+  eval "a = [#{(1..10_000).to_a.join(", ")}]"
+  :ok
+}
+
+assert_equal 'ok', %q{ #  long hash literal
+  x = nil
+  eval "a = {#{(1..10_000).map{|n| "#{n} => x"}.join(', ')}}"
+  :ok
+}
+
+assert_equal 'ok', %q{ #  long hash literal (optimized)
+  eval "a = {#{(1..10_000).map{|n| "#{n} => #{n}"}.join(', ')}}"
+  :ok
+}

compile.c

@@ -2274,65 +2274,139 @@ compile_branch_condition(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE * cond,
     return COMPILE_OK;
 }
 
+enum compile_array_type_t {
+    COMPILE_ARRAY_TYPE_ARRAY,
+    COMPILE_ARRAY_TYPE_HASH,
+    COMPILE_ARRAY_TYPE_ARGS,
+};
+
 static int
 compile_array_(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE* node_root,
-	       VALUE opt_p, int poped)
+	       enum compile_array_type_t type, int poped)
 {
     NODE *node = node_root;
-    int len = (int)node->nd_alen, line = (int)nd_line(node), i=0;
+    int line = (int)nd_line(node);
-    DECL_ANCHOR(anchor);
+    int len = 0;
 
-    INIT_ANCHOR(anchor);
+    if (nd_type(node) == NODE_ZARRAY) {
-    if (nd_type(node) != NODE_ZARRAY) {
-	while (node) {
-	    if (nd_type(node) != NODE_ARRAY) {
-		rb_bug("compile_array: This node is not NODE_ARRAY, but %s",
-		       ruby_node_name(nd_type(node)));
-	    }
-
-	    i++;
-	    if (opt_p && nd_type(node->nd_head) != NODE_LIT) {
-		opt_p = Qfalse;
-	    }
-	    COMPILE_(anchor, "array element", node->nd_head, poped);
-	    node = node->nd_next;
-	}
-    }
-
-    if (len != i) {
-	if (0) {
-	    rb_bug("node error: compile_array (%d: %d-%d)",
-		   (int)nd_line(node_root), len, i);
-	}
-	len = i;
-    }
-
-    if (opt_p == Qtrue) {
 	if (!poped) {
-	    VALUE ary = rb_ary_tmp_new(len);
+	    switch (type) {
-	    node = node_root;
+	      case COMPILE_ARRAY_TYPE_ARRAY: ADD_INSN1(ret, line, newarray, INT2FIX(0)); break;
-	    while (node) {
+	      case COMPILE_ARRAY_TYPE_HASH: ADD_INSN1(ret, line, newhash, INT2FIX(0)); break;
-		rb_ary_push(ary, node->nd_head->nd_lit);
+	      case COMPILE_ARRAY_TYPE_ARGS: /* do nothing */ break;
-		node = node->nd_next;
 	    }
-	    OBJ_FREEZE(ary);
-	    iseq_add_mark_object_compile_time(iseq, ary);
-	    ADD_INSN1(ret, nd_line(node_root), duparray, ary);
 	}
     }
     else {
-	if (!poped) {
+	int opt_p = 1;
-	    ADD_INSN1(anchor, line, newarray, INT2FIX(len));
+	int first = 1, i;
+
+	while (node) {
+	    NODE *start_node = node, *end_node;
+	    const int max = 0x100;
+	    DECL_ANCHOR(anchor);
+	    INIT_ANCHOR(anchor);
+
+	    for (i=0; i<max && node; i++, len++) {
+		if (CPDEBUG > 0 && nd_type(node) != NODE_ARRAY) {
+		    rb_bug("compile_array: This node is not NODE_ARRAY, but %s", ruby_node_name(nd_type(node)));
+		}
+
+		if (opt_p && nd_type(node->nd_head) != NODE_LIT) {
+		    opt_p = 0;
+		}
+
+		COMPILE_(anchor, "array element", node->nd_head, poped);
+		node = node->nd_next;
+	    }
+
+	    if (opt_p && type != COMPILE_ARRAY_TYPE_ARGS) {
+		if (!poped) {
+		    VALUE ary = rb_ary_tmp_new(i);
+
+		    end_node = node;
+		    node = start_node;
+
+		    while (node != end_node) {
+			rb_ary_push(ary, node->nd_head->nd_lit);
+			node = node->nd_next;
+		    }
+		    while (node && nd_type(node->nd_head) == NODE_LIT) {
+			rb_ary_push(ary, node->nd_head->nd_lit);
+			node = node->nd_next;
+			len++;
+		    }
+
+		    OBJ_FREEZE(ary);
+
+		    iseq_add_mark_object_compile_time(iseq, ary);
+
+		    if (first) {
+			first = 0;
+			if (type == COMPILE_ARRAY_TYPE_ARRAY) {
+			    ADD_INSN1(ret, line, duparray, ary);
+			}
+			else { /* COMPILE_ARRAY_TYPE_HASH */
+			    ADD_INSN1(ret, line, putspecialobject, INT2FIX(VM_SPECIAL_OBJECT_VMCORE));
+			    ADD_INSN1(ret, line, putobject, ary);
+			    ADD_SEND(ret, line, ID2SYM(id_core_hash_from_ary), INT2FIX(1));
+			}
+		    }
+		    else {
+			if (type == COMPILE_ARRAY_TYPE_ARRAY) {
+			    ADD_INSN1(ret, line, putobject, ary);
+			    ADD_INSN(ret, line, concatarray);
+			}
+			else {
+			    ADD_INSN1(ret, line, putspecialobject, INT2FIX(VM_SPECIAL_OBJECT_VMCORE));
+			    ADD_INSN1(ret, line, putobject, ary);
+			    ADD_SEND(ret, line, ID2SYM(id_core_hash_merge_ary), INT2FIX(1));
+			}
+		    }
+		}
+	    }
+	    else {
+		if (!poped) {
+		    switch (type) {
+		      case COMPILE_ARRAY_TYPE_ARRAY:
+			ADD_INSN1(anchor, line, newarray, INT2FIX(i));
+
+			if (first) {
+			    first = 0;
+			}
+			else {
+			    ADD_INSN(anchor, line, concatarray);
+			}
+			APPEND_LIST(ret, anchor);
+			break;
+		      case COMPILE_ARRAY_TYPE_HASH:
+			if (first) {
+			    first = 0;
+			    ADD_INSN1(anchor, line, newhash, INT2FIX(i));
+			    APPEND_LIST(ret, anchor);
+			}
+			else {
+			    ADD_INSN1(ret, line, putspecialobject, INT2FIX(VM_SPECIAL_OBJECT_VMCORE));
+			    ADD_INSN(ret, line, swap);
+			    APPEND_LIST(ret, anchor);
+			    ADD_SEND(ret, line, ID2SYM(id_core_hash_merge_ptr), INT2FIX(i + 1));
+			}
+			break;
+		      case COMPILE_ARRAY_TYPE_ARGS:
+			APPEND_LIST(ret, anchor);
+			break;
+		    }
+		}
+	    }
 	}
-	APPEND_LIST(ret, anchor);
     }
     return len;
 }
 
 static VALUE
-compile_array(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE* node_root, VALUE opt_p)
+compile_array(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE* node_root, enum compile_array_type_t type)
 {
-    return compile_array_(iseq, ret, node_root, opt_p, 0);
+    return compile_array_(iseq, ret, node_root, type, 0);
 }
 
 static VALUE
@@ -2963,8 +3037,7 @@ setup_args(rb_iseq_t *iseq, LINK_ANCHOR *args, NODE *argn, VALUE *flag)
 	    *flag |= VM_CALL_ARGS_SPLAT_BIT;
 
 	    if (next_is_array) {
-		argc = INT2FIX(compile_array(iseq, args, argn->nd_head, Qfalse) + 1);
+		argc = INT2FIX(compile_array(iseq, args, argn->nd_head, COMPILE_ARRAY_TYPE_ARGS) + 1);
-		POP_ELEMENT(args);
 	    }
 	    else {
 		argn = argn->nd_head;
@@ -2973,8 +3046,7 @@ setup_args(rb_iseq_t *iseq, LINK_ANCHOR *args, NODE *argn, VALUE *flag)
 	    break;
 	  }
 	  case NODE_ARRAY: {
-	    argc = INT2FIX(compile_array(iseq, args, argn, Qfalse));
+	    argc = INT2FIX(compile_array(iseq, args, argn, COMPILE_ARRAY_TYPE_ARGS));
-	    POP_ELEMENT(args);
 	    break;
 	  }
 	  default: {
@@ -4265,7 +4337,7 @@ iseq_compile_each(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE * node, int poped)
 	break;
       }
       case NODE_ARRAY:{
-	compile_array_(iseq, ret, node, Qtrue, poped);
+	compile_array_(iseq, ret, node, COMPILE_ARRAY_TYPE_ARRAY, poped);
 	break;
       }
       case NODE_ZARRAY:{
@@ -4293,22 +4365,19 @@ iseq_compile_each(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE * node, int poped)
 
 	INIT_ANCHOR(list);
 	switch (type) {
-	  case NODE_ARRAY:{
+	  case NODE_ARRAY:
-	    compile_array(iseq, list, node->nd_head, Qfalse);
+	    size = INT2FIX(compile_array(iseq, list, node->nd_head, COMPILE_ARRAY_TYPE_HASH));
-	    size = OPERAND_AT(POP_ELEMENT(list), 0);
 	    ADD_SEQ(ret, list);
 	    break;
-	  }
+
 	  case NODE_ZARRAY:
-	    size = INT2FIX(0);
+	    ADD_INSN1(ret, nd_line(node), newhash, INT2FIX(0));
 	    break;
 
 	  default:
 	    rb_bug("can't make hash with this node: %s", ruby_node_name(type));
 	}
 
-	ADD_INSN1(ret, nd_line(node), newhash, size);
-
 	if (poped) {
 	    ADD_INSN(ret, nd_line(node), pop);
 	}

id.c

@@ -31,6 +31,9 @@ Init_id(void)
     REGISTER_SYMID(id_core_define_method,      	    "core#define_method");
     REGISTER_SYMID(id_core_define_singleton_method, "core#define_singleton_method");
     REGISTER_SYMID(id_core_set_postexe,             "core#set_postexe");
+    REGISTER_SYMID(id_core_hash_from_ary,           "core#hash_from_ary");
+    REGISTER_SYMID(id_core_hash_merge_ary,          "core#hash_merge_ary");
+    REGISTER_SYMID(id_core_hash_merge_ptr,          "core#hash_merge_ptr");
 
     REGISTER_SYMID(idEach, "each");
     REGISTER_SYMID(idLength, "length");

parse.y

@@ -802,6 +802,9 @@ static void token_info_pop(struct parser_params*, const char *token);
 %nonassoc id_core_define_method
 %nonassoc id_core_define_singleton_method
 %nonassoc id_core_set_postexe
+%nonassoc id_core_hash_from_ary
+%nonassoc id_core_hash_merge_ary
+%nonassoc id_core_hash_merge_ptr
 
 %token tLAST_TOKEN
 

test/ruby/test_literal.rb

@@ -1,4 +1,5 @@
 require 'test/unit'
+require_relative 'envutil'
 
 class TestRubyLiteral < Test::Unit::TestCase
 
@@ -186,6 +187,13 @@ class TestRubyLiteral < Test::Unit::TestCase
     assert_equal "literal", h["string"]
   end
 
+  def test_big_array_and_hash_literal
+    assert_normal_exit %q{x = nil; raise if eval("[#{(1..1_000_000).map{'x'}.join(", ")}]").size != 1_000_000}, "", timeout: 300
+    assert_normal_exit %q{x = nil; raise if eval("[#{(1..1_000_000).to_a.join(", ")}]").size != 1_000_000}, "", timeout: 300
+    assert_normal_exit %q{x = nil; raise if eval("{#{(1..1_000_000).map{|n| "#{n} => x"}.join(', ')}}").size != 1_000_000}, "", timeout: 300
+    assert_normal_exit %q{x = nil; raise if eval("{#{(1..1_000_000).map{|n| "#{n} => #{n}"}.join(', ')}}").size != 1_000_000}, "", timeout: 300
+  end
+
   def test_range
     assert_instance_of Range, (1..2)
     assert_equal(1..2, 1..2)

vm.c

@@ -2030,6 +2030,44 @@ m_core_set_postexe(VALUE self, VALUE iseqval)
     return Qnil;
 }
 
+static VALUE
+m_core_hash_from_ary(VALUE self, VALUE ary)
+{
+    VALUE hash = rb_hash_new();
+    int i;
+
+    for (i=0; i<RARRAY_LEN(ary); i+=2) {
+	rb_hash_aset(hash, RARRAY_PTR(ary)[i], RARRAY_PTR(ary)[i+1]);
+    }
+
+    return hash;
+}
+
+static VALUE
+m_core_hash_merge_ary(VALUE self, VALUE hash, VALUE ary)
+{
+    int i;
+
+    for (i=0; i<RARRAY_LEN(ary); i+=2) {
+	rb_hash_aset(hash, RARRAY_PTR(ary)[i], RARRAY_PTR(ary)[i+1]);
+    }
+
+    return hash;
+}
+
+static VALUE
+m_core_hash_merge_ptr(int argc, VALUE *argv, VALUE recv)
+{
+    int i;
+    VALUE hash = argv[0];
+
+    for (i=1; i<argc; i+=2) {
+	rb_hash_aset(hash, argv[i], argv[i+1]);
+    }
+
+    return hash;
+}
+
 extern VALUE *rb_gc_stack_start;
 extern size_t rb_gc_stack_maxsize;
 #ifdef __ia64
@@ -2093,6 +2131,9 @@ Init_VM(void)
     rb_define_method_id(klass, id_core_define_method, m_core_define_method, 3);
     rb_define_method_id(klass, id_core_define_singleton_method, m_core_define_singleton_method, 3);
     rb_define_method_id(klass, id_core_set_postexe, m_core_set_postexe, 1);
+    rb_define_method_id(klass, id_core_hash_from_ary, m_core_hash_from_ary, 1);
+    rb_define_method_id(klass, id_core_hash_merge_ary, m_core_hash_merge_ary, 2);
+    rb_define_method_id(klass, id_core_hash_merge_ptr, m_core_hash_merge_ptr, -1);
     rb_obj_freeze(fcore);
     rb_gc_register_mark_object(fcore);
     rb_mRubyVMFrozenCore = fcore;