* parse.y: support Symbol GC. [ruby-trunk Feature #9634]

  See this ticket about Symbol GC.

* include/ruby/ruby.h:
  Declare few functions.
  * rb_sym2id: almost same as old SYM2ID but support dynamic symbols.
  * rb_id2sym: almost same as old ID2SYM but support dynamic symbols.
  * rb_sym2str: almost same as `rb_id2str(SYM2ID(sym))` but not
    pin down a dynamic symbol.
  Declare a new struct.
  * struct RSymbol: represents a dynamic symbol as object in
    Ruby's heaps.
  Add few macros.
  * STATIC_SYM_P: check a static symbol.
  * DYNAMIC_SYM_P: check a dynamic symbol.
  * RSYMBOL: cast to RSymbol

* gc.c: declare RSymbol. support T_SYMBOL.

* internal.h: Declare few functions.
  * rb_gc_free_dsymbol: free up a dynamic symbol. GC call this
    function at a sweep phase.
  * rb_str_dynamic_intern: convert a string to a dynamic symbol.
  * rb_check_id_without_pindown: not pinning function.
  * rb_sym2id_without_pindown: ditto.
  * rb_check_id_cstr_without_pindown: ditto.

* string.c (Init_String): String#intern and String#to_sym use
  rb_str_dynamic_intern.

* template/id.h.tmpl: use LSB of ID as a flag for determining a
  static symbol, so we shift left other ruby_id_types.

* string.c: use rb_sym2str instead `rb_id2str(SYM2ID(sym))` to
  avoid pinning.

* load.c: use xx_without_pindown function at creating temporary ID
  to avoid pinning.

* object.c: ditto.

* sprintf.c: ditto.

* struct.c: ditto.

* thread.c: ditto.

* variable.c: ditto.

* vm_method.c: ditto.

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@45426 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

ChangeLog

@@ -1,3 +1,56 @@
+Tue Mar 25 22:57:11 2014  Narihiro Nakamura  <authornari@gmail.com>
+
+	* parse.y: support Symbol GC. [ruby-trunk Feature #9634]
+	  See this ticket about Symbol GC.
+
+	* include/ruby/ruby.h:
+	  Declare few functions.
+	  * rb_sym2id: almost same as old SYM2ID but support dynamic symbols.
+	  * rb_id2sym: almost same as old ID2SYM but support dynamic symbols.
+	  * rb_sym2str: almost same as `rb_id2str(SYM2ID(sym))` but not
+	    pin down a dynamic symbol.
+	  Declare a new struct.
+	  * struct RSymbol: represents a dynamic symbol as object in
+	    Ruby's heaps.
+	  Add few macros.
+	  * STATIC_SYM_P: check a static symbol.
+	  * DYNAMIC_SYM_P: check a dynamic symbol.
+	  * RSYMBOL: cast to RSymbol
+
+	* gc.c: declare RSymbol. support T_SYMBOL.
+
+	* internal.h: Declare few functions.
+	  * rb_gc_free_dsymbol: free up a dynamic symbol. GC call this
+	    function at a sweep phase.
+	  * rb_str_dynamic_intern: convert a string to a dynamic symbol.
+	  * rb_check_id_without_pindown: not pinning function.
+	  * rb_sym2id_without_pindown: ditto.
+	  * rb_check_id_cstr_without_pindown: ditto.
+
+	* string.c (Init_String): String#intern and String#to_sym use
+	  rb_str_dynamic_intern.
+
+	* template/id.h.tmpl: use LSB of ID as a flag for determining a
+	  static symbol, so we shift left other ruby_id_types.
+
+	* string.c: use rb_sym2str instead `rb_id2str(SYM2ID(sym))` to
+	  avoid pinning.
+
+	* load.c: use xx_without_pindown function at creating temporary ID
+	  to avoid pinning.
+
+	* object.c: ditto.
+
+	* sprintf.c: ditto.
+
+	* struct.c: ditto.
+
+	* thread.c: ditto.
+
+	* variable.c: ditto.
+
+	* vm_method.c: ditto.
+
 Wed Mar 26 13:25:54 2014  NARUSE, Yui  <naruse@ruby-lang.org>
 
 	* addr2line.c (fill_lines): loop reverse order not to overwrite

gc.c

@@ -354,6 +354,7 @@ typedef struct RVALUE {
 	struct RMatch  match;
 	struct RRational rational;
 	struct RComplex complex;
+	struct RSymbol symbol;
 	struct {
 	    struct RBasic basic;
 	    VALUE v1;
@@ -1652,6 +1653,12 @@ obj_free(rb_objspace_t *objspace, VALUE obj)
 	}
 	break;
 
+      case T_SYMBOL:
+	{
+            rb_gc_free_dsymbol(obj);
+	}
+	break;
+
       default:
 	rb_bug("gc_sweep(): unknown data type 0x%x(%p) 0x%"PRIxVALUE,
 	       BUILTIN_TYPE(obj), (void*)obj, RBASIC(obj)->flags);
@@ -2393,7 +2400,7 @@ rb_obj_id(VALUE obj)
      *  24 if 32-bit, double is 8-byte aligned
      *  40 if 64-bit
      */
-    if (SYMBOL_P(obj)) {
+    if (STATIC_SYM_P(obj)) {
         return (SYM2ID(obj) * sizeof(RVALUE) + (4 << 2)) | FIXNUM_FLAG;
     }
     else if (FLONUM_P(obj)) {
@@ -2499,6 +2506,7 @@ obj_memsize_of(VALUE obj, int use_tdata)
 	break;
 
       case T_FLOAT:
+      case T_SYMBOL:
 	break;
 
       case T_BIGNUM:
@@ -3918,6 +3926,7 @@ gc_mark_children(rb_objspace_t *objspace, VALUE ptr)
 
       case T_FLOAT:
       case T_BIGNUM:
+      case T_SYMBOL:
 	break;
 
       case T_MATCH:

include/ruby/ruby.h

@@ -351,9 +351,13 @@ rb_long2int_inline(long n)
 
 #define IMMEDIATE_P(x) ((VALUE)(x) & IMMEDIATE_MASK)
 
-#define SYMBOL_P(x) (((VALUE)(x)&~((~(VALUE)0)<<RUBY_SPECIAL_SHIFT))==SYMBOL_FLAG)
-#define ID2SYM(x) (((VALUE)(x)<<RUBY_SPECIAL_SHIFT)|SYMBOL_FLAG)
-#define SYM2ID(x) RSHIFT((unsigned long)(x),RUBY_SPECIAL_SHIFT)
+ID rb_sym2id(VALUE);
+VALUE rb_id2sym(ID);
+#define STATIC_SYM_P(x) (((VALUE)(x)&~((~(VALUE)0)<<RUBY_SPECIAL_SHIFT))==SYMBOL_FLAG)
+#define DYNAMIC_SYM_P(x) (!SPECIAL_CONST_P(x) && BUILTIN_TYPE(x) == (T_SYMBOL))
+#define SYMBOL_P(x) (STATIC_SYM_P(x)||DYNAMIC_SYM_P(x))
+#define ID2SYM(x) (rb_id2sym(x))
+#define SYM2ID(x) (rb_sym2id(x))
 
 #ifndef USE_FLONUM
 #if SIZEOF_VALUE >= SIZEOF_DOUBLE
@@ -957,6 +961,12 @@ struct RComplex {
 #define RCOMPLEX_SET_REAL(cmp, r) RB_OBJ_WRITE((cmp), &((struct RComplex *)(cmp))->real,(r))
 #define RCOMPLEX_SET_IMAG(cmp, i) RB_OBJ_WRITE((cmp), &((struct RComplex *)(cmp))->imag,(i))
 
+struct RSymbol {
+    struct RBasic basic;
+    VALUE fstr;
+    ID type;
+};
+
 struct RData {
     struct RBasic basic;
     void (*dmark)(void*);
@@ -1093,6 +1103,7 @@ struct RStruct {
 #define RFILE(obj)   (R_CAST(RFile)(obj))
 #define RRATIONAL(obj) (R_CAST(RRational)(obj))
 #define RCOMPLEX(obj) (R_CAST(RComplex)(obj))
+#define RSYMBOL(obj) (R_CAST(RSymbol)(obj))
 
 #define FL_SINGLETON FL_USER0
 #define FL_WB_PROTECTED (((VALUE)1)<<5)
@@ -1146,7 +1157,7 @@ struct RStruct {
     (OBJ_TAINTABLE(x) && FL_ABLE(s)) ? \
     RBASIC(x)->flags |= RBASIC(s)->flags & FL_TAINT : 0)
 
-#define OBJ_FROZEN(x) (!!(FL_ABLE(x)?(RBASIC(x)->flags&(FL_FREEZE)):(FIXNUM_P(x)||FLONUM_P(x)||SYMBOL_P(x))))
+#define OBJ_FROZEN(x) (!!(FL_ABLE(x)?(RBASIC(x)->flags&(FL_FREEZE)):(FIXNUM_P(x)||FLONUM_P(x)||STATIC_SYM_P(x))))
 #define OBJ_FREEZE(x) FL_SET((x), FL_FREEZE)
 
 #if USE_RGENGC
@@ -1381,6 +1392,7 @@ const char *rb_id2name(ID);
 ID rb_check_id(volatile VALUE *);
 ID rb_to_id(VALUE);
 VALUE rb_id2str(ID);
+VALUE rb_sym2str(VALUE);
 
 #define CONST_ID_CACHE(result, str)			\
     {							\
@@ -1597,7 +1609,7 @@ rb_class_of(VALUE obj)
 	if (FIXNUM_P(obj)) return rb_cFixnum;
 	if (FLONUM_P(obj)) return rb_cFloat;
 	if (obj == Qtrue)  return rb_cTrueClass;
-	if (SYMBOL_P(obj)) return rb_cSymbol;
+	if (STATIC_SYM_P(obj)) return rb_cSymbol;
     }
     else if (!RTEST(obj)) {
 	if (obj == Qnil)   return rb_cNilClass;
@@ -1613,7 +1625,7 @@ rb_type(VALUE obj)
 	if (FIXNUM_P(obj)) return T_FIXNUM;
         if (FLONUM_P(obj)) return T_FLOAT;
         if (obj == Qtrue)  return T_TRUE;
-	if (SYMBOL_P(obj)) return T_SYMBOL;
+	if (STATIC_SYM_P(obj)) return T_SYMBOL;
 	if (obj == Qundef) return T_UNDEF;
     }
     else if (!RTEST(obj)) {

internal.h

@@ -742,6 +742,13 @@ int rb_is_junk_name(VALUE name);
 void rb_gc_mark_parser(void);
 void rb_gc_mark_symbols(int full_mark);
 ID rb_make_internal_id(void);
+void rb_gc_free_dsymbol(VALUE);
+VALUE rb_str_dynamic_intern(VALUE);
+ID rb_check_id_without_pindown(VALUE *);
+ID rb_sym2id_without_pindown(VALUE);
+#ifdef RUBY_ENCODING_H
+ID rb_check_id_cstr_without_pindown(const char *, long, rb_encoding *);
+#endif
 
 /* proc.c */
 VALUE rb_proc_location(VALUE self);

load.c

@@ -1108,7 +1108,7 @@ rb_mod_autoload(VALUE mod, VALUE sym, VALUE file)
 static VALUE
 rb_mod_autoload_p(VALUE mod, VALUE sym)
 {
-    ID id = rb_check_id(&sym);
+    ID id = rb_check_id_without_pindown(&sym);
     if (!id) {
 	return Qnil;
     }

object.c

@@ -2125,7 +2125,7 @@ rb_mod_const_get(int argc, VALUE *argv, VALUE mod)
 
 	if (pbeg == p) goto wrong_name;
 
-	id = rb_check_id_cstr(pbeg, len = p-pbeg, enc);
+	id = rb_check_id_cstr_without_pindown(pbeg, len = p-pbeg, enc);
 	beglen = pbeg-path;
 
 	if (p < pend && p[0] == ':') {
@@ -2267,7 +2267,7 @@ rb_mod_const_defined(int argc, VALUE *argv, VALUE mod)
 
 	if (pbeg == p) goto wrong_name;
 
-	id = rb_check_id_cstr(pbeg, len = p-pbeg, enc);
+	id = rb_check_id_cstr_without_pindown(pbeg, len = p-pbeg, enc);
 	beglen = pbeg-path;
 
 	if (p < pend && p[0] == ':') {
@@ -2338,7 +2338,7 @@ rb_mod_const_defined(int argc, VALUE *argv, VALUE mod)
 static VALUE
 rb_obj_ivar_get(VALUE obj, VALUE iv)
 {
-    ID id = rb_check_id(&iv);
+    ID id = rb_check_id_without_pindown(&iv);
 
     if (!id) {
 	if (rb_is_instance_name(iv)) {
@@ -2409,7 +2409,7 @@ rb_obj_ivar_set(VALUE obj, VALUE iv, VALUE val)
 static VALUE
 rb_obj_ivar_defined(VALUE obj, VALUE iv)
 {
-    ID id = rb_check_id(&iv);
+    ID id = rb_check_id_without_pindown(&iv);
 
     if (!id) {
 	if (rb_is_instance_name(iv)) {
@@ -2446,7 +2446,7 @@ rb_obj_ivar_defined(VALUE obj, VALUE iv)
 static VALUE
 rb_mod_cvar_get(VALUE obj, VALUE iv)
 {
-    ID id = rb_check_id(&iv);
+    ID id = rb_check_id_without_pindown(&iv);
 
     if (!id) {
 	if (rb_is_class_name(iv)) {
@@ -2512,7 +2512,7 @@ rb_mod_cvar_set(VALUE obj, VALUE iv, VALUE val)
 static VALUE
 rb_mod_cvar_defined(VALUE obj, VALUE iv)
 {
-    ID id = rb_check_id(&iv);
+    ID id = rb_check_id_without_pindown(&iv);
 
     if (!id) {
 	if (rb_is_class_name(iv)) {