ruby/shape.c

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

1289 строки
37 KiB
C
Исходник Обычный вид История

#include "vm_core.h"
#include "vm_sync.h"
#include "shape.h"
#include "symbol.h"
#include "id_table.h"
#include "internal/class.h"
#include "internal/gc.h"
#include "internal/symbol.h"
#include "internal/variable.h"
#include "internal/error.h"
#include "variable.h"
#include <stdbool.h>
#ifndef _WIN32
#include <sys/mman.h>
#endif
#ifndef SHAPE_DEBUG
#define SHAPE_DEBUG (VM_CHECK_MODE > 0)
#endif
#if SIZEOF_SHAPE_T == 4
#if RUBY_DEBUG
#define SHAPE_BUFFER_SIZE 0x8000
#else
#define SHAPE_BUFFER_SIZE 0x80000
#endif
#else
#define SHAPE_BUFFER_SIZE 0x8000
#endif
#define REDBLACK_CACHE_SIZE (SHAPE_BUFFER_SIZE * 32)
#define SINGLE_CHILD_TAG 0x1
#define TAG_SINGLE_CHILD(x) (struct rb_id_table *)((uintptr_t)x | SINGLE_CHILD_TAG)
#define SINGLE_CHILD_MASK (~((uintptr_t)SINGLE_CHILD_TAG))
#define SINGLE_CHILD_P(x) (((uintptr_t)x) & SINGLE_CHILD_TAG)
#define SINGLE_CHILD(x) (rb_shape_t *)((uintptr_t)x & SINGLE_CHILD_MASK)
#define ANCESTOR_CACHE_THRESHOLD 10
#define MAX_SHAPE_ID (SHAPE_BUFFER_SIZE - 1)
vm_getivar: assume the cached shape_id like have a common ancestor When an inline cache misses, it is very likely that the stale shape_id and the current instance shape_id have a close common ancestor. For example if the instance variable is sometimes frozen sometimes not, one of the two shape will be the direct parent of the other. Another pattern that commonly cause IC misses is "memoization", in such case the object will have a "base common shape" and then a number of close descendants. In addition, when we find a common ancestor, we store it in the inline cache instead of the current shape. This help prevent the cache from flip-flopping, ensuring the next lookup will be marginally faster and more generally avoid writing in memory too much. However, now that shapes have an ancestors index, we only check for a few ancestors before falling back to use the index. So overall this change speeds up what is assumed to be the more common case, but makes what is assumed to be the less common case a bit slower. ``` compare-ruby: ruby 3.3.0dev (2023-10-26T05:30:17Z master 701ca070b4) [arm64-darwin22] built-ruby: ruby 3.3.0dev (2023-10-26T09:25:09Z shapes_double_sear.. a723a85235) [arm64-darwin22] warming up...... | |compare-ruby|built-ruby| |:------------------------------------|-----------:|---------:| |vm_ivar_stable_shape | 11.672M| 11.679M| | | -| 1.00x| |vm_ivar_memoize_unstable_shape | 7.551M| 10.506M| | | -| 1.39x| |vm_ivar_memoize_unstable_shape_miss | 11.591M| 11.624M| | | -| 1.00x| |vm_ivar_unstable_undef | 9.037M| 7.981M| | | 1.13x| -| |vm_ivar_divergent_shape | 8.034M| 6.657M| | | 1.21x| -| |vm_ivar_divergent_shape_imbalanced | 10.471M| 9.231M| | | 1.13x| -| ``` Co-Authored-By: John Hawthorn <john@hawthorn.email>
2023-10-26 12:08:05 +03:00
#define ANCESTOR_SEARCH_MAX_DEPTH 2
static ID id_frozen;
static ID id_t_object;
static ID size_pool_edge_names[SIZE_POOL_COUNT];
#define LEAF 0
#define BLACK 0x0
#define RED 0x1
static redblack_node_t *
redblack_left(redblack_node_t * node)
{
if (node->l == LEAF) {
return LEAF;
}
else {
RUBY_ASSERT(node->l < GET_SHAPE_TREE()->cache_size);
redblack_node_t * left = &GET_SHAPE_TREE()->shape_cache[node->l - 1];
return left;
}
}
static redblack_node_t *
redblack_right(redblack_node_t * node)
{
if (node->r == LEAF) {
return LEAF;
}
else {
RUBY_ASSERT(node->r < GET_SHAPE_TREE()->cache_size);
redblack_node_t * right = &GET_SHAPE_TREE()->shape_cache[node->r - 1];
return right;
}
}
static redblack_node_t *
redblack_find(redblack_node_t * tree, ID key)
{
2023-10-24 18:46:54 +03:00
if (tree == LEAF) {
return LEAF;
}
else {
if (tree->key == key) {
return tree;
}
else {
if (key < tree->key) {
return redblack_find(redblack_left(tree), key);
}
else {
return redblack_find(redblack_right(tree), key);
}
}
}
}
static inline char
redblack_color(redblack_node_t * node)
{
return node && ((uintptr_t)node->value & RED);
}
static inline bool
redblack_red_p(redblack_node_t * node)
{
return redblack_color(node) == RED;
}
static inline rb_shape_t *
redblack_value(redblack_node_t * node)
{
// Color is stored in the bottom bit of the shape pointer
// Mask away the bit so we get the actual pointer back
return (rb_shape_t *)((uintptr_t)node->value & (((uintptr_t)-1) - 1));
}
static redblack_id_t
redblack_id_for(redblack_node_t * node)
{
RUBY_ASSERT(node || node == LEAF);
if (node == LEAF) {
return 0;
}
else {
redblack_node_t * redblack_nodes = GET_SHAPE_TREE()->shape_cache;
redblack_id_t id = (redblack_id_t)(node - redblack_nodes);
return id + 1;
}
}
static redblack_node_t *
redblack_new(char color, ID key, rb_shape_t * value, redblack_node_t * left, redblack_node_t * right)
{
if (GET_SHAPE_TREE()->cache_size + 1 >= REDBLACK_CACHE_SIZE) {
// We're out of cache, just quit
return LEAF;
}
redblack_node_t * redblack_nodes = GET_SHAPE_TREE()->shape_cache;
redblack_node_t * node = &redblack_nodes[(GET_SHAPE_TREE()->cache_size)++];
node->key = key;
node->value = (rb_shape_t *)((uintptr_t)value | color);
node->l = redblack_id_for(left);
node->r = redblack_id_for(right);
return node;
}
static redblack_node_t *
redblack_balance(char color, ID key, rb_shape_t * value, redblack_node_t * left, redblack_node_t * right)
{
if (color == BLACK) {
ID z, y, x;
rb_shape_t * z_, * y_, * x_;
redblack_node_t * a, * b, * c, * d;
if (redblack_red_p(left) && redblack_red_p(redblack_left(left))) {
z = key;
z_ = value;
d = right;
y = left->key;
y_ = redblack_value(left);
c = redblack_right(left);
x = redblack_left(left)->key;
x_ = redblack_value(redblack_left(left));
a = redblack_left(redblack_left(left));
b = redblack_right(redblack_left(left));
}
else if (redblack_red_p(left) && redblack_red_p(redblack_right(left))) {
z = key;
z_ = value;
d = right;
x = left->key;
x_ = redblack_value(left);
a = redblack_left(left);
y = redblack_right(left)->key;
y_ = redblack_value(redblack_right(left));
b = redblack_left(redblack_right(left));
c = redblack_right(redblack_right(left));
}
else if (redblack_red_p(right) && redblack_red_p(redblack_left(right))) {
x = key;
x_ = value;
a = left;
z = right->key;
z_ = redblack_value(right);
d = redblack_right(right);
y = redblack_left(right)->key;
y_ = redblack_value(redblack_left(right));
b = redblack_left(redblack_left(right));
c = redblack_right(redblack_left(right));
}
else if (redblack_red_p(right) && redblack_red_p(redblack_right(right))) {
x = key;
x_ = value;
a = left;
y = right->key;
y_ = redblack_value(right);
b = redblack_left(right);
z = redblack_right(right)->key;
z_ = redblack_value(redblack_right(right));
c = redblack_left(redblack_right(right));
d = redblack_right(redblack_right(right));
}
else {
return redblack_new(color, key, value, left, right);
}
return redblack_new(
RED, y, y_,
redblack_new(BLACK, x, x_, a, b),
redblack_new(BLACK, z, z_, c, d));
}
return redblack_new(color, key, value, left, right);
}
static redblack_node_t *
redblack_insert_aux(redblack_node_t * tree, ID key, rb_shape_t * value)
{
if (tree == LEAF) {
return redblack_new(RED, key, value, LEAF, LEAF);
}
else {
if (key < tree->key) {
return redblack_balance(redblack_color(tree),
tree->key,
redblack_value(tree),
redblack_insert_aux(redblack_left(tree), key, value),
redblack_right(tree));
}
else {
if (key > tree->key) {
return redblack_balance(redblack_color(tree),
tree->key,
redblack_value(tree),
redblack_left(tree),
redblack_insert_aux(redblack_right(tree), key, value));
}
else {
return tree;
}
}
}
}
static redblack_node_t *
redblack_force_black(redblack_node_t * node)
{
node->value = redblack_value(node);
return node;
}
static redblack_node_t *
redblack_insert(redblack_node_t * tree, ID key, rb_shape_t * value)
{
redblack_node_t * root = redblack_insert_aux(tree, key, value);
if (redblack_red_p(root)) {
return redblack_force_black(root);
}
else {
return root;
}
}
rb_shape_tree_t *rb_shape_tree_ptr = NULL;
/*
* Shape getters
*/
rb_shape_t *
2022-10-12 12:27:23 +03:00
rb_shape_get_root_shape(void)
{
return GET_SHAPE_TREE()->root_shape;
}
shape_id_t
rb_shape_id(rb_shape_t * shape)
{
return (shape_id_t)(shape - GET_SHAPE_TREE()->shape_list);
}
void
rb_shape_each_shape(each_shape_callback callback, void *data)
{
rb_shape_t *cursor = rb_shape_get_root_shape();
rb_shape_t *end = rb_shape_get_shape_by_id(GET_SHAPE_TREE()->next_shape_id);
while (cursor < end) {
callback(cursor, data);
cursor += 1;
}
}
2023-03-07 08:56:40 +03:00
RUBY_FUNC_EXPORTED rb_shape_t*
rb_shape_get_shape_by_id(shape_id_t shape_id)
{
RUBY_ASSERT(shape_id != INVALID_SHAPE_ID);
rb_shape_t *shape = &GET_SHAPE_TREE()->shape_list[shape_id];
return shape;
}
rb_shape_t *
rb_shape_get_parent(rb_shape_t * shape)
{
return rb_shape_get_shape_by_id(shape->parent_id);
}
#if !SHAPE_IN_BASIC_FLAGS
shape_id_t rb_generic_shape_id(VALUE obj);
#endif
2023-03-07 08:56:40 +03:00
RUBY_FUNC_EXPORTED shape_id_t
rb_shape_get_shape_id(VALUE obj)
{
if (RB_SPECIAL_CONST_P(obj)) {
return SPECIAL_CONST_SHAPE_ID;
}
#if SHAPE_IN_BASIC_FLAGS
return RBASIC_SHAPE_ID(obj);
#else
switch (BUILTIN_TYPE(obj)) {
case T_OBJECT:
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
return ROBJECT_SHAPE_ID(obj);
break;
case T_CLASS:
case T_MODULE:
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
return RCLASS_SHAPE_ID(obj);
default:
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
return rb_generic_shape_id(obj);
}
#endif
}
size_t
rb_shape_depth(rb_shape_t * shape)
{
size_t depth = 1;
while (shape->parent_id != INVALID_SHAPE_ID) {
depth++;
shape = rb_shape_get_parent(shape);
}
return depth;
}
rb_shape_t*
rb_shape_get_shape(VALUE obj)
{
return rb_shape_get_shape_by_id(rb_shape_get_shape_id(obj));
}
static rb_shape_t *
shape_alloc(void)
{
shape_id_t shape_id = GET_SHAPE_TREE()->next_shape_id;
GET_SHAPE_TREE()->next_shape_id++;
if (shape_id == (MAX_SHAPE_ID + 1)) {
// TODO: Make an OutOfShapesError ??
rb_bug("Out of shapes");
}
return &GET_SHAPE_TREE()->shape_list[shape_id];
}
static rb_shape_t *
rb_shape_alloc_with_parent_id(ID edge_name, shape_id_t parent_id)
{
rb_shape_t * shape = shape_alloc();
shape->edge_name = edge_name;
shape->next_iv_index = 0;
shape->parent_id = parent_id;
2023-03-22 18:47:29 +03:00
shape->edges = NULL;
return shape;
}
static rb_shape_t *
2023-03-22 18:47:29 +03:00
rb_shape_alloc(ID edge_name, rb_shape_t * parent, enum shape_type type)
{
rb_shape_t * shape = rb_shape_alloc_with_parent_id(edge_name, rb_shape_id(parent));
2023-03-22 18:47:29 +03:00
shape->type = (uint8_t)type;
shape->size_pool_index = parent->size_pool_index;
shape->capacity = parent->capacity;
shape->edges = 0;
return shape;
}
#ifdef HAVE_MMAP
static redblack_node_t *
redblack_cache_ancestors(rb_shape_t * shape)
{
2023-10-24 21:59:48 +03:00
if (!(shape->ancestor_index || shape->parent_id == INVALID_SHAPE_ID)) {
redblack_node_t * parent_index;
2023-10-24 21:59:48 +03:00
parent_index = redblack_cache_ancestors(rb_shape_get_parent(shape));
2023-10-24 21:59:48 +03:00
if (shape->type == SHAPE_IVAR) {
shape->ancestor_index = redblack_insert(parent_index, shape->edge_name, shape);
}
else {
shape->ancestor_index = parent_index;
}
}
2023-10-24 21:59:48 +03:00
return shape->ancestor_index;
}
#else
static redblack_node_t *
redblack_cache_ancestors(rb_shape_t * shape)
{
return LEAF;
}
#endif
static rb_shape_t *
rb_shape_alloc_new_child(ID id, rb_shape_t * shape, enum shape_type shape_type)
{
rb_shape_t * new_shape = rb_shape_alloc(id, shape, shape_type);
switch (shape_type) {
case SHAPE_IVAR:
new_shape->next_iv_index = shape->next_iv_index + 1;
if (new_shape->next_iv_index > ANCESTOR_CACHE_THRESHOLD) {
redblack_cache_ancestors(new_shape);
}
break;
case SHAPE_CAPACITY_CHANGE:
case SHAPE_FROZEN:
case SHAPE_T_OBJECT:
new_shape->next_iv_index = shape->next_iv_index;
break;
case SHAPE_OBJ_TOO_COMPLEX:
case SHAPE_ROOT:
rb_bug("Unreachable");
break;
}
return new_shape;
}
static rb_shape_t*
get_next_shape_internal(rb_shape_t * shape, ID id, enum shape_type shape_type, bool * variation_created, bool new_variations_allowed)
{
rb_shape_t *res = NULL;
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
// There should never be outgoing edges from "too complex"
RUBY_ASSERT(rb_shape_id(shape) != OBJ_TOO_COMPLEX_SHAPE_ID);
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
*variation_created = false;
RB_VM_LOCK_ENTER();
{
// If the current shape has children
if (shape->edges) {
// Check if it only has one child
if (SINGLE_CHILD_P(shape->edges)) {
rb_shape_t * child = SINGLE_CHILD(shape->edges);
// If the one child has a matching edge name, then great,
// we found what we want.
if (child->edge_name == id) {
res = child;
}
}
else {
// If it has more than one child, do a hash lookup to find it.
VALUE lookup_result;
if (rb_id_table_lookup(shape->edges, id, &lookup_result)) {
res = (rb_shape_t *)lookup_result;
}
}
}
// If we didn't find the shape we're looking for we create it.
if (!res) {
// If we're not allowed to create a new variation, of if we're out of shapes
// we return TOO_COMPLEX_SHAPE.
if (!new_variations_allowed || GET_SHAPE_TREE()->next_shape_id > MAX_SHAPE_ID) {
res = rb_shape_get_shape_by_id(OBJ_TOO_COMPLEX_SHAPE_ID);
}
else {
rb_shape_t * new_shape = rb_shape_alloc_new_child(id, shape, shape_type);
if (!shape->edges) {
// If the shape had no edge yet, we can directly set the new child
shape->edges = TAG_SINGLE_CHILD(new_shape);
}
else {
// If the edge was single child we need to allocate a table.
if (SINGLE_CHILD_P(shape->edges)) {
rb_shape_t * old_child = SINGLE_CHILD(shape->edges);
shape->edges = rb_id_table_create(2);
rb_id_table_insert(shape->edges, old_child->edge_name, (VALUE)old_child);
}
rb_id_table_insert(shape->edges, new_shape->edge_name, (VALUE)new_shape);
*variation_created = true;
}
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
res = new_shape;
}
}
}
RB_VM_LOCK_LEAVE();
return res;
}
2023-03-07 08:34:31 +03:00
int
rb_shape_frozen_shape_p(rb_shape_t* shape)
{
return SHAPE_FROZEN == (enum shape_type)shape->type;
}
static void
move_iv(VALUE obj, ID id, attr_index_t from, attr_index_t to)
{
switch(BUILTIN_TYPE(obj)) {
case T_CLASS:
case T_MODULE:
RCLASS_IVPTR(obj)[to] = RCLASS_IVPTR(obj)[from];
break;
case T_OBJECT:
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
RUBY_ASSERT(!rb_shape_obj_too_complex(obj));
ROBJECT_IVPTR(obj)[to] = ROBJECT_IVPTR(obj)[from];
break;
default: {
struct gen_ivtbl *ivtbl;
rb_gen_ivtbl_get(obj, id, &ivtbl);
ivtbl->as.shape.ivptr[to] = ivtbl->as.shape.ivptr[from];
break;
}
}
}
static rb_shape_t *
remove_shape_recursive(VALUE obj, ID id, rb_shape_t * shape, VALUE * removed)
{
if (shape->parent_id == INVALID_SHAPE_ID) {
// We've hit the top of the shape tree and couldn't find the
// IV we wanted to remove, so return NULL
return NULL;
}
else {
if (shape->type == SHAPE_IVAR && shape->edge_name == id) {
// We've hit the edge we wanted to remove, return it's _parent_
// as the new parent while we go back down the stack.
attr_index_t index = shape->next_iv_index - 1;
switch(BUILTIN_TYPE(obj)) {
case T_CLASS:
case T_MODULE:
*removed = RCLASS_IVPTR(obj)[index];
break;
case T_OBJECT:
*removed = ROBJECT_IVPTR(obj)[index];
break;
default: {
struct gen_ivtbl *ivtbl;
rb_gen_ivtbl_get(obj, id, &ivtbl);
*removed = ivtbl->as.shape.ivptr[index];
break;
}
}
return rb_shape_get_parent(shape);
}
else {
// This isn't the IV we want to remove, keep walking up.
rb_shape_t * new_parent = remove_shape_recursive(obj, id, rb_shape_get_parent(shape), removed);
// We found a new parent. Create a child of the new parent that
// has the same attributes as this shape.
if (new_parent) {
if (UNLIKELY(new_parent->type == SHAPE_OBJ_TOO_COMPLEX)) {
return new_parent;
}
bool dont_care;
rb_shape_t * new_child = get_next_shape_internal(new_parent, shape->edge_name, shape->type, &dont_care, true);
if (UNLIKELY(new_child->type == SHAPE_OBJ_TOO_COMPLEX)) {
return new_child;
}
new_child->capacity = shape->capacity;
if (new_child->type == SHAPE_IVAR) {
move_iv(obj, id, shape->next_iv_index - 1, new_child->next_iv_index - 1);
}
return new_child;
}
else {
// We went all the way to the top of the shape tree and couldn't
// find an IV to remove, so return NULL
return NULL;
}
}
}
}
bool
rb_shape_transition_shape_remove_ivar(VALUE obj, ID id, rb_shape_t *shape, VALUE * removed)
{
if (UNLIKELY(shape->type == SHAPE_OBJ_TOO_COMPLEX)) {
return false;
}
rb_shape_t * new_shape = remove_shape_recursive(obj, id, shape, removed);
if (new_shape) {
if (UNLIKELY(new_shape->type == SHAPE_OBJ_TOO_COMPLEX)) {
return false;
}
rb_shape_set_shape(obj, new_shape);
}
return true;
}
2023-10-20 02:01:35 +03:00
rb_shape_t *
rb_shape_transition_shape_frozen(VALUE obj)
{
rb_shape_t* shape = rb_shape_get_shape(obj);
RUBY_ASSERT(shape);
RUBY_ASSERT(RB_OBJ_FROZEN(obj));
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
if (rb_shape_frozen_shape_p(shape) || rb_shape_obj_too_complex(obj)) {
2023-10-20 02:01:35 +03:00
return shape;
}
rb_shape_t* next_shape;
if (shape == rb_shape_get_root_shape()) {
2023-10-20 02:01:35 +03:00
return rb_shape_get_shape_by_id(SPECIAL_CONST_SHAPE_ID);
}
bool dont_care;
next_shape = get_next_shape_internal(shape, (ID)id_frozen, SHAPE_FROZEN, &dont_care, true);
RUBY_ASSERT(next_shape);
2023-10-20 02:01:35 +03:00
return next_shape;
}
/*
* This function is used for assertions where we don't want to increment
* max_iv_count
*/
rb_shape_t *
rb_shape_get_next_iv_shape(rb_shape_t* shape, ID id)
{
RUBY_ASSERT(!is_instance_id(id) || RTEST(rb_sym2str(ID2SYM(id))));
bool dont_care;
return get_next_shape_internal(shape, id, SHAPE_IVAR, &dont_care, true);
}
rb_shape_t *
rb_shape_get_next(rb_shape_t* shape, VALUE obj, ID id)
{
RUBY_ASSERT(!is_instance_id(id) || RTEST(rb_sym2str(ID2SYM(id))));
RUBY_ASSERT(shape->type != SHAPE_OBJ_TOO_COMPLEX);
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
bool allow_new_shape = true;
if (BUILTIN_TYPE(obj) == T_OBJECT) {
VALUE klass = rb_obj_class(obj);
allow_new_shape = RCLASS_EXT(klass)->variation_count < SHAPE_MAX_VARIATIONS;
}
bool variation_created = false;
rb_shape_t * new_shape = get_next_shape_internal(shape, id, SHAPE_IVAR, &variation_created, allow_new_shape);
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
// Check if we should update max_iv_count on the object's class
if (BUILTIN_TYPE(obj) == T_OBJECT) {
VALUE klass = rb_obj_class(obj);
if (new_shape->next_iv_index > RCLASS_EXT(klass)->max_iv_count) {
RCLASS_EXT(klass)->max_iv_count = new_shape->next_iv_index;
}
if (variation_created) {
RCLASS_EXT(klass)->variation_count++;
if (rb_warning_category_enabled_p(RB_WARN_CATEGORY_PERFORMANCE)) {
if (RCLASS_EXT(klass)->variation_count >= SHAPE_MAX_VARIATIONS) {
rb_category_warn(
RB_WARN_CATEGORY_PERFORMANCE,
"Maximum shapes variations (%d) reached by %"PRIsVALUE", instance variables accesses will be slower.",
SHAPE_MAX_VARIATIONS,
rb_class_path(klass)
);
}
}
}
}
return new_shape;
}
static inline rb_shape_t *
rb_shape_transition_shape_capa: use optimal sizes transitions Previously the growth was 3(embed), 6, 12, 24, ... With this change it's now 3(embed), 8, 16, 32, 64, ... by default. However, since power of two isn't the best size for all allocators, if `malloc_usable_size` is vailable, we use it to discover the best offset. On Linux/glibc 2.35 for instance, the growth will be 3(embed), 7, 15, 31 to avoid wasting 8B per object. Test program: ```c size_t test(size_t slots) { size_t allocated = slots * VALUE_SIZE; void *test_ptr = malloc(allocated); size_t wasted = malloc_usable_size(test_ptr) - allocated; free(test_ptr); fprintf(stderr, "slots = %lu, wasted_bytes = %lu\n", slots, wasted); return wasted; } int main(int argc, char *argv[]) { size_t best_padding = 0; size_t padding = 0; for (padding = 0; padding <= 2; padding++) { size_t wasted = test(8 - padding); if (wasted == 0) { best_padding = padding; break; } } size_t index = 0; fprintf(stderr, "=============== naive ================\n"); size_t list_size = 4; for (index = 0; index < 10; index++) { test(list_size); list_size *= 2; } fprintf(stderr, "=============== auto-padded (-%lu) ================\n", best_padding); list_size = 4; for (index = 0; index < 10; index ++) { test(list_size - best_padding); list_size *= 2; } fprintf(stderr, "\n\n"); return 0; } ``` ``` ===== glibc ====== slots = 8, wasted_bytes = 8 slots = 7, wasted_bytes = 0 =============== naive ================ slots = 4, wasted_bytes = 8 slots = 8, wasted_bytes = 8 slots = 16, wasted_bytes = 8 slots = 32, wasted_bytes = 8 slots = 64, wasted_bytes = 8 slots = 128, wasted_bytes = 8 slots = 256, wasted_bytes = 8 slots = 512, wasted_bytes = 8 slots = 1024, wasted_bytes = 8 slots = 2048, wasted_bytes = 8 =============== auto-padded (-1) ================ slots = 3, wasted_bytes = 0 slots = 7, wasted_bytes = 0 slots = 15, wasted_bytes = 0 slots = 31, wasted_bytes = 0 slots = 63, wasted_bytes = 0 slots = 127, wasted_bytes = 0 slots = 255, wasted_bytes = 0 slots = 511, wasted_bytes = 0 slots = 1023, wasted_bytes = 0 slots = 2047, wasted_bytes = 0 ``` ``` ========== jemalloc ======= slots = 8, wasted_bytes = 0 =============== naive ================ slots = 4, wasted_bytes = 0 slots = 8, wasted_bytes = 0 slots = 16, wasted_bytes = 0 slots = 32, wasted_bytes = 0 slots = 64, wasted_bytes = 0 slots = 128, wasted_bytes = 0 slots = 256, wasted_bytes = 0 slots = 512, wasted_bytes = 0 slots = 1024, wasted_bytes = 0 slots = 2048, wasted_bytes = 0 =============== auto-padded (-0) ================ slots = 4, wasted_bytes = 0 slots = 8, wasted_bytes = 0 slots = 16, wasted_bytes = 0 slots = 32, wasted_bytes = 0 slots = 64, wasted_bytes = 0 slots = 128, wasted_bytes = 0 slots = 256, wasted_bytes = 0 slots = 512, wasted_bytes = 0 slots = 1024, wasted_bytes = 0 slots = 2048, wasted_bytes = 0 ```
2023-10-10 16:32:12 +03:00
rb_shape_transition_shape_capa_create(rb_shape_t* shape, size_t new_capacity)
{
rb_shape_transition_shape_capa: use optimal sizes transitions Previously the growth was 3(embed), 6, 12, 24, ... With this change it's now 3(embed), 8, 16, 32, 64, ... by default. However, since power of two isn't the best size for all allocators, if `malloc_usable_size` is vailable, we use it to discover the best offset. On Linux/glibc 2.35 for instance, the growth will be 3(embed), 7, 15, 31 to avoid wasting 8B per object. Test program: ```c size_t test(size_t slots) { size_t allocated = slots * VALUE_SIZE; void *test_ptr = malloc(allocated); size_t wasted = malloc_usable_size(test_ptr) - allocated; free(test_ptr); fprintf(stderr, "slots = %lu, wasted_bytes = %lu\n", slots, wasted); return wasted; } int main(int argc, char *argv[]) { size_t best_padding = 0; size_t padding = 0; for (padding = 0; padding <= 2; padding++) { size_t wasted = test(8 - padding); if (wasted == 0) { best_padding = padding; break; } } size_t index = 0; fprintf(stderr, "=============== naive ================\n"); size_t list_size = 4; for (index = 0; index < 10; index++) { test(list_size); list_size *= 2; } fprintf(stderr, "=============== auto-padded (-%lu) ================\n", best_padding); list_size = 4; for (index = 0; index < 10; index ++) { test(list_size - best_padding); list_size *= 2; } fprintf(stderr, "\n\n"); return 0; } ``` ``` ===== glibc ====== slots = 8, wasted_bytes = 8 slots = 7, wasted_bytes = 0 =============== naive ================ slots = 4, wasted_bytes = 8 slots = 8, wasted_bytes = 8 slots = 16, wasted_bytes = 8 slots = 32, wasted_bytes = 8 slots = 64, wasted_bytes = 8 slots = 128, wasted_bytes = 8 slots = 256, wasted_bytes = 8 slots = 512, wasted_bytes = 8 slots = 1024, wasted_bytes = 8 slots = 2048, wasted_bytes = 8 =============== auto-padded (-1) ================ slots = 3, wasted_bytes = 0 slots = 7, wasted_bytes = 0 slots = 15, wasted_bytes = 0 slots = 31, wasted_bytes = 0 slots = 63, wasted_bytes = 0 slots = 127, wasted_bytes = 0 slots = 255, wasted_bytes = 0 slots = 511, wasted_bytes = 0 slots = 1023, wasted_bytes = 0 slots = 2047, wasted_bytes = 0 ``` ``` ========== jemalloc ======= slots = 8, wasted_bytes = 0 =============== naive ================ slots = 4, wasted_bytes = 0 slots = 8, wasted_bytes = 0 slots = 16, wasted_bytes = 0 slots = 32, wasted_bytes = 0 slots = 64, wasted_bytes = 0 slots = 128, wasted_bytes = 0 slots = 256, wasted_bytes = 0 slots = 512, wasted_bytes = 0 slots = 1024, wasted_bytes = 0 slots = 2048, wasted_bytes = 0 =============== auto-padded (-0) ================ slots = 4, wasted_bytes = 0 slots = 8, wasted_bytes = 0 slots = 16, wasted_bytes = 0 slots = 32, wasted_bytes = 0 slots = 64, wasted_bytes = 0 slots = 128, wasted_bytes = 0 slots = 256, wasted_bytes = 0 slots = 512, wasted_bytes = 0 slots = 1024, wasted_bytes = 0 slots = 2048, wasted_bytes = 0 ```
2023-10-10 16:32:12 +03:00
RUBY_ASSERT(new_capacity < (size_t)MAX_IVARS);
ID edge_name = rb_make_temporary_id(new_capacity);
bool dont_care;
rb_shape_t * new_shape = get_next_shape_internal(shape, edge_name, SHAPE_CAPACITY_CHANGE, &dont_care, true);
if (rb_shape_id(new_shape) != OBJ_TOO_COMPLEX_SHAPE_ID) {
new_shape->capacity = (uint32_t)new_capacity;
}
return new_shape;
}
rb_shape_t *
rb_shape_transition_shape_capa(rb_shape_t* shape)
{
2023-10-23 13:28:14 +03:00
return rb_shape_transition_shape_capa_create(shape, rb_malloc_grow_capa(shape->capacity, sizeof(VALUE)));
}
vm_getivar: assume the cached shape_id like have a common ancestor When an inline cache misses, it is very likely that the stale shape_id and the current instance shape_id have a close common ancestor. For example if the instance variable is sometimes frozen sometimes not, one of the two shape will be the direct parent of the other. Another pattern that commonly cause IC misses is "memoization", in such case the object will have a "base common shape" and then a number of close descendants. In addition, when we find a common ancestor, we store it in the inline cache instead of the current shape. This help prevent the cache from flip-flopping, ensuring the next lookup will be marginally faster and more generally avoid writing in memory too much. However, now that shapes have an ancestors index, we only check for a few ancestors before falling back to use the index. So overall this change speeds up what is assumed to be the more common case, but makes what is assumed to be the less common case a bit slower. ``` compare-ruby: ruby 3.3.0dev (2023-10-26T05:30:17Z master 701ca070b4) [arm64-darwin22] built-ruby: ruby 3.3.0dev (2023-10-26T09:25:09Z shapes_double_sear.. a723a85235) [arm64-darwin22] warming up...... | |compare-ruby|built-ruby| |:------------------------------------|-----------:|---------:| |vm_ivar_stable_shape | 11.672M| 11.679M| | | -| 1.00x| |vm_ivar_memoize_unstable_shape | 7.551M| 10.506M| | | -| 1.39x| |vm_ivar_memoize_unstable_shape_miss | 11.591M| 11.624M| | | -| 1.00x| |vm_ivar_unstable_undef | 9.037M| 7.981M| | | 1.13x| -| |vm_ivar_divergent_shape | 8.034M| 6.657M| | | 1.21x| -| |vm_ivar_divergent_shape_imbalanced | 10.471M| 9.231M| | | 1.13x| -| ``` Co-Authored-By: John Hawthorn <john@hawthorn.email>
2023-10-26 12:08:05 +03:00
// Same as rb_shape_get_iv_index, but uses a provided valid shape id and index
// to return a result faster if branches of the shape tree are closely related.
bool
rb_shape_get_iv_index_with_hint(shape_id_t shape_id, ID id, attr_index_t *value, shape_id_t *shape_id_hint)
{
attr_index_t index_hint = *value;
rb_shape_t *shape = rb_shape_get_shape_by_id(shape_id);
rb_shape_t *initial_shape = shape;
if (*shape_id_hint == INVALID_SHAPE_ID) {
*shape_id_hint = shape_id;
return rb_shape_get_iv_index(shape, id, value);
}
rb_shape_t * shape_hint = rb_shape_get_shape_by_id(*shape_id_hint);
// We assume it's likely shape_id_hint and shape_id have a close common
// ancestor, so we check up to ANCESTOR_SEARCH_MAX_DEPTH ancestors before
// eventually using the index, as in case of a match it will be faster.
// However if the shape doesn't have an index, we walk the entire tree.
int depth = INT_MAX;
if (shape->ancestor_index && shape->next_iv_index >= ANCESTOR_CACHE_THRESHOLD) {
depth = ANCESTOR_SEARCH_MAX_DEPTH;
}
while (depth > 0 && shape->next_iv_index > index_hint) {
while (shape_hint->next_iv_index > shape->next_iv_index) {
shape_hint = rb_shape_get_parent(shape_hint);
}
if (shape_hint == shape) {
// We've found a common ancestor so use the index hint
*value = index_hint;
*shape_id_hint = rb_shape_id(shape);
return true;
}
if (shape->edge_name == id) {
// We found the matching id before a common ancestor
*value = shape->next_iv_index - 1;
*shape_id_hint = rb_shape_id(shape);
return true;
}
shape = rb_shape_get_parent(shape);
depth--;
}
// If the original shape had an index but its ancestor doesn't
// we switch back to the original one as it will be faster.
if (!shape->ancestor_index && initial_shape->ancestor_index) {
shape = initial_shape;
}
*shape_id_hint = shape_id;
return rb_shape_get_iv_index(shape, id, value);
}
bool
2022-10-12 12:27:23 +03:00
rb_shape_get_iv_index(rb_shape_t * shape, ID id, attr_index_t *value)
{
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
// It doesn't make sense to ask for the index of an IV that's stored
// on an object that is "too complex" as it uses a hash for storing IVs
RUBY_ASSERT(rb_shape_id(shape) != OBJ_TOO_COMPLEX_SHAPE_ID);
while (shape->parent_id != INVALID_SHAPE_ID) {
// Try the ancestor cache if it's available
if (shape->ancestor_index && shape->next_iv_index >= ANCESTOR_CACHE_THRESHOLD) {
redblack_node_t * node = redblack_find(shape->ancestor_index, id);
if (node) {
rb_shape_t * shape = redblack_value(node);
*value = shape->next_iv_index - 1;
2022-10-12 12:27:23 +03:00
return true;
}
else {
2022-10-12 12:27:23 +03:00
return false;
}
}
else {
if (shape->edge_name == id) {
enum shape_type shape_type;
shape_type = (enum shape_type)shape->type;
switch (shape_type) {
case SHAPE_IVAR:
RUBY_ASSERT(shape->next_iv_index > 0);
*value = shape->next_iv_index - 1;
return true;
case SHAPE_CAPACITY_CHANGE:
case SHAPE_ROOT:
case SHAPE_T_OBJECT:
return false;
case SHAPE_OBJ_TOO_COMPLEX:
case SHAPE_FROZEN:
rb_bug("Ivar should not exist on transition");
}
}
}
shape = rb_shape_get_parent(shape);
}
return false;
}
2023-03-07 08:34:31 +03:00
void
rb_shape_set_shape(VALUE obj, rb_shape_t* shape)
{
rb_shape_set_shape_id(obj, rb_shape_id(shape));
}
2022-12-02 20:33:20 +03:00
int32_t
rb_shape_id_offset(void)
{
return sizeof(uintptr_t) - SHAPE_ID_NUM_BITS / sizeof(uintptr_t);
2022-12-02 20:33:20 +03:00
}
rb_shape_t *
rb_shape_traverse_from_new_root(rb_shape_t *initial_shape, rb_shape_t *dest_shape)
{
RUBY_ASSERT(initial_shape->type == SHAPE_T_OBJECT);
rb_shape_t *next_shape = initial_shape;
if (dest_shape->type != initial_shape->type) {
next_shape = rb_shape_traverse_from_new_root(initial_shape, rb_shape_get_parent(dest_shape));
if (!next_shape) {
return NULL;
}
}
switch ((enum shape_type)dest_shape->type) {
case SHAPE_IVAR:
case SHAPE_FROZEN:
if (!next_shape->edges) {
return NULL;
}
VALUE lookup_result;
if (SINGLE_CHILD_P(next_shape->edges)) {
rb_shape_t * child = SINGLE_CHILD(next_shape->edges);
if (child->edge_name == dest_shape->edge_name) {
return child;
}
else {
return NULL;
}
}
else {
if (rb_id_table_lookup(next_shape->edges, dest_shape->edge_name, &lookup_result)) {
next_shape = (rb_shape_t *)lookup_result;
}
else {
return NULL;
}
}
break;
case SHAPE_ROOT:
case SHAPE_CAPACITY_CHANGE:
case SHAPE_T_OBJECT:
break;
case SHAPE_OBJ_TOO_COMPLEX:
rb_bug("Unreachable");
break;
}
return next_shape;
}
rb_shape_t *
rb_shape_rebuild_shape(rb_shape_t * initial_shape, rb_shape_t * dest_shape)
{
RUBY_ASSERT(rb_shape_id(initial_shape) != OBJ_TOO_COMPLEX_SHAPE_ID);
RUBY_ASSERT(rb_shape_id(dest_shape) != OBJ_TOO_COMPLEX_SHAPE_ID);
rb_shape_t * midway_shape;
RUBY_ASSERT(initial_shape->type == SHAPE_T_OBJECT);
if (dest_shape->type != initial_shape->type) {
midway_shape = rb_shape_rebuild_shape(initial_shape, rb_shape_get_parent(dest_shape));
if (UNLIKELY(rb_shape_id(midway_shape) == OBJ_TOO_COMPLEX_SHAPE_ID)) {
return midway_shape;
}
}
else {
midway_shape = initial_shape;
}
switch ((enum shape_type)dest_shape->type) {
case SHAPE_IVAR:
if (midway_shape->capacity <= midway_shape->next_iv_index) {
// There isn't enough room to write this IV, so we need to increase the capacity
midway_shape = rb_shape_transition_shape_capa(midway_shape);
}
if (LIKELY(rb_shape_id(midway_shape) != OBJ_TOO_COMPLEX_SHAPE_ID)) {
midway_shape = rb_shape_get_next_iv_shape(midway_shape, dest_shape->edge_name);
}
break;
case SHAPE_ROOT:
case SHAPE_FROZEN:
case SHAPE_CAPACITY_CHANGE:
case SHAPE_T_OBJECT:
break;
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
case SHAPE_OBJ_TOO_COMPLEX:
rb_bug("Unreachable");
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
break;
}
return midway_shape;
}
2023-03-07 08:56:40 +03:00
RUBY_FUNC_EXPORTED bool
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
rb_shape_obj_too_complex(VALUE obj)
{
return rb_shape_get_shape_id(obj) == OBJ_TOO_COMPLEX_SHAPE_ID;
}
void
rb_shape_set_too_complex(VALUE obj)
{
RUBY_ASSERT(!rb_shape_obj_too_complex(obj));
rb_shape_set_shape_id(obj, OBJ_TOO_COMPLEX_SHAPE_ID);
}
size_t
rb_shape_edges_count(rb_shape_t *shape)
{
if (shape->edges) {
if (SINGLE_CHILD_P(shape->edges)) {
return 1;
}
else {
return rb_id_table_size(shape->edges);
}
}
return 0;
}
size_t
rb_shape_memsize(rb_shape_t *shape)
{
size_t memsize = sizeof(rb_shape_t);
if (shape->edges && !SINGLE_CHILD_P(shape->edges)) {
memsize += rb_id_table_memsize(shape->edges);
}
return memsize;
}
#if SHAPE_DEBUG
/*
* Exposing Shape to Ruby via RubyVM.debug_shape
*/
/* :nodoc: */
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
static VALUE
rb_shape_too_complex(VALUE self)
{
rb_shape_t * shape;
shape = rb_shape_get_shape_by_id(NUM2INT(rb_struct_getmember(self, rb_intern("id"))));
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
if (rb_shape_id(shape) == OBJ_TOO_COMPLEX_SHAPE_ID) {
return Qtrue;
}
else {
return Qfalse;
}
}
2022-10-12 12:27:23 +03:00
static VALUE
parse_key(ID key)
{
if (is_instance_id(key)) {
return ID2SYM(key);
}
return LONG2NUM(key);
}
static VALUE rb_shape_edge_name(rb_shape_t * shape);
static VALUE
2022-10-12 12:27:23 +03:00
rb_shape_t_to_rb_cShape(rb_shape_t *shape)
{
VALUE rb_cShape = rb_const_get(rb_cRubyVM, rb_intern("Shape"));
VALUE obj = rb_struct_new(rb_cShape,
INT2NUM(rb_shape_id(shape)),
INT2NUM(shape->parent_id),
rb_shape_edge_name(shape),
INT2NUM(shape->next_iv_index),
INT2NUM(shape->size_pool_index),
INT2NUM(shape->type),
INT2NUM(shape->capacity));
rb_obj_freeze(obj);
return obj;
}
2022-10-12 12:27:23 +03:00
static enum rb_id_table_iterator_result
rb_edges_to_hash(ID key, VALUE value, void *ref)
{
rb_hash_aset(*(VALUE *)ref, parse_key(key), rb_shape_t_to_rb_cShape((rb_shape_t*)value));
return ID_TABLE_CONTINUE;
}
/* :nodoc: */
static VALUE
rb_shape_edges(VALUE self)
{
rb_shape_t* shape;
shape = rb_shape_get_shape_by_id(NUM2INT(rb_struct_getmember(self, rb_intern("id"))));
VALUE hash = rb_hash_new();
if (shape->edges) {
if (SINGLE_CHILD_P(shape->edges)) {
rb_shape_t * child = SINGLE_CHILD(shape->edges);
rb_edges_to_hash(child->edge_name, (VALUE)child, &hash);
}
else {
rb_id_table_foreach(shape->edges, rb_edges_to_hash, &hash);
}
}
return hash;
}
static VALUE
rb_shape_edge_name(rb_shape_t * shape)
{
if (shape->edge_name) {
if (is_instance_id(shape->edge_name)) {
return ID2SYM(shape->edge_name);
}
return INT2NUM(shape->capacity);
}
return Qnil;
}
/* :nodoc: */
static VALUE
rb_shape_export_depth(VALUE self)
{
rb_shape_t* shape;
shape = rb_shape_get_shape_by_id(NUM2INT(rb_struct_getmember(self, rb_intern("id"))));
return SIZET2NUM(rb_shape_depth(shape));
}
/* :nodoc: */
static VALUE
rb_shape_parent(VALUE self)
{
rb_shape_t * shape;
shape = rb_shape_get_shape_by_id(NUM2INT(rb_struct_getmember(self, rb_intern("id"))));
if (shape->parent_id != INVALID_SHAPE_ID) {
return rb_shape_t_to_rb_cShape(rb_shape_get_parent(shape));
}
else {
return Qnil;
}
}
/* :nodoc: */
static VALUE
2022-10-12 12:27:23 +03:00
rb_shape_debug_shape(VALUE self, VALUE obj)
{
return rb_shape_t_to_rb_cShape(rb_shape_get_shape(obj));
}
/* :nodoc: */
static VALUE
2022-10-12 12:27:23 +03:00
rb_shape_root_shape(VALUE self)
{
return rb_shape_t_to_rb_cShape(rb_shape_get_root_shape());
}
/* :nodoc: */
static VALUE
rb_shape_shapes_available(VALUE self)
{
return INT2NUM(MAX_SHAPE_ID - (GET_SHAPE_TREE()->next_shape_id - 1));
}
VALUE rb_obj_shape(rb_shape_t* shape);
static enum rb_id_table_iterator_result collect_keys_and_values(ID key, VALUE value, void *ref)
{
rb_hash_aset(*(VALUE *)ref, parse_key(key), rb_obj_shape((rb_shape_t*)value));
return ID_TABLE_CONTINUE;
}
static VALUE edges(struct rb_id_table* edges)
{
VALUE hash = rb_hash_new();
if (SINGLE_CHILD_P(edges)) {
rb_shape_t * child = SINGLE_CHILD(edges);
collect_keys_and_values(child->edge_name, (VALUE)child, &hash);
}
else {
rb_id_table_foreach(edges, collect_keys_and_values, &hash);
}
return hash;
}
/* :nodoc: */
2022-10-12 12:27:23 +03:00
VALUE
rb_obj_shape(rb_shape_t* shape)
{
VALUE rb_shape = rb_hash_new();
rb_hash_aset(rb_shape, ID2SYM(rb_intern("id")), INT2NUM(rb_shape_id(shape)));
rb_hash_aset(rb_shape, ID2SYM(rb_intern("edges")), edges(shape->edges));
if (shape == rb_shape_get_root_shape()) {
rb_hash_aset(rb_shape, ID2SYM(rb_intern("parent_id")), INT2NUM(ROOT_SHAPE_ID));
}
else {
rb_hash_aset(rb_shape, ID2SYM(rb_intern("parent_id")), INT2NUM(shape->parent_id));
}
rb_hash_aset(rb_shape, ID2SYM(rb_intern("edge_name")), rb_id2str(shape->edge_name));
return rb_shape;
}
/* :nodoc: */
2022-10-12 12:27:23 +03:00
static VALUE
shape_transition_tree(VALUE self)
{
return rb_obj_shape(rb_shape_get_root_shape());
}
/* :nodoc: */
static VALUE
rb_shape_find_by_id(VALUE mod, VALUE id)
{
shape_id_t shape_id = NUM2UINT(id);
if (shape_id >= GET_SHAPE_TREE()->next_shape_id) {
rb_raise(rb_eArgError, "Shape ID %d is out of bounds\n", shape_id);
}
return rb_shape_t_to_rb_cShape(rb_shape_get_shape_by_id(shape_id));
}
#endif
#ifdef HAVE_MMAP
#include <sys/mman.h>
#endif
void
Init_default_shapes(void)
{
rb_shape_tree_t *st = ruby_mimmalloc(sizeof(rb_shape_tree_t));
memset(st, 0, sizeof(rb_shape_tree_t));
rb_shape_tree_ptr = st;
#ifdef HAVE_MMAP
rb_shape_tree_ptr->shape_list = (rb_shape_t *)mmap(NULL, rb_size_mul_or_raise(SHAPE_BUFFER_SIZE, sizeof(rb_shape_t), rb_eRuntimeError),
PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
if (GET_SHAPE_TREE()->shape_list == MAP_FAILED) {
GET_SHAPE_TREE()->shape_list = 0;
}
#else
GET_SHAPE_TREE()->shape_list = xcalloc(SHAPE_BUFFER_SIZE, sizeof(rb_shape_t));
#endif
if (!GET_SHAPE_TREE()->shape_list) {
rb_memerror();
}
id_frozen = rb_make_internal_id();
id_t_object = rb_make_internal_id();
#ifdef HAVE_MMAP
rb_shape_tree_ptr->shape_cache = (redblack_node_t *)mmap(NULL, rb_size_mul_or_raise(REDBLACK_CACHE_SIZE, sizeof(redblack_node_t), rb_eRuntimeError),
PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
rb_shape_tree_ptr->cache_size = 0;
#endif
// Shapes by size pool
for (int i = 0; i < SIZE_POOL_COUNT; i++) {
size_pool_edge_names[i] = rb_make_internal_id();
}
// Root shape
rb_shape_t *root = rb_shape_alloc_with_parent_id(0, INVALID_SHAPE_ID);
root->capacity = 0;
root->type = SHAPE_ROOT;
root->size_pool_index = 0;
GET_SHAPE_TREE()->root_shape = root;
RUBY_ASSERT(rb_shape_id(GET_SHAPE_TREE()->root_shape) == ROOT_SHAPE_ID);
// Shapes by size pool
for (int i = 1; i < SIZE_POOL_COUNT; i++) {
rb_shape_t *new_shape = rb_shape_alloc_with_parent_id(0, INVALID_SHAPE_ID);
new_shape->type = SHAPE_ROOT;
new_shape->size_pool_index = i;
new_shape->ancestor_index = LEAF;
RUBY_ASSERT(rb_shape_id(new_shape) == (shape_id_t)i);
}
// Make shapes for T_OBJECT
for (int i = 0; i < SIZE_POOL_COUNT; i++) {
rb_shape_t * shape = rb_shape_get_shape_by_id(i);
bool dont_care;
rb_shape_t * t_object_shape =
get_next_shape_internal(shape, id_t_object, SHAPE_T_OBJECT, &dont_care, true);
t_object_shape->capacity = (uint32_t)((rb_size_pool_slot_size(i) - offsetof(struct RObject, as.ary)) / sizeof(VALUE));
t_object_shape->edges = rb_id_table_create(0);
t_object_shape->ancestor_index = LEAF;
RUBY_ASSERT(rb_shape_id(t_object_shape) == (shape_id_t)(i + SIZE_POOL_COUNT));
}
bool dont_care;
// Special const shape
#if RUBY_DEBUG
rb_shape_t * special_const_shape =
#endif
get_next_shape_internal(root, (ID)id_frozen, SHAPE_FROZEN, &dont_care, true);
RUBY_ASSERT(rb_shape_id(special_const_shape) == SPECIAL_CONST_SHAPE_ID);
RUBY_ASSERT(SPECIAL_CONST_SHAPE_ID == (GET_SHAPE_TREE()->next_shape_id - 1));
RUBY_ASSERT(rb_shape_frozen_shape_p(special_const_shape));
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
rb_shape_t * hash_fallback_shape = rb_shape_alloc_with_parent_id(0, ROOT_SHAPE_ID);
hash_fallback_shape->type = SHAPE_OBJ_TOO_COMPLEX;
hash_fallback_shape->size_pool_index = 0;
RUBY_ASSERT(OBJ_TOO_COMPLEX_SHAPE_ID == (GET_SHAPE_TREE()->next_shape_id - 1));
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
RUBY_ASSERT(rb_shape_id(hash_fallback_shape) == OBJ_TOO_COMPLEX_SHAPE_ID);
}
void
Init_shape(void)
{
#if SHAPE_DEBUG
VALUE rb_cShape = rb_struct_define_under(rb_cRubyVM, "Shape",
"id",
"parent_id",
"edge_name",
"next_iv_index",
"size_pool_index",
"type",
"capacity",
NULL);
rb_define_method(rb_cShape, "parent", rb_shape_parent, 0);
rb_define_method(rb_cShape, "edges", rb_shape_edges, 0);
rb_define_method(rb_cShape, "depth", rb_shape_export_depth, 0);
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
rb_define_method(rb_cShape, "too_complex?", rb_shape_too_complex, 0);
rb_define_const(rb_cShape, "SHAPE_ROOT", INT2NUM(SHAPE_ROOT));
rb_define_const(rb_cShape, "SHAPE_IVAR", INT2NUM(SHAPE_IVAR));
rb_define_const(rb_cShape, "SHAPE_T_OBJECT", INT2NUM(SHAPE_T_OBJECT));
rb_define_const(rb_cShape, "SHAPE_FROZEN", INT2NUM(SHAPE_FROZEN));
2022-11-18 21:29:41 +03:00
rb_define_const(rb_cShape, "SHAPE_ID_NUM_BITS", INT2NUM(SHAPE_ID_NUM_BITS));
rb_define_const(rb_cShape, "SHAPE_FLAG_SHIFT", INT2NUM(SHAPE_FLAG_SHIFT));
rb_define_const(rb_cShape, "SPECIAL_CONST_SHAPE_ID", INT2NUM(SPECIAL_CONST_SHAPE_ID));
Transition complex objects to "too complex" shape When an object becomes "too complex" (in other words it has too many variations in the shape tree), we transition it to use a "too complex" shape and use a hash for storing instance variables. Without this patch, there were rare cases where shape tree growth could "explode" and cause performance degradation on what would otherwise have been cached fast paths. This patch puts a limit on shape tree growth, and gracefully degrades in the rare case where there could be a factorial growth in the shape tree. For example: ```ruby class NG; end HUGE_NUMBER.times do NG.new.instance_variable_set(:"@unique_ivar_#{_1}", 1) end ``` We consider objects to be "too complex" when the object's class has more than SHAPE_MAX_VARIATIONS (currently 8) leaf nodes in the shape tree and the object introduces a new variation (a new leaf node) associated with that class. For example, new variations on instances of the following class would be considered "too complex" because those instances create more than 8 leaves in the shape tree: ```ruby class Foo; end 9.times { Foo.new.instance_variable_set(":@uniq_#{_1}", 1) } ``` However, the following class is *not* too complex because it only has one leaf in the shape tree: ```ruby class Foo def initialize @a = @b = @c = @d = @e = @f = @g = @h = @i = nil end end 9.times { Foo.new } `` This case is rare, so we don't expect this change to impact performance of most applications, but it needs to be handled. Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
2022-12-09 01:16:52 +03:00
rb_define_const(rb_cShape, "OBJ_TOO_COMPLEX_SHAPE_ID", INT2NUM(OBJ_TOO_COMPLEX_SHAPE_ID));
rb_define_const(rb_cShape, "SHAPE_MAX_VARIATIONS", INT2NUM(SHAPE_MAX_VARIATIONS));
rb_define_const(rb_cShape, "SIZEOF_RB_SHAPE_T", INT2NUM(sizeof(rb_shape_t)));
rb_define_const(rb_cShape, "SIZEOF_REDBLACK_NODE_T", INT2NUM(sizeof(redblack_node_t)));
rb_define_const(rb_cShape, "SHAPE_BUFFER_SIZE", INT2NUM(sizeof(rb_shape_t) * SHAPE_BUFFER_SIZE));
rb_define_const(rb_cShape, "REDBLACK_CACHE_SIZE", INT2NUM(sizeof(redblack_node_t) * REDBLACK_CACHE_SIZE));
rb_define_singleton_method(rb_cShape, "transition_tree", shape_transition_tree, 0);
rb_define_singleton_method(rb_cShape, "find_by_id", rb_shape_find_by_id, 1);
rb_define_singleton_method(rb_cShape, "of", rb_shape_debug_shape, 1);
rb_define_singleton_method(rb_cShape, "root_shape", rb_shape_root_shape, 0);
rb_define_singleton_method(rb_cShape, "shapes_available", rb_shape_shapes_available, 0);
#endif
}