ruby/gc.c

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

5602 lines
153 KiB
C
Raw Permalink Normal View History

/**********************************************************************
gc.c -
$Author$
created at: Tue Oct 5 09:44:46 JST 1993
* encoding.c: provide basic features for M17N. * parse.y: encoding aware parsing. * parse.y (pragma_encoding): encoding specification pragma. * parse.y (rb_intern3): encoding specified symbols. * string.c (rb_str_length): length based on characters. for older behavior, bytesize method added. * string.c (rb_str_index_m): index based on characters. rindex as well. * string.c (succ_char): encoding aware succeeding string. * string.c (rb_str_reverse): reverse based on characters. * string.c (rb_str_inspect): encoding aware string description. * string.c (rb_str_upcase_bang): encoding aware case conversion. downcase, capitalize, swapcase as well. * string.c (rb_str_tr_bang): tr based on characters. delete, squeeze, tr_s, count as well. * string.c (rb_str_split_m): split based on characters. * string.c (rb_str_each_line): encoding aware each_line. * string.c (rb_str_each_char): added. iteration based on characters. * string.c (rb_str_strip_bang): encoding aware whitespace stripping. lstrip, rstrip as well. * string.c (rb_str_justify): encoding aware justifying (ljust, rjust, center). * string.c (str_encoding): get encoding attribute from a string. * re.c (rb_reg_initialize): encoding aware regular expression * sprintf.c (rb_str_format): formatting (i.e. length count) based on characters. * io.c (rb_io_getc): getc to return one-character string. for older behavior, getbyte method added. * ext/stringio/stringio.c (strio_getc): ditto. * io.c (rb_io_ungetc): allow pushing arbitrary string at the current reading point. * ext/stringio/stringio.c (strio_ungetc): ditto. * ext/strscan/strscan.c: encoding support. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@13261 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2007-08-25 03:29:39 +00:00
Copyright (C) 1993-2007 Yukihiro Matsumoto
Copyright (C) 2000 Network Applied Communication Laboratory, Inc.
Copyright (C) 2000 Information-technology Promotion Agency, Japan
**********************************************************************/
#define rb_data_object_alloc rb_data_object_alloc
#define rb_data_typed_object_alloc rb_data_typed_object_alloc
#include "ruby/internal/config.h"
#ifdef _WIN32
# include "ruby/ruby.h"
#endif
#if defined(__wasm__) && !defined(__EMSCRIPTEN__)
# include "wasm/setjmp.h"
# include "wasm/machine.h"
#else
# include <setjmp.h>
#endif
#include <stdarg.h>
#include <stdio.h>
/* MALLOC_HEADERS_BEGIN */
#ifndef HAVE_MALLOC_USABLE_SIZE
# ifdef _WIN32
# define HAVE_MALLOC_USABLE_SIZE
# define malloc_usable_size(a) _msize(a)
# elif defined HAVE_MALLOC_SIZE
# define HAVE_MALLOC_USABLE_SIZE
# define malloc_usable_size(a) malloc_size(a)
# endif
#endif
#ifdef HAVE_MALLOC_USABLE_SIZE
# ifdef RUBY_ALTERNATIVE_MALLOC_HEADER
2021-06-13 14:34:46 +09:00
/* Alternative malloc header is included in ruby/missing.h */
# elif defined(HAVE_MALLOC_H)
# include <malloc.h>
# elif defined(HAVE_MALLOC_NP_H)
# include <malloc_np.h>
# elif defined(HAVE_MALLOC_MALLOC_H)
# include <malloc/malloc.h>
# endif
#endif
/* MALLOC_HEADERS_END */
#ifdef HAVE_SYS_TIME_H
# include <sys/time.h>
#endif
#ifdef HAVE_SYS_RESOURCE_H
# include <sys/resource.h>
#endif
#if defined _WIN32 || defined __CYGWIN__
# include <windows.h>
#elif defined(HAVE_POSIX_MEMALIGN)
#elif defined(HAVE_MEMALIGN)
# include <malloc.h>
#endif
#include <sys/types.h>
#ifdef __EMSCRIPTEN__
#include <emscripten.h>
#endif
/* For ruby_annotate_mmap */
#ifdef HAVE_SYS_PRCTL_H
#include <sys/prctl.h>
#endif
#undef LIST_HEAD /* ccan/list conflicts with BSD-origin sys/queue.h. */
#include "constant.h"
#include "darray.h"
#include "debug_counter.h"
#include "eval_intern.h"
#include "gc/gc.h"
#include "id_table.h"
#include "internal.h"
#include "internal/class.h"
#include "internal/compile.h"
#include "internal/complex.h"
#include "internal/cont.h"
#include "internal/error.h"
#include "internal/eval.h"
#include "internal/gc.h"
#include "internal/hash.h"
#include "internal/imemo.h"
#include "internal/io.h"
#include "internal/numeric.h"
#include "internal/object.h"
#include "internal/proc.h"
#include "internal/rational.h"
#include "internal/sanitizers.h"
#include "internal/struct.h"
#include "internal/symbol.h"
#include "internal/thread.h"
#include "internal/variable.h"
#include "internal/warnings.h"
#include "probes.h"
#include "regint.h"
#include "ruby/debug.h"
#include "ruby/io.h"
#include "ruby/re.h"
#include "ruby/st.h"
#include "ruby/thread.h"
#include "ruby/util.h"
#include "ruby/vm.h"
#include "ruby_assert.h"
#include "ruby_atomic.h"
#include "symbol.h"
#include "variable.h"
#include "vm_core.h"
#include "vm_sync.h"
#include "vm_callinfo.h"
#include "ractor_core.h"
#include "yjit.h"
#include "builtin.h"
#include "shape.h"
unsigned int
rb_gc_vm_lock(const char *file, int line)
{
unsigned int lev = 0;
rb_vm_lock_enter(&lev, file, line);
return lev;
}
void
rb_gc_vm_unlock(unsigned int lev, const char *file, int line)
{
rb_vm_lock_leave(&lev, file, line);
}
unsigned int
rb_gc_cr_lock(const char *file, int line)
{
unsigned int lev;
rb_vm_lock_enter_cr(GET_RACTOR(), &lev, file, line);
return lev;
}
void
rb_gc_cr_unlock(unsigned int lev, const char *file, int line)
{
rb_vm_lock_leave_cr(GET_RACTOR(), &lev, file, line);
}
unsigned int
rb_gc_vm_lock_no_barrier(const char *file, int line)
{
unsigned int lev = 0;
rb_vm_lock_enter_nb(&lev, file, line);
return lev;
}
void
rb_gc_vm_unlock_no_barrier(unsigned int lev, const char *file, int line)
{
rb_vm_lock_leave_nb(&lev, file, line);
}
void
rb_gc_vm_barrier(void)
{
rb_vm_barrier();
}
#if USE_MODULAR_GC
void *
rb_gc_get_ractor_newobj_cache(void)
{
return GET_RACTOR()->newobj_cache;
}
void
rb_gc_initialize_vm_context(struct rb_gc_vm_context *context)
{
rb_native_mutex_initialize(&context->lock);
context->ec = GET_EC();
}
void
rb_gc_worker_thread_set_vm_context(struct rb_gc_vm_context *context)
{
rb_native_mutex_lock(&context->lock);
GC_ASSERT(rb_current_execution_context(false) == NULL);
#ifdef RB_THREAD_LOCAL_SPECIFIER
rb_current_ec_set(context->ec);
#else
native_tls_set(ruby_current_ec_key, context->ec);
#endif
}
void
rb_gc_worker_thread_unset_vm_context(struct rb_gc_vm_context *context)
{
rb_native_mutex_unlock(&context->lock);
GC_ASSERT(rb_current_execution_context(true) == context->ec);
#ifdef RB_THREAD_LOCAL_SPECIFIER
rb_current_ec_set(NULL);
#else
native_tls_set(ruby_current_ec_key, NULL);
#endif
}
#endif
bool
rb_gc_event_hook_required_p(rb_event_flag_t event)
{
return ruby_vm_event_flags & event;
}
void
rb_gc_event_hook(VALUE obj, rb_event_flag_t event)
{
if (LIKELY(!rb_gc_event_hook_required_p(event))) return;
rb_execution_context_t *ec = GET_EC();
if (!ec->cfp) return;
EXEC_EVENT_HOOK(ec, event, ec->cfp->self, 0, 0, 0, obj);
}
void *
rb_gc_get_objspace(void)
{
return GET_VM()->gc.objspace;
}
void
rb_gc_ractor_newobj_cache_foreach(void (*func)(void *cache, void *data), void *data)
{
rb_ractor_t *r = NULL;
if (RB_LIKELY(ruby_single_main_ractor)) {
GC_ASSERT(
ccan_list_empty(&GET_VM()->ractor.set) ||
(ccan_list_top(&GET_VM()->ractor.set, rb_ractor_t, vmlr_node) == ruby_single_main_ractor &&
ccan_list_tail(&GET_VM()->ractor.set, rb_ractor_t, vmlr_node) == ruby_single_main_ractor)
);
func(ruby_single_main_ractor->newobj_cache, data);
}
else {
ccan_list_for_each(&GET_VM()->ractor.set, r, vmlr_node) {
func(r->newobj_cache, data);
}
}
}
void
rb_gc_run_obj_finalizer(VALUE objid, long count, VALUE (*callback)(long i, void *data), void *data)
{
volatile struct {
VALUE errinfo;
VALUE final;
rb_control_frame_t *cfp;
VALUE *sp;
long finished;
} saved;
rb_execution_context_t * volatile ec = GET_EC();
#define RESTORE_FINALIZER() (\
ec->cfp = saved.cfp, \
ec->cfp->sp = saved.sp, \
ec->errinfo = saved.errinfo)
saved.errinfo = ec->errinfo;
saved.cfp = ec->cfp;
saved.sp = ec->cfp->sp;
saved.finished = 0;
saved.final = Qundef;
rb_ractor_ignore_belonging(true);
EC_PUSH_TAG(ec);
enum ruby_tag_type state = EC_EXEC_TAG();
if (state != TAG_NONE) {
++saved.finished; /* skip failed finalizer */
VALUE failed_final = saved.final;
saved.final = Qundef;
if (!UNDEF_P(failed_final) && !NIL_P(ruby_verbose)) {
rb_warn("Exception in finalizer %+"PRIsVALUE, failed_final);
rb_ec_error_print(ec, ec->errinfo);
}
}
for (long i = saved.finished; RESTORE_FINALIZER(), i < count; saved.finished = ++i) {
saved.final = callback(i, data);
rb_check_funcall(saved.final, idCall, 1, &objid);
}
EC_POP_TAG();
rb_ractor_ignore_belonging(false);
#undef RESTORE_FINALIZER
}
void
rb_gc_set_pending_interrupt(void)
{
rb_execution_context_t *ec = GET_EC();
ec->interrupt_mask |= PENDING_INTERRUPT_MASK;
}
void
rb_gc_unset_pending_interrupt(void)
{
rb_execution_context_t *ec = GET_EC();
ec->interrupt_mask &= ~PENDING_INTERRUPT_MASK;
}
bool
rb_gc_multi_ractor_p(void)
{
return rb_multi_ractor_p();
}
bool rb_obj_is_main_ractor(VALUE gv);
bool
rb_gc_shutdown_call_finalizer_p(VALUE obj)
{
switch (BUILTIN_TYPE(obj)) {
case T_DATA:
if (!ruby_free_at_exit_p() && (!DATA_PTR(obj) || !RDATA(obj)->dfree)) return false;
if (rb_obj_is_thread(obj)) return false;
if (rb_obj_is_mutex(obj)) return false;
if (rb_obj_is_fiber(obj)) return false;
if (rb_obj_is_main_ractor(obj)) return false;
Lock-free hash set for fstrings [Feature #21268] This implements a hash set which is wait-free for lookup and lock-free for insert (unless resizing) to use for fstring de-duplication. As highlighted in https://bugs.ruby-lang.org/issues/19288, heavy use of fstrings (frozen interned strings) can significantly reduce the parallelism of Ractors. I tried a few other approaches first: using an RWLock, striping a series of RWlocks (partitioning the hash N-ways to reduce lock contention), and putting a cache in front of it. All of these improved the situation, but were unsatisfying as all still required locks for writes (and granular locks are awkward, since we run the risk of needing to reach a vm barrier) and this table is somewhat write-heavy. My main reference for this was Cliff Click's talk on a lock free hash-table for java https://www.youtube.com/watch?v=HJ-719EGIts. It turns out this lock-free hash set is made easier to implement by a few properties: * We only need a hash set rather than a hash table (we only need keys, not values), and so the full entry can be written as a single VALUE * As a set we only need lookup/insert/delete, no update * Delete is only run inside GC so does not need to be atomic (It could be made concurrent) * I use rb_vm_barrier for the (rare) table rebuilds (It could be made concurrent) We VM lock (but don't require other threads to stop) for table rebuilds, as those are rare * The conservative garbage collector makes deferred replication easy, using a T_DATA object Another benefits of having a table specific to fstrings is that we compare by value on lookup/insert, but by identity on delete, as we only want to remove the exact string which is being freed. This is faster and provides a second way to avoid the race condition in https://bugs.ruby-lang.org/issues/21172. This is a pretty standard open-addressing hash table with quadratic probing. Similar to our existing st_table or id_table. Deletes (which happen on GC) replace existing keys with a tombstone, which is the only type of update which can occur. Tombstones are only cleared out on resize. Unlike st_table, the VALUEs are stored in the hash table itself (st_table's bins) rather than as a compact index. This avoids an extra pointer dereference and is possible because we don't need to preserve insertion order. The table targets a load factor of 2 (it is enlarged once it is half full).
2025-03-06 16:16:35 -08:00
if (rb_obj_is_fstring_table(obj)) return false;
return true;
case T_FILE:
return true;
case T_SYMBOL:
if (RSYMBOL(obj)->fstr &&
(BUILTIN_TYPE(RSYMBOL(obj)->fstr) == T_NONE ||
BUILTIN_TYPE(RSYMBOL(obj)->fstr) == T_ZOMBIE)) {
RSYMBOL(obj)->fstr = 0;
}
return true;
case T_NONE:
return false;
default:
return ruby_free_at_exit_p();
}
}
uint32_t
rb_gc_get_shape(VALUE obj)
{
return (uint32_t)rb_obj_shape_id(obj);
}
void
rb_gc_set_shape(VALUE obj, uint32_t shape_id)
{
rb_obj_set_shape_id(obj, (uint32_t)shape_id);
}
uint32_t
rb_gc_rebuild_shape(VALUE obj, size_t heap_id)
{
RUBY_ASSERT(RB_TYPE_P(obj, T_OBJECT));
return (uint32_t)rb_shape_transition_heap(obj, heap_id);
}
void rb_vm_update_references(void *ptr);
#define rb_setjmp(env) RUBY_SETJMP(env)
#define rb_jmp_buf rb_jmpbuf_t
#undef rb_data_object_wrap
#if !defined(MAP_ANONYMOUS) && defined(MAP_ANON)
#define MAP_ANONYMOUS MAP_ANON
#endif
#define unless_objspace(objspace) \
void *objspace; \
rb_vm_t *unless_objspace_vm = GET_VM(); \
if (unless_objspace_vm) objspace = unless_objspace_vm->gc.objspace; \
else /* return; or objspace will be warned uninitialized */
#define RMOVED(obj) ((struct RMoved *)(obj))
#define TYPED_UPDATE_IF_MOVED(_objspace, _type, _thing) do { \
if (rb_gc_impl_object_moved_p((_objspace), (VALUE)(_thing))) { \
*(_type *)&(_thing) = (_type)gc_location_internal(_objspace, (VALUE)_thing); \
} \
} while (0)
#define UPDATE_IF_MOVED(_objspace, _thing) TYPED_UPDATE_IF_MOVED(_objspace, VALUE, _thing)
#if RUBY_MARK_FREE_DEBUG
int ruby_gc_debug_indent = 0;
#endif
#ifndef RGENGC_OBJ_INFO
# define RGENGC_OBJ_INFO RGENGC_CHECK_MODE
#endif
#ifndef CALC_EXACT_MALLOC_SIZE
# define CALC_EXACT_MALLOC_SIZE 0
#endif
VALUE rb_mGC;
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 15:32:12 +02:00
static size_t malloc_offset = 0;
#if defined(HAVE_MALLOC_USABLE_SIZE)
static size_t
gc_compute_malloc_offset(void)
{
// Different allocators use different metadata storage strategies which result in different
// ideal sizes.
// For instance malloc(64) will waste 8B with glibc, but waste 0B with jemalloc.
// But malloc(56) will waste 0B with glibc, but waste 8B with jemalloc.
// So we try allocating 64, 56 and 48 bytes and select the first offset that doesn't
// waste memory.
// This was tested on Linux with glibc 2.35 and jemalloc 5, and for both it result in
// no wasted memory.
size_t offset = 0;
for (offset = 0; offset <= 16; offset += 8) {
size_t allocated = (64 - offset);
void *test_ptr = malloc(allocated);
size_t wasted = malloc_usable_size(test_ptr) - allocated;
free(test_ptr);
if (wasted == 0) {
return offset;
}
}
return 0;
}
#else
static size_t
gc_compute_malloc_offset(void)
{
// If we don't have malloc_usable_size, we use powers of 2.
return 0;
}
#endif
size_t
rb_malloc_grow_capa(size_t current, size_t type_size)
{
size_t current_capacity = current;
if (current_capacity < 4) {
current_capacity = 4;
}
current_capacity *= type_size;
// We double the current capacity.
size_t new_capacity = (current_capacity * 2);
// And round up to the next power of 2 if it's not already one.
if (rb_popcount64(new_capacity) != 1) {
new_capacity = (size_t)(1 << (64 - nlz_int64(new_capacity)));
}
new_capacity -= malloc_offset;
new_capacity /= type_size;
if (current > new_capacity) {
rb_bug("rb_malloc_grow_capa: current_capacity=%zu, new_capacity=%zu, malloc_offset=%zu", current, new_capacity, malloc_offset);
}
RUBY_ASSERT(new_capacity > current);
return new_capacity;
}
2020-05-04 16:27:48 +09:00
static inline struct rbimpl_size_mul_overflow_tag
size_mul_add_overflow(size_t x, size_t y, size_t z) /* x * y + z */
{
2020-05-04 16:27:48 +09:00
struct rbimpl_size_mul_overflow_tag t = rbimpl_size_mul_overflow(x, y);
struct rbimpl_size_mul_overflow_tag u = rbimpl_size_add_overflow(t.right, z);
2020-05-04 16:27:48 +09:00
return (struct rbimpl_size_mul_overflow_tag) { t.left || u.left, u.right };
}
2020-05-04 16:27:48 +09:00
static inline struct rbimpl_size_mul_overflow_tag
size_mul_add_mul_overflow(size_t x, size_t y, size_t z, size_t w) /* x * y + z * w */
{
2020-05-04 16:27:48 +09:00
struct rbimpl_size_mul_overflow_tag t = rbimpl_size_mul_overflow(x, y);
struct rbimpl_size_mul_overflow_tag u = rbimpl_size_mul_overflow(z, w);
struct rbimpl_size_mul_overflow_tag v = rbimpl_size_add_overflow(t.right, u.right);
2020-05-04 16:27:48 +09:00
return (struct rbimpl_size_mul_overflow_tag) { t.left || u.left || v.left, v.right };
}
PRINTF_ARGS(NORETURN(static void gc_raise(VALUE, const char*, ...)), 2, 3);
static inline size_t
size_mul_or_raise(size_t x, size_t y, VALUE exc)
{
2020-05-04 16:27:48 +09:00
struct rbimpl_size_mul_overflow_tag t = rbimpl_size_mul_overflow(x, y);
if (LIKELY(!t.left)) {
return t.right;
}
else if (rb_during_gc()) {
rb_memerror(); /* or...? */
}
else {
gc_raise(
exc,
"integer overflow: %"PRIuSIZE
" * %"PRIuSIZE
" > %"PRIuSIZE,
x, y, (size_t)SIZE_MAX);
}
}
size_t
rb_size_mul_or_raise(size_t x, size_t y, VALUE exc)
{
return size_mul_or_raise(x, y, exc);
}
static inline size_t
size_mul_add_or_raise(size_t x, size_t y, size_t z, VALUE exc)
{
2020-05-04 16:27:48 +09:00
struct rbimpl_size_mul_overflow_tag t = size_mul_add_overflow(x, y, z);
if (LIKELY(!t.left)) {
return t.right;
}
else if (rb_during_gc()) {
rb_memerror(); /* or...? */
}
else {
gc_raise(
exc,
"integer overflow: %"PRIuSIZE
" * %"PRIuSIZE
" + %"PRIuSIZE
" > %"PRIuSIZE,
x, y, z, (size_t)SIZE_MAX);
}
}
size_t
rb_size_mul_add_or_raise(size_t x, size_t y, size_t z, VALUE exc)
{
return size_mul_add_or_raise(x, y, z, exc);
}
static inline size_t
size_mul_add_mul_or_raise(size_t x, size_t y, size_t z, size_t w, VALUE exc)
{
2020-05-04 16:27:48 +09:00
struct rbimpl_size_mul_overflow_tag t = size_mul_add_mul_overflow(x, y, z, w);
if (LIKELY(!t.left)) {
return t.right;
}
else if (rb_during_gc()) {
rb_memerror(); /* or...? */
}
else {
gc_raise(
exc,
"integer overflow: %"PRIdSIZE
" * %"PRIdSIZE
" + %"PRIdSIZE
" * %"PRIdSIZE
" > %"PRIdSIZE,
x, y, z, w, (size_t)SIZE_MAX);
}
}
#if defined(HAVE_RB_GC_GUARDED_PTR_VAL) && HAVE_RB_GC_GUARDED_PTR_VAL
/* trick the compiler into thinking a external signal handler uses this */
volatile VALUE rb_gc_guarded_val;
volatile VALUE *
rb_gc_guarded_ptr_val(volatile VALUE *ptr, VALUE val)
{
rb_gc_guarded_val = val;
return ptr;
}
#endif
static const char *obj_type_name(VALUE obj);
#include "gc/default/default.c"
#if USE_MODULAR_GC && !defined(HAVE_DLOPEN)
# error "Modular GC requires dlopen"
#elif USE_MODULAR_GC
#include <dlfcn.h>
typedef struct gc_function_map {
// Bootup
void *(*objspace_alloc)(void);
void (*objspace_init)(void *objspace_ptr);
void *(*ractor_cache_alloc)(void *objspace_ptr, void *ractor);
void (*set_params)(void *objspace_ptr);
void (*init)(void);
size_t *(*heap_sizes)(void *objspace_ptr);
// Shutdown
void (*shutdown_free_objects)(void *objspace_ptr);
void (*objspace_free)(void *objspace_ptr);
void (*ractor_cache_free)(void *objspace_ptr, void *cache);
// GC
void (*start)(void *objspace_ptr, bool full_mark, bool immediate_mark, bool immediate_sweep, bool compact);
bool (*during_gc_p)(void *objspace_ptr);
void (*prepare_heap)(void *objspace_ptr);
void (*gc_enable)(void *objspace_ptr);
void (*gc_disable)(void *objspace_ptr, bool finish_current_gc);
bool (*gc_enabled_p)(void *objspace_ptr);
Provide GC.config to disable major GC collections This feature provides a new method `GC.config` that configures internal GC configuration variables provided by an individual GC implementation. Implemented in this PR is the option `full_mark`: a boolean value that will determine whether the Ruby GC is allowed to run a major collection while the process is running. It has the following semantics This feature configures Ruby's GC to only run minor GC's. It's designed to give users relying on Out of Band GC complete control over when a major GC is run. Configuring `full_mark: false` does two main things: * Never runs a Major GC. When the heap runs out of space during a minor and when a major would traditionally be run, instead we allocate more heap pages, and mark objspace as needing a major GC. * Don't increment object ages. We don't promote objects during GC, this will cause every object to be scanned on every minor. This is an intentional trade-off between minor GC's doing more work every time, and potentially promoting objects that will then never be GC'd. The intention behind not aging objects is that users of this feature should use a preforking web server, or some other method of pre-warming the oldgen (like Nakayoshi fork)before disabling Majors. That way most objects that are going to be old will have already been promoted. This will interleave major and minor GC collections in exactly the same what that the Ruby GC runs in versions previously to this. This is the default behaviour. * This new method has the following extra semantics: - `GC.config` with no arguments returns a hash of the keys of the currently configured GC - `GC.config` with a key pair (eg. `GC.config(full_mark: true)` sets the matching config key to the corresponding value and returns the entire known config hash, including the new values. If the key does not exist, `nil` is returned * When a minor GC is run, Ruby sets an internal status flag to determine whether the next GC will be a major or a minor. When `full_mark: false` this flag is ignored and every GC will be a minor. This status flag can be accessed at `GC.latest_gc_info(:needs_major_by)`. Any value other than `nil` means that the next collection would have been a major. Thus it's possible to use this feature to check at a predetermined time, whether a major GC is necessary and run one if it is. eg. After a request has finished processing. ```ruby if GC.latest_gc_info(:needs_major_by) GC.start(full_mark: true) end ``` [Feature #20443]
2024-07-04 15:21:09 +01:00
VALUE (*config_get)(void *objpace_ptr);
void (*config_set)(void *objspace_ptr, VALUE hash);
void (*stress_set)(void *objspace_ptr, VALUE flag);
VALUE (*stress_get)(void *objspace_ptr);
// Object allocation
VALUE (*new_obj)(void *objspace_ptr, void *cache_ptr, VALUE klass, VALUE flags, VALUE v1, VALUE v2, VALUE v3, bool wb_protected, size_t alloc_size);
size_t (*obj_slot_size)(VALUE obj);
size_t (*heap_id_for_size)(void *objspace_ptr, size_t size);
bool (*size_allocatable_p)(size_t size);
// Malloc
void *(*malloc)(void *objspace_ptr, size_t size);
void *(*calloc)(void *objspace_ptr, size_t size);
void *(*realloc)(void *objspace_ptr, void *ptr, size_t new_size, size_t old_size);
void (*free)(void *objspace_ptr, void *ptr, size_t old_size);
void (*adjust_memory_usage)(void *objspace_ptr, ssize_t diff);
// Marking
void (*mark)(void *objspace_ptr, VALUE obj);
void (*mark_and_move)(void *objspace_ptr, VALUE *ptr);
void (*mark_and_pin)(void *objspace_ptr, VALUE obj);
void (*mark_maybe)(void *objspace_ptr, VALUE obj);
void (*mark_weak)(void *objspace_ptr, VALUE *ptr);
void (*remove_weak)(void *objspace_ptr, VALUE parent_obj, VALUE *ptr);
// Compaction
bool (*object_moved_p)(void *objspace_ptr, VALUE obj);
VALUE (*location)(void *objspace_ptr, VALUE value);
// Write barriers
void (*writebarrier)(void *objspace_ptr, VALUE a, VALUE b);
void (*writebarrier_unprotect)(void *objspace_ptr, VALUE obj);
void (*writebarrier_remember)(void *objspace_ptr, VALUE obj);
// Heap walking
void (*each_objects)(void *objspace_ptr, int (*callback)(void *, void *, size_t, void *), void *data);
void (*each_object)(void *objspace_ptr, void (*func)(VALUE obj, void *data), void *data);
// Finalizers
void (*make_zombie)(void *objspace_ptr, VALUE obj, void (*dfree)(void *), void *data);
VALUE (*define_finalizer)(void *objspace_ptr, VALUE obj, VALUE block);
void (*undefine_finalizer)(void *objspace_ptr, VALUE obj);
void (*copy_finalizer)(void *objspace_ptr, VALUE dest, VALUE obj);
void (*shutdown_call_finalizer)(void *objspace_ptr);
// Forking
void (*before_fork)(void *objspace_ptr);
void (*after_fork)(void *objspace_ptr, rb_pid_t pid);
// Statistics
void (*set_measure_total_time)(void *objspace_ptr, VALUE flag);
bool (*get_measure_total_time)(void *objspace_ptr);
unsigned long long (*get_total_time)(void *objspace_ptr);
size_t (*gc_count)(void *objspace_ptr);
VALUE (*latest_gc_info)(void *objspace_ptr, VALUE key);
VALUE (*stat)(void *objspace_ptr, VALUE hash_or_sym);
VALUE (*stat_heap)(void *objspace_ptr, VALUE heap_name, VALUE hash_or_sym);
const char *(*active_gc_name)(void);
// Miscellaneous
struct rb_gc_object_metadata_entry *(*object_metadata)(void *objspace_ptr, VALUE obj);
bool (*pointer_to_heap_p)(void *objspace_ptr, const void *ptr);
bool (*garbage_object_p)(void *objspace_ptr, VALUE obj);
void (*set_event_hook)(void *objspace_ptr, const rb_event_flag_t event);
void (*copy_attributes)(void *objspace_ptr, VALUE dest, VALUE obj);
bool modular_gc_loaded_p;
} rb_gc_function_map_t;
static rb_gc_function_map_t rb_gc_functions;
# define RUBY_GC_LIBRARY "RUBY_GC_LIBRARY"
# define MODULAR_GC_DIR STRINGIZE(modular_gc_dir)
static void
ruby_modular_gc_init(void)
{
// Assert that the directory path ends with a /
RUBY_ASSERT_ALWAYS(MODULAR_GC_DIR[sizeof(MODULAR_GC_DIR) - 2] == '/');
const char *gc_so_file = getenv(RUBY_GC_LIBRARY);
rb_gc_function_map_t gc_functions = { 0 };
char *gc_so_path = NULL;
void *handle = NULL;
if (gc_so_file) {
/* Check to make sure that gc_so_file matches /[\w-_]+/ so that it does
* not load a shared object outside of the directory. */
for (size_t i = 0; i < strlen(gc_so_file); i++) {
char c = gc_so_file[i];
if (isalnum(c)) continue;
switch (c) {
case '-':
case '_':
break;
default:
fprintf(stderr, "Only alphanumeric, dash, and underscore is allowed in "RUBY_GC_LIBRARY"\n");
exit(1);
}
}
size_t gc_so_path_size = strlen(MODULAR_GC_DIR "librubygc." DLEXT) + strlen(gc_so_file) + 1;
#ifdef LOAD_RELATIVE
Dl_info dli;
size_t prefix_len = 0;
if (dladdr((void *)(uintptr_t)ruby_modular_gc_init, &dli)) {
const char *base = strrchr(dli.dli_fname, '/');
if (base) {
size_t tail = 0;
# define end_with_p(lit) \
(prefix_len >= (tail = rb_strlen_lit(lit)) && \
memcmp(base - tail, lit, tail) == 0)
prefix_len = base - dli.dli_fname;
if (end_with_p("/bin") || end_with_p("/lib")) {
prefix_len -= tail;
}
prefix_len += MODULAR_GC_DIR[0] != '/';
gc_so_path_size += prefix_len;
}
}
#endif
gc_so_path = alloca(gc_so_path_size);
{
size_t gc_so_path_idx = 0;
#define GC_SO_PATH_APPEND(str) do { \
gc_so_path_idx += strlcpy(gc_so_path + gc_so_path_idx, str, gc_so_path_size - gc_so_path_idx); \
} while (0)
#ifdef LOAD_RELATIVE
if (prefix_len > 0) {
memcpy(gc_so_path, dli.dli_fname, prefix_len);
gc_so_path_idx = prefix_len;
}
#endif
GC_SO_PATH_APPEND(MODULAR_GC_DIR "librubygc.");
GC_SO_PATH_APPEND(gc_so_file);
GC_SO_PATH_APPEND(DLEXT);
GC_ASSERT(gc_so_path_idx == gc_so_path_size - 1);
#undef GC_SO_PATH_APPEND
}
handle = dlopen(gc_so_path, RTLD_LAZY | RTLD_GLOBAL);
if (!handle) {
fprintf(stderr, "ruby_modular_gc_init: Shared library %s cannot be opened: %s\n", gc_so_path, dlerror());
exit(1);
}
gc_functions.modular_gc_loaded_p = true;
}
# define load_modular_gc_func(name) do { \
if (handle) { \
const char *func_name = "rb_gc_impl_" #name; \
gc_functions.name = dlsym(handle, func_name); \
if (!gc_functions.name) { \
fprintf(stderr, "ruby_modular_gc_init: %s function not exported by library %s\n", func_name, gc_so_path); \
exit(1); \
} \
} \
else { \
gc_functions.name = rb_gc_impl_##name; \
} \
} while (0)
// Bootup
load_modular_gc_func(objspace_alloc);
load_modular_gc_func(objspace_init);
load_modular_gc_func(ractor_cache_alloc);
load_modular_gc_func(set_params);
load_modular_gc_func(init);
load_modular_gc_func(heap_sizes);
// Shutdown
load_modular_gc_func(shutdown_free_objects);
load_modular_gc_func(objspace_free);
load_modular_gc_func(ractor_cache_free);
// GC
load_modular_gc_func(start);
load_modular_gc_func(during_gc_p);
load_modular_gc_func(prepare_heap);
load_modular_gc_func(gc_enable);
load_modular_gc_func(gc_disable);
load_modular_gc_func(gc_enabled_p);
load_modular_gc_func(config_set);
load_modular_gc_func(config_get);
load_modular_gc_func(stress_set);
load_modular_gc_func(stress_get);
// Object allocation
load_modular_gc_func(new_obj);
load_modular_gc_func(obj_slot_size);
load_modular_gc_func(heap_id_for_size);
load_modular_gc_func(size_allocatable_p);
// Malloc
load_modular_gc_func(malloc);
load_modular_gc_func(calloc);
load_modular_gc_func(realloc);
load_modular_gc_func(free);
load_modular_gc_func(adjust_memory_usage);
// Marking
load_modular_gc_func(mark);
load_modular_gc_func(mark_and_move);
load_modular_gc_func(mark_and_pin);
load_modular_gc_func(mark_maybe);
load_modular_gc_func(mark_weak);
load_modular_gc_func(remove_weak);
// Compaction
load_modular_gc_func(object_moved_p);
load_modular_gc_func(location);
// Write barriers
load_modular_gc_func(writebarrier);
load_modular_gc_func(writebarrier_unprotect);
load_modular_gc_func(writebarrier_remember);
// Heap walking
load_modular_gc_func(each_objects);
load_modular_gc_func(each_object);
// Finalizers
load_modular_gc_func(make_zombie);
load_modular_gc_func(define_finalizer);
load_modular_gc_func(undefine_finalizer);
load_modular_gc_func(copy_finalizer);
load_modular_gc_func(shutdown_call_finalizer);
// Forking
load_modular_gc_func(before_fork);
load_modular_gc_func(after_fork);
// Statistics
load_modular_gc_func(set_measure_total_time);
load_modular_gc_func(get_measure_total_time);
load_modular_gc_func(get_total_time);
load_modular_gc_func(gc_count);
load_modular_gc_func(latest_gc_info);
load_modular_gc_func(stat);
load_modular_gc_func(stat_heap);
load_modular_gc_func(active_gc_name);
// Miscellaneous
load_modular_gc_func(object_metadata);
load_modular_gc_func(pointer_to_heap_p);
load_modular_gc_func(garbage_object_p);
load_modular_gc_func(set_event_hook);
load_modular_gc_func(copy_attributes);
# undef load_modular_gc_func
rb_gc_functions = gc_functions;
}
// Bootup
# define rb_gc_impl_objspace_alloc rb_gc_functions.objspace_alloc
# define rb_gc_impl_objspace_init rb_gc_functions.objspace_init
# define rb_gc_impl_ractor_cache_alloc rb_gc_functions.ractor_cache_alloc
# define rb_gc_impl_set_params rb_gc_functions.set_params
# define rb_gc_impl_init rb_gc_functions.init
# define rb_gc_impl_heap_sizes rb_gc_functions.heap_sizes
// Shutdown
# define rb_gc_impl_shutdown_free_objects rb_gc_functions.shutdown_free_objects
# define rb_gc_impl_objspace_free rb_gc_functions.objspace_free
# define rb_gc_impl_ractor_cache_free rb_gc_functions.ractor_cache_free
// GC
# define rb_gc_impl_start rb_gc_functions.start
# define rb_gc_impl_during_gc_p rb_gc_functions.during_gc_p
# define rb_gc_impl_prepare_heap rb_gc_functions.prepare_heap
# define rb_gc_impl_gc_enable rb_gc_functions.gc_enable
# define rb_gc_impl_gc_disable rb_gc_functions.gc_disable
# define rb_gc_impl_gc_enabled_p rb_gc_functions.gc_enabled_p
Provide GC.config to disable major GC collections This feature provides a new method `GC.config` that configures internal GC configuration variables provided by an individual GC implementation. Implemented in this PR is the option `full_mark`: a boolean value that will determine whether the Ruby GC is allowed to run a major collection while the process is running. It has the following semantics This feature configures Ruby's GC to only run minor GC's. It's designed to give users relying on Out of Band GC complete control over when a major GC is run. Configuring `full_mark: false` does two main things: * Never runs a Major GC. When the heap runs out of space during a minor and when a major would traditionally be run, instead we allocate more heap pages, and mark objspace as needing a major GC. * Don't increment object ages. We don't promote objects during GC, this will cause every object to be scanned on every minor. This is an intentional trade-off between minor GC's doing more work every time, and potentially promoting objects that will then never be GC'd. The intention behind not aging objects is that users of this feature should use a preforking web server, or some other method of pre-warming the oldgen (like Nakayoshi fork)before disabling Majors. That way most objects that are going to be old will have already been promoted. This will interleave major and minor GC collections in exactly the same what that the Ruby GC runs in versions previously to this. This is the default behaviour. * This new method has the following extra semantics: - `GC.config` with no arguments returns a hash of the keys of the currently configured GC - `GC.config` with a key pair (eg. `GC.config(full_mark: true)` sets the matching config key to the corresponding value and returns the entire known config hash, including the new values. If the key does not exist, `nil` is returned * When a minor GC is run, Ruby sets an internal status flag to determine whether the next GC will be a major or a minor. When `full_mark: false` this flag is ignored and every GC will be a minor. This status flag can be accessed at `GC.latest_gc_info(:needs_major_by)`. Any value other than `nil` means that the next collection would have been a major. Thus it's possible to use this feature to check at a predetermined time, whether a major GC is necessary and run one if it is. eg. After a request has finished processing. ```ruby if GC.latest_gc_info(:needs_major_by) GC.start(full_mark: true) end ``` [Feature #20443]
2024-07-04 15:21:09 +01:00
# define rb_gc_impl_config_get rb_gc_functions.config_get
# define rb_gc_impl_config_set rb_gc_functions.config_set
# define rb_gc_impl_stress_set rb_gc_functions.stress_set
# define rb_gc_impl_stress_get rb_gc_functions.stress_get
// Object allocation
# define rb_gc_impl_new_obj rb_gc_functions.new_obj
# define rb_gc_impl_obj_slot_size rb_gc_functions.obj_slot_size
# define rb_gc_impl_heap_id_for_size rb_gc_functions.heap_id_for_size
# define rb_gc_impl_size_allocatable_p rb_gc_functions.size_allocatable_p
// Malloc
# define rb_gc_impl_malloc rb_gc_functions.malloc
# define rb_gc_impl_calloc rb_gc_functions.calloc
# define rb_gc_impl_realloc rb_gc_functions.realloc
# define rb_gc_impl_free rb_gc_functions.free
# define rb_gc_impl_adjust_memory_usage rb_gc_functions.adjust_memory_usage
// Marking
# define rb_gc_impl_mark rb_gc_functions.mark
# define rb_gc_impl_mark_and_move rb_gc_functions.mark_and_move
# define rb_gc_impl_mark_and_pin rb_gc_functions.mark_and_pin
# define rb_gc_impl_mark_maybe rb_gc_functions.mark_maybe
# define rb_gc_impl_mark_weak rb_gc_functions.mark_weak
# define rb_gc_impl_remove_weak rb_gc_functions.remove_weak
// Compaction
# define rb_gc_impl_object_moved_p rb_gc_functions.object_moved_p
# define rb_gc_impl_location rb_gc_functions.location
// Write barriers
# define rb_gc_impl_writebarrier rb_gc_functions.writebarrier
# define rb_gc_impl_writebarrier_unprotect rb_gc_functions.writebarrier_unprotect
# define rb_gc_impl_writebarrier_remember rb_gc_functions.writebarrier_remember
// Heap walking
# define rb_gc_impl_each_objects rb_gc_functions.each_objects
# define rb_gc_impl_each_object rb_gc_functions.each_object
// Finalizers
# define rb_gc_impl_make_zombie rb_gc_functions.make_zombie
# define rb_gc_impl_define_finalizer rb_gc_functions.define_finalizer
# define rb_gc_impl_undefine_finalizer rb_gc_functions.undefine_finalizer
# define rb_gc_impl_copy_finalizer rb_gc_functions.copy_finalizer
# define rb_gc_impl_shutdown_call_finalizer rb_gc_functions.shutdown_call_finalizer
// Forking
# define rb_gc_impl_before_fork rb_gc_functions.before_fork
# define rb_gc_impl_after_fork rb_gc_functions.after_fork
// Statistics
# define rb_gc_impl_set_measure_total_time rb_gc_functions.set_measure_total_time
# define rb_gc_impl_get_measure_total_time rb_gc_functions.get_measure_total_time
# define rb_gc_impl_get_total_time rb_gc_functions.get_total_time
# define rb_gc_impl_gc_count rb_gc_functions.gc_count
# define rb_gc_impl_latest_gc_info rb_gc_functions.latest_gc_info
# define rb_gc_impl_stat rb_gc_functions.stat
# define rb_gc_impl_stat_heap rb_gc_functions.stat_heap
# define rb_gc_impl_active_gc_name rb_gc_functions.active_gc_name
// Miscellaneous
# define rb_gc_impl_object_metadata rb_gc_functions.object_metadata
# define rb_gc_impl_pointer_to_heap_p rb_gc_functions.pointer_to_heap_p
# define rb_gc_impl_garbage_object_p rb_gc_functions.garbage_object_p
# define rb_gc_impl_set_event_hook rb_gc_functions.set_event_hook
# define rb_gc_impl_copy_attributes rb_gc_functions.copy_attributes
#endif
#ifdef RUBY_ASAN_ENABLED
static void
asan_death_callback(void)
{
if (GET_VM()) {
rb_bug_without_die("ASAN error");
}
}
#endif
2024-09-18 14:54:52 -04:00
static VALUE initial_stress = Qfalse;
void *
rb_objspace_alloc(void)
{
#if USE_MODULAR_GC
ruby_modular_gc_init();
#endif
void *objspace = rb_gc_impl_objspace_alloc();
ruby_current_vm_ptr->gc.objspace = objspace;
rb_gc_impl_objspace_init(objspace);
2024-09-18 14:54:52 -04:00
rb_gc_impl_stress_set(objspace, initial_stress);
#ifdef RUBY_ASAN_ENABLED
__sanitizer_set_death_callback(asan_death_callback);
#endif
return objspace;
}
void
rb_objspace_free(void *objspace)
{
rb_gc_impl_objspace_free(objspace);
}
size_t
rb_gc_obj_slot_size(VALUE obj)
{
return rb_gc_impl_obj_slot_size(obj);
}
static inline void
2025-04-19 22:02:10 +09:00
gc_validate_pc(void)
{
#if RUBY_DEBUG
rb_execution_context_t *ec = GET_EC();
const rb_control_frame_t *cfp = ec->cfp;
if (cfp && VM_FRAME_RUBYFRAME_P(cfp) && cfp->pc) {
RUBY_ASSERT(cfp->pc >= ISEQ_BODY(cfp->iseq)->iseq_encoded);
RUBY_ASSERT(cfp->pc <= ISEQ_BODY(cfp->iseq)->iseq_encoded + ISEQ_BODY(cfp->iseq)->iseq_size);
}
#endif
}
static inline VALUE
newobj_of(rb_ractor_t *cr, VALUE klass, VALUE flags, VALUE v1, VALUE v2, VALUE v3, bool wb_protected, size_t size)
{
VALUE obj = rb_gc_impl_new_obj(rb_gc_get_objspace(), cr->newobj_cache, klass, flags, v1, v2, v3, wb_protected, size);
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
gc_validate_pc();
if (UNLIKELY(rb_gc_event_hook_required_p(RUBY_INTERNAL_EVENT_NEWOBJ))) {
unsigned int lev;
RB_VM_LOCK_ENTER_CR_LEV(cr, &lev);
{
memset((char *)obj + RVALUE_SIZE, 0, rb_gc_obj_slot_size(obj) - RVALUE_SIZE);
/* We must disable GC here because the callback could call xmalloc
* which could potentially trigger a GC, and a lot of code is unsafe
* to trigger a GC right after an object has been allocated because
* they perform initialization for the object and assume that the
* GC does not trigger before then. */
bool gc_disabled = RTEST(rb_gc_disable_no_rest());
{
rb_gc_event_hook(obj, RUBY_INTERNAL_EVENT_NEWOBJ);
}
if (!gc_disabled) rb_gc_enable();
}
RB_VM_LOCK_LEAVE_CR_LEV(cr, &lev);
}
return obj;
}
VALUE
rb_wb_unprotected_newobj_of(VALUE klass, VALUE flags, size_t size)
{
GC_ASSERT((flags & FL_WB_PROTECTED) == 0);
return newobj_of(GET_RACTOR(), klass, flags, 0, 0, 0, FALSE, size);
}
VALUE
rb_wb_protected_newobj_of(rb_execution_context_t *ec, VALUE klass, VALUE flags, size_t size)
{
GC_ASSERT((flags & FL_WB_PROTECTED) == 0);
return newobj_of(rb_ec_ractor_ptr(ec), klass, flags, 0, 0, 0, TRUE, size);
}
#define UNEXPECTED_NODE(func) \
rb_bug(#func"(): GC does not handle T_NODE 0x%x(%p) 0x%"PRIxVALUE, \
BUILTIN_TYPE(obj), (void*)(obj), RBASIC(obj)->flags)
static inline void
rb_data_object_check(VALUE klass)
{
if (klass != rb_cObject && (rb_get_alloc_func(klass) == rb_class_allocate_instance)) {
rb_undef_alloc_func(klass);
rb_warn("undefining the allocator of T_DATA class %"PRIsVALUE, klass);
}
}
VALUE
rb_data_object_wrap(VALUE klass, void *datap, RUBY_DATA_FUNC dmark, RUBY_DATA_FUNC dfree)
{
RUBY_ASSERT_ALWAYS(dfree != (RUBY_DATA_FUNC)1);
if (klass) rb_data_object_check(klass);
Save one VALUE per embedded RTypedData This halves the amount of memory used for embedded RTypedData if they are one VALUE (8 bytes on 64-bit platforms) over the slot size limit. For Set, on 64-bit it uses an embedded 56-byte struct. With the previous implementation, the embedded structs starts at offset 32, resulting in a total size of 88. Since that is over the 80 byte limit, it goes to the next highest bucket, 160 bytes, wasting 72 bytes. This allows it to fit in a 80 byte bucket, which reduces the total size for small sets of from 224 bytes (160 bytes embedded, 64 bytes malloc, 72 bytes wasted in embedding) to 144 bytes (80 bytes embedded, 64 bytes malloc, 0 bytes wasted in embedding). Any other embedded RTypedData will see similar advantages if they are currently one VALUE over the limit. To implement this, remove the typed_flag from struct RTypedData. Embed the typed_flag information in the type member, which is now a tagged pointer using VALUE type, using the bottom low 2 bits as flags (1 bit for typed flag, the other for the embedded flag). To get the actual pointer, RTYPEDDATA_TYPE masks out the low 2 bits and then casts. That moves the RTypedData data pointer from offset 32 to offset 24 (on 64-bit). Vast amount of code in the internals (and probably external C extensions) expects the following code to work for both RData and non-embedded RTypedData: ```c DATA_PTR(obj) = some_pointer; ``` Allow this to work by moving the data pointer in RData between the dmark and dfree pointers, so it is at the same offset (24 on 64-bit). Other than these changes to the include files, the only changes needed were to gc.c, to account for the new struct layouts, handle setting the low bits in the type member, and to use RTYPEDDATA_TYPE(obj) instead of RTYPEDDATA(obj)->type.
2025-04-27 17:23:04 -07:00
return newobj_of(GET_RACTOR(), klass, T_DATA, (VALUE)dmark, (VALUE)datap, (VALUE)dfree, !dmark, sizeof(struct RTypedData));
}
VALUE
rb_data_object_zalloc(VALUE klass, size_t size, RUBY_DATA_FUNC dmark, RUBY_DATA_FUNC dfree)
{
VALUE obj = rb_data_object_wrap(klass, 0, dmark, dfree);
DATA_PTR(obj) = xcalloc(1, size);
return obj;
}
static VALUE
typed_data_alloc(VALUE klass, VALUE typed_flag, void *datap, const rb_data_type_t *type, size_t size)
{
RBIMPL_NONNULL_ARG(type);
if (klass) rb_data_object_check(klass);
bool wb_protected = (type->flags & RUBY_FL_WB_PROTECTED) || !type->function.dmark;
Save one VALUE per embedded RTypedData This halves the amount of memory used for embedded RTypedData if they are one VALUE (8 bytes on 64-bit platforms) over the slot size limit. For Set, on 64-bit it uses an embedded 56-byte struct. With the previous implementation, the embedded structs starts at offset 32, resulting in a total size of 88. Since that is over the 80 byte limit, it goes to the next highest bucket, 160 bytes, wasting 72 bytes. This allows it to fit in a 80 byte bucket, which reduces the total size for small sets of from 224 bytes (160 bytes embedded, 64 bytes malloc, 72 bytes wasted in embedding) to 144 bytes (80 bytes embedded, 64 bytes malloc, 0 bytes wasted in embedding). Any other embedded RTypedData will see similar advantages if they are currently one VALUE over the limit. To implement this, remove the typed_flag from struct RTypedData. Embed the typed_flag information in the type member, which is now a tagged pointer using VALUE type, using the bottom low 2 bits as flags (1 bit for typed flag, the other for the embedded flag). To get the actual pointer, RTYPEDDATA_TYPE masks out the low 2 bits and then casts. That moves the RTypedData data pointer from offset 32 to offset 24 (on 64-bit). Vast amount of code in the internals (and probably external C extensions) expects the following code to work for both RData and non-embedded RTypedData: ```c DATA_PTR(obj) = some_pointer; ``` Allow this to work by moving the data pointer in RData between the dmark and dfree pointers, so it is at the same offset (24 on 64-bit). Other than these changes to the include files, the only changes needed were to gc.c, to account for the new struct layouts, handle setting the low bits in the type member, and to use RTYPEDDATA_TYPE(obj) instead of RTYPEDDATA(obj)->type.
2025-04-27 17:23:04 -07:00
return newobj_of(GET_RACTOR(), klass, T_DATA, ((VALUE)type) | IS_TYPED_DATA | typed_flag, (VALUE)datap, 0, wb_protected, size);
}
VALUE
rb_data_typed_object_wrap(VALUE klass, void *datap, const rb_data_type_t *type)
{
if (UNLIKELY(type->flags & RUBY_TYPED_EMBEDDABLE)) {
rb_raise(rb_eTypeError, "Cannot wrap an embeddable TypedData");
}
return typed_data_alloc(klass, 0, datap, type, sizeof(struct RTypedData));
}
VALUE
rb_data_typed_object_zalloc(VALUE klass, size_t size, const rb_data_type_t *type)
{
if (type->flags & RUBY_TYPED_EMBEDDABLE) {
if (!(type->flags & RUBY_TYPED_FREE_IMMEDIATELY)) {
rb_raise(rb_eTypeError, "Embeddable TypedData must be freed immediately");
}
size_t embed_size = offsetof(struct RTypedData, data) + size;
if (rb_gc_size_allocatable_p(embed_size)) {
VALUE obj = typed_data_alloc(klass, TYPED_DATA_EMBEDDED, 0, type, embed_size);
memset((char *)obj + offsetof(struct RTypedData, data), 0, size);
return obj;
}
}
VALUE obj = typed_data_alloc(klass, 0, NULL, type, sizeof(struct RTypedData));
DATA_PTR(obj) = xcalloc(1, size);
return obj;
}
static size_t
rb_objspace_data_type_memsize(VALUE obj)
{
size_t size = 0;
if (RTYPEDDATA_P(obj)) {
const rb_data_type_t *type = RTYPEDDATA_TYPE(obj);
const void *ptr = RTYPEDDATA_GET_DATA(obj);
if (RTYPEDDATA_TYPE(obj)->flags & RUBY_TYPED_EMBEDDABLE && !RTYPEDDATA_EMBEDDED_P(obj)) {
#ifdef HAVE_MALLOC_USABLE_SIZE
size += malloc_usable_size((void *)ptr);
#endif
}
if (ptr && type->function.dsize) {
size += type->function.dsize(ptr);
}
}
return size;
}
const char *
rb_objspace_data_type_name(VALUE obj)
2022-02-03 11:51:20 -05:00
{
if (RTYPEDDATA_P(obj)) {
return RTYPEDDATA_TYPE(obj)->wrap_struct_name;
}
else {
return 0;
2022-02-03 11:51:20 -05:00
}
}
static enum rb_id_table_iterator_result
cvar_table_free_i(VALUE value, void *ctx)
{
xfree((void *)value);
return ID_TABLE_CONTINUE;
}
static void
io_fptr_finalize(void *fptr)
{
rb_io_fptr_finalize((struct rb_io *)fptr);
}
static inline void
make_io_zombie(void *objspace, VALUE obj)
{
rb_io_t *fptr = RFILE(obj)->fptr;
rb_gc_impl_make_zombie(objspace, obj, io_fptr_finalize, fptr);
}
static bool
rb_data_free(void *objspace, VALUE obj)
{
void *data = RTYPEDDATA_P(obj) ? RTYPEDDATA_GET_DATA(obj) : DATA_PTR(obj);
if (data) {
int free_immediately = false;
void (*dfree)(void *);
if (RTYPEDDATA_P(obj)) {
Save one VALUE per embedded RTypedData This halves the amount of memory used for embedded RTypedData if they are one VALUE (8 bytes on 64-bit platforms) over the slot size limit. For Set, on 64-bit it uses an embedded 56-byte struct. With the previous implementation, the embedded structs starts at offset 32, resulting in a total size of 88. Since that is over the 80 byte limit, it goes to the next highest bucket, 160 bytes, wasting 72 bytes. This allows it to fit in a 80 byte bucket, which reduces the total size for small sets of from 224 bytes (160 bytes embedded, 64 bytes malloc, 72 bytes wasted in embedding) to 144 bytes (80 bytes embedded, 64 bytes malloc, 0 bytes wasted in embedding). Any other embedded RTypedData will see similar advantages if they are currently one VALUE over the limit. To implement this, remove the typed_flag from struct RTypedData. Embed the typed_flag information in the type member, which is now a tagged pointer using VALUE type, using the bottom low 2 bits as flags (1 bit for typed flag, the other for the embedded flag). To get the actual pointer, RTYPEDDATA_TYPE masks out the low 2 bits and then casts. That moves the RTypedData data pointer from offset 32 to offset 24 (on 64-bit). Vast amount of code in the internals (and probably external C extensions) expects the following code to work for both RData and non-embedded RTypedData: ```c DATA_PTR(obj) = some_pointer; ``` Allow this to work by moving the data pointer in RData between the dmark and dfree pointers, so it is at the same offset (24 on 64-bit). Other than these changes to the include files, the only changes needed were to gc.c, to account for the new struct layouts, handle setting the low bits in the type member, and to use RTYPEDDATA_TYPE(obj) instead of RTYPEDDATA(obj)->type.
2025-04-27 17:23:04 -07:00
free_immediately = (RTYPEDDATA_TYPE(obj)->flags & RUBY_TYPED_FREE_IMMEDIATELY) != 0;
dfree = RTYPEDDATA_TYPE(obj)->function.dfree;
}
else {
dfree = RDATA(obj)->dfree;
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
if (dfree) {
if (dfree == RUBY_DEFAULT_FREE) {
if (!RTYPEDDATA_P(obj) || !RTYPEDDATA_EMBEDDED_P(obj)) {
xfree(data);
RB_DEBUG_COUNTER_INC(obj_data_xfree);
}
}
else if (free_immediately) {
(*dfree)(data);
if (RTYPEDDATA_TYPE(obj)->flags & RUBY_TYPED_EMBEDDABLE && !RTYPEDDATA_EMBEDDED_P(obj)) {
xfree(data);
}
RB_DEBUG_COUNTER_INC(obj_data_imm_free);
}
else {
rb_gc_impl_make_zombie(objspace, obj, dfree, data);
RB_DEBUG_COUNTER_INC(obj_data_zombie);
return FALSE;
}
}
else {
RB_DEBUG_COUNTER_INC(obj_data_empty);
}
}
return true;
}
2025-04-30 13:48:02 +09:00
struct classext_foreach_args {
VALUE klass;
rb_objspace_t *objspace; // used for update_*
};
static void
classext_free(rb_classext_t *ext, bool is_prime, VALUE namespace, void *arg)
{
struct rb_id_table *tbl;
struct classext_foreach_args *args = (struct classext_foreach_args *)arg;
rb_id_table_free(RCLASSEXT_M_TBL(ext));
rb_cc_tbl_free(RCLASSEXT_CC_TBL(ext), args->klass);
if (!RCLASSEXT_SHARED_CONST_TBL(ext) && (tbl = RCLASSEXT_CONST_TBL(ext)) != NULL) {
rb_free_const_table(tbl);
}
if ((tbl = RCLASSEXT_CVC_TBL(ext)) != NULL) {
rb_id_table_foreach_values(tbl, cvar_table_free_i, NULL);
rb_id_table_free(tbl);
}
rb_class_classext_free_subclasses(ext, args->klass);
if (RCLASSEXT_SUPERCLASSES_WITH_SELF(ext)) {
RUBY_ASSERT(is_prime); // superclasses should only be used on prime
2025-04-30 13:48:02 +09:00
xfree(RCLASSEXT_SUPERCLASSES(ext));
}
if (!is_prime) { // the prime classext will be freed with RClass
xfree(ext);
}
}
static void
classext_iclass_free(rb_classext_t *ext, bool is_prime, VALUE namespace, void *arg)
{
struct classext_foreach_args *args = (struct classext_foreach_args *)arg;
if (RCLASSEXT_ICLASS_IS_ORIGIN(ext) && !RCLASSEXT_ICLASS_ORIGIN_SHARED_MTBL(ext)) {
/* Method table is not shared for origin iclasses of classes */
rb_id_table_free(RCLASSEXT_M_TBL(ext));
}
if (RCLASSEXT_CALLABLE_M_TBL(ext) != NULL) {
rb_id_table_free(RCLASSEXT_CALLABLE_M_TBL(ext));
}
rb_cc_tbl_free(RCLASSEXT_CC_TBL(ext), args->klass);
rb_class_classext_free_subclasses(ext, args->klass);
if (!is_prime) { // the prime classext will be freed with RClass
xfree(ext);
}
}
bool
rb_gc_obj_free(void *objspace, VALUE obj)
{
2025-04-30 13:48:02 +09:00
struct classext_foreach_args args;
RB_DEBUG_COUNTER_INC(obj_free);
switch (BUILTIN_TYPE(obj)) {
case T_NIL:
case T_FIXNUM:
case T_TRUE:
case T_FALSE:
rb_bug("obj_free() called for broken object");
break;
default:
break;
}
2015-07-21 22:52:59 +00:00
switch (BUILTIN_TYPE(obj)) {
case T_OBJECT:
if (rb_shape_obj_too_complex_p(obj)) {
RB_DEBUG_COUNTER_INC(obj_obj_too_complex);
st_free_table(ROBJECT_FIELDS_HASH(obj));
}
else if (RBASIC(obj)->flags & ROBJECT_EMBED) {
RB_DEBUG_COUNTER_INC(obj_obj_embed);
}
else {
xfree(ROBJECT(obj)->as.heap.fields);
RB_DEBUG_COUNTER_INC(obj_obj_ptr);
}
break;
case T_MODULE:
case T_CLASS:
2025-04-30 13:48:02 +09:00
args.klass = obj;
rb_class_classext_foreach(obj, classext_free, (void *)&args);
if (RCLASS(obj)->ns_classext_tbl) {
st_free_table(RCLASS(obj)->ns_classext_tbl);
}
(void)RB_DEBUG_COUNTER_INC_IF(obj_module_ptr, BUILTIN_TYPE(obj) == T_MODULE);
(void)RB_DEBUG_COUNTER_INC_IF(obj_class_ptr, BUILTIN_TYPE(obj) == T_CLASS);
break;
case T_STRING:
rb_str_free(obj);
break;
case T_ARRAY:
rb_ary_free(obj);
break;
case T_HASH:
#if USE_DEBUG_COUNTER
switch (RHASH_SIZE(obj)) {
case 0:
RB_DEBUG_COUNTER_INC(obj_hash_empty);
break;
case 1:
RB_DEBUG_COUNTER_INC(obj_hash_1);
break;
case 2:
RB_DEBUG_COUNTER_INC(obj_hash_2);
break;
case 3:
RB_DEBUG_COUNTER_INC(obj_hash_3);
break;
case 4:
RB_DEBUG_COUNTER_INC(obj_hash_4);
break;
case 5:
case 6:
case 7:
case 8:
RB_DEBUG_COUNTER_INC(obj_hash_5_8);
break;
default:
GC_ASSERT(RHASH_SIZE(obj) > 8);
RB_DEBUG_COUNTER_INC(obj_hash_g8);
}
2020-12-17 17:03:05 +09:00
if (RHASH_AR_TABLE_P(obj)) {
if (RHASH_AR_TABLE(obj) == NULL) {
RB_DEBUG_COUNTER_INC(obj_hash_null);
}
else {
RB_DEBUG_COUNTER_INC(obj_hash_ar);
}
}
else {
RB_DEBUG_COUNTER_INC(obj_hash_st);
}
#endif
2020-12-17 17:03:05 +09:00
rb_hash_free(obj);
break;
case T_REGEXP:
if (RREGEXP(obj)->ptr) {
onig_free(RREGEXP(obj)->ptr);
RB_DEBUG_COUNTER_INC(obj_regexp_ptr);
}
break;
case T_DATA:
if (!rb_data_free(objspace, obj)) return false;
break;
case T_MATCH:
{
rb_matchext_t *rm = RMATCH_EXT(obj);
#if USE_DEBUG_COUNTER
if (rm->regs.num_regs >= 8) {
RB_DEBUG_COUNTER_INC(obj_match_ge8);
}
else if (rm->regs.num_regs >= 4) {
RB_DEBUG_COUNTER_INC(obj_match_ge4);
}
else if (rm->regs.num_regs >= 1) {
RB_DEBUG_COUNTER_INC(obj_match_under4);
}
#endif
onig_region_free(&rm->regs, 0);
xfree(rm->char_offset);
RB_DEBUG_COUNTER_INC(obj_match_ptr);
}
break;
case T_FILE:
if (RFILE(obj)->fptr) {
make_io_zombie(objspace, obj);
RB_DEBUG_COUNTER_INC(obj_file_ptr);
return FALSE;
}
break;
case T_RATIONAL:
RB_DEBUG_COUNTER_INC(obj_rational);
break;
case T_COMPLEX:
RB_DEBUG_COUNTER_INC(obj_complex);
break;
case T_MOVED:
break;
case T_ICLASS:
2025-04-30 13:48:02 +09:00
args.klass = obj;
rb_class_classext_foreach(obj, classext_iclass_free, (void *)&args);
if (RCLASS(obj)->ns_classext_tbl) {
st_free_table(RCLASS(obj)->ns_classext_tbl);
}
RB_DEBUG_COUNTER_INC(obj_iclass_ptr);
break;
case T_FLOAT:
RB_DEBUG_COUNTER_INC(obj_float);
break;
case T_BIGNUM:
if (!BIGNUM_EMBED_P(obj) && BIGNUM_DIGITS(obj)) {
xfree(BIGNUM_DIGITS(obj));
RB_DEBUG_COUNTER_INC(obj_bignum_ptr);
}
else {
RB_DEBUG_COUNTER_INC(obj_bignum_embed);
}
break;
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
case T_NODE:
UNEXPECTED_NODE(obj_free);
break;
Change the semantics of rb_postponed_job_register Our current implementation of rb_postponed_job_register suffers from some safety issues that can lead to interpreter crashes (see bug #1991). Essentially, the issue is that jobs can be called with the wrong arguments. We made two attempts to fix this whilst keeping the promised semantics, but: * The first one involved masking/unmasking when flushing jobs, which was believed to be too expensive * The second one involved a lock-free, multi-producer, single-consumer ringbuffer, which was too complex The critical insight behind this third solution is that essentially the only user of these APIs are a) internal, or b) profiling gems. For a), none of the usages actually require variable data; they will work just fine with the preregistration interface. For b), generally profiling gems only call a single callback with a single piece of data (which is actually usually just zero) for the life of the program. The ringbuffer is complex because it needs to support multi-word inserts of job & data (which can't be atomic); but nobody actually even needs that functionality, really. So, this comit: * Introduces a pre-registration API for jobs, with a GVL-requiring rb_postponed_job_prereigster, which returns a handle which can be used with an async-signal-safe rb_postponed_job_trigger. * Deprecates rb_postponed_job_register (and re-implements it on top of the preregister function for compatability) * Moves all the internal usages of postponed job register pre-registration
2023-11-19 22:54:57 +11:00
case T_STRUCT:
if ((RBASIC(obj)->flags & RSTRUCT_EMBED_LEN_MASK) ||
RSTRUCT(obj)->as.heap.ptr == NULL) {
RB_DEBUG_COUNTER_INC(obj_struct_embed);
}
else {
xfree((void *)RSTRUCT(obj)->as.heap.ptr);
RB_DEBUG_COUNTER_INC(obj_struct_ptr);
}
break;
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
case T_SYMBOL:
RB_DEBUG_COUNTER_INC(obj_symbol);
break;
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
case T_IMEMO:
rb_imemo_free((VALUE)obj);
break;
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
default:
rb_bug("gc_sweep(): unknown data type 0x%x(%p) 0x%"PRIxVALUE,
BUILTIN_TYPE(obj), (void*)obj, RBASIC(obj)->flags);
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
if (FL_TEST_RAW(obj, FL_FINALIZE)) {
rb_gc_impl_make_zombie(objspace, obj, 0, 0);
return FALSE;
}
else {
return TRUE;
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
}
void
rb_objspace_set_event_hook(const rb_event_flag_t event)
{
rb_gc_impl_set_event_hook(rb_gc_get_objspace(), event);
}
static int
internal_object_p(VALUE obj)
{
void *ptr = asan_unpoison_object_temporary(obj);
if (RBASIC(obj)->flags) {
switch (BUILTIN_TYPE(obj)) {
case T_NODE:
UNEXPECTED_NODE(internal_object_p);
break;
case T_NONE:
case T_MOVED:
case T_IMEMO:
case T_ICLASS:
case T_ZOMBIE:
break;
case T_CLASS:
2025-05-28 10:11:58 -07:00
if (obj == rb_mRubyVMFrozenCore)
return 1;
if (!RBASIC_CLASS(obj)) break;
if (RCLASS_SINGLETON_P(obj)) {
return rb_singleton_class_internal_p(obj);
}
return 0;
default:
if (!RBASIC(obj)->klass) break;
return 0;
}
}
if (ptr || !RBASIC(obj)->flags) {
2024-12-18 12:01:09 -05:00
rb_asan_poison_object(obj);
}
return 1;
}
int
rb_objspace_internal_object_p(VALUE obj)
{
return internal_object_p(obj);
}
struct os_each_struct {
size_t num;
VALUE of;
};
static int
os_obj_of_i(void *vstart, void *vend, size_t stride, void *data)
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
{
struct os_each_struct *oes = (struct os_each_struct *)data;
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
VALUE v = (VALUE)vstart;
for (; v != (VALUE)vend; v += stride) {
if (!internal_object_p(v)) {
if (!oes->of || rb_obj_is_kind_of(v, oes->of)) {
if (!rb_multi_ractor_p() || rb_ractor_shareable_p(v)) {
rb_yield(v);
oes->num++;
}
}
}
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
return 0;
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
}
static VALUE
os_obj_of(VALUE of)
{
struct os_each_struct oes;
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
oes.num = 0;
oes.of = of;
rb_objspace_each_objects(os_obj_of_i, &oes);
return SIZET2NUM(oes.num);
}
/*
* call-seq:
* ObjectSpace.each_object([module]) {|obj| ... } -> integer
* ObjectSpace.each_object([module]) -> an_enumerator
*
* Calls the block once for each living, nonimmediate object in this
* Ruby process. If <i>module</i> is specified, calls the block
* for only those classes or modules that match (or are a subclass of)
* <i>module</i>. Returns the number of objects found. Immediate
* objects (<code>Fixnum</code>s, <code>Symbol</code>s
* <code>true</code>, <code>false</code>, and <code>nil</code>) are
* never returned. In the example below, #each_object returns both
* the numbers we defined and several constants defined in the Math
* module.
*
* If no block is given, an enumerator is returned instead.
*
* a = 102.7
* b = 95 # Won't be returned
* c = 12345678987654321
* count = ObjectSpace.each_object(Numeric) {|x| p x }
* puts "Total count: #{count}"
*
* <em>produces:</em>
*
* 12345678987654321
* 102.7
* 2.71828182845905
* 3.14159265358979
* 2.22044604925031e-16
* 1.7976931348623157e+308
* 2.2250738585072e-308
* Total count: 7
*
* Due to a current known Ractor implementation issue, this method will not yield
* Ractor-unshareable objects in multi-Ractor mode (when
* <code>Ractor.new</code> has been called within the process at least once).
* See https://bugs.ruby-lang.org/issues/19387 for more information.
*
* a = 12345678987654321 # shareable
* b = [].freeze # shareable
* c = {} # not shareable
* ObjectSpace.each_object {|x| x } # yields a, b, and c
* Ractor.new {} # enter multi-Ractor mode
* ObjectSpace.each_object {|x| x } # does not yield c
*
*/
static VALUE
os_each_obj(int argc, VALUE *argv, VALUE os)
{
VALUE of;
of = (!rb_check_arity(argc, 0, 1) ? 0 : argv[0]);
RETURN_ENUMERATOR(os, 1, &of);
return os_obj_of(of);
}
/*
* call-seq:
* ObjectSpace.undefine_finalizer(obj)
*
* Removes all finalizers for <i>obj</i>.
*
*/
static VALUE
undefine_final(VALUE os, VALUE obj)
{
return rb_undefine_finalizer(obj);
}
VALUE
rb_undefine_finalizer(VALUE obj)
{
rb_check_frozen(obj);
rb_gc_impl_undefine_finalizer(rb_gc_get_objspace(), obj);
return obj;
}
static void
should_be_callable(VALUE block)
{
if (!rb_obj_respond_to(block, idCall, TRUE)) {
rb_raise(rb_eArgError, "wrong type argument %"PRIsVALUE" (should be callable)",
rb_obj_class(block));
}
}
* gc.c: support RGENGC. [ruby-trunk - Feature #8339] See this ticet about RGENGC. * gc.c: Add several flags: * RGENGC_DEBUG: if >0, then prints debug information. * RGENGC_CHECK_MODE: if >0, add assertions. * RGENGC_PROFILE: if >0, add profiling features. check GC.stat and GC::Profiler. * include/ruby/ruby.h: disable RGENGC by default (USE_RGENGC == 0). * array.c: add write barriers for T_ARRAY and generate sunny objects. * include/ruby/ruby.h (RARRAY_PTR_USE): added. Use this macro if you want to access raw pointers. If you modify the contents which pointer pointed, then you need to care write barrier. * bignum.c, marshal.c, random.c: generate T_BIGNUM sunny objects. * complex.c, include/ruby/ruby.h: add write barriers for T_COMPLEX and generate sunny objects. * rational.c (nurat_s_new_internal), include/ruby/ruby.h: add write barriers for T_RATIONAL and generate sunny objects. * internal.h: add write barriers for RBasic::klass. * numeric.c (rb_float_new_in_heap): generate sunny T_FLOAT objects. * object.c (rb_class_allocate_instance), range.c: generate sunny T_OBJECT objects. * string.c: add write barriers for T_STRING and generate sunny objects. * variable.c: add write barriers for ivars. * vm_insnhelper.c (vm_setivar): ditto. * include/ruby/ruby.h, debug.c: use two flags FL_WB_PROTECTED and FL_OLDGEN. * node.h (NODE_FL_CREF_PUSHED_BY_EVAL, NODE_FL_CREF_OMOD_SHARED): move flag bits. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40703 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2013-05-13 18:07:47 +00:00
static void
should_be_finalizable(VALUE obj)
{
if (!FL_ABLE(obj)) {
rb_raise(rb_eArgError, "cannot define finalizer for %s",
rb_obj_classname(obj));
}
rb_check_frozen(obj);
}
void
rb_gc_copy_finalizer(VALUE dest, VALUE obj)
{
rb_gc_impl_copy_finalizer(rb_gc_get_objspace(), dest, obj);
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
/*
* call-seq:
* ObjectSpace.define_finalizer(obj, aProc=proc())
*
* Adds <i>aProc</i> as a finalizer, to be called after <i>obj</i>
* was destroyed. The object ID of the <i>obj</i> will be passed
* as an argument to <i>aProc</i>. If <i>aProc</i> is a lambda or
* method, make sure it can be called with a single argument.
*
* The return value is an array <code>[0, aProc]</code>.
*
* The two recommended patterns are to either create the finaliser proc
* in a non-instance method where it can safely capture the needed state,
* or to use a custom callable object that stores the needed state
* explicitly as instance variables.
*
* class Foo
* def initialize(data_needed_for_finalization)
* ObjectSpace.define_finalizer(self, self.class.create_finalizer(data_needed_for_finalization))
* end
*
* def self.create_finalizer(data_needed_for_finalization)
* proc {
* puts "finalizing #{data_needed_for_finalization}"
* }
* end
* end
*
* class Bar
* class Remover
* def initialize(data_needed_for_finalization)
* @data_needed_for_finalization = data_needed_for_finalization
* end
*
* def call(id)
* puts "finalizing #{@data_needed_for_finalization}"
* end
* end
*
* def initialize(data_needed_for_finalization)
* ObjectSpace.define_finalizer(self, Remover.new(data_needed_for_finalization))
* end
* end
*
* Note that if your finalizer references the object to be
* finalized it will never be run on GC, although it will still be
* run at exit. You will get a warning if you capture the object
* to be finalized as the receiver of the finalizer.
*
* class CapturesSelf
* def initialize(name)
* ObjectSpace.define_finalizer(self, proc {
* # this finalizer will only be run on exit
* puts "finalizing #{name}"
* })
* end
* end
*
* Also note that finalization can be unpredictable and is never guaranteed
* to be run except on exit.
*/
* gc.c: support RGENGC. [ruby-trunk - Feature #8339] See this ticet about RGENGC. * gc.c: Add several flags: * RGENGC_DEBUG: if >0, then prints debug information. * RGENGC_CHECK_MODE: if >0, add assertions. * RGENGC_PROFILE: if >0, add profiling features. check GC.stat and GC::Profiler. * include/ruby/ruby.h: disable RGENGC by default (USE_RGENGC == 0). * array.c: add write barriers for T_ARRAY and generate sunny objects. * include/ruby/ruby.h (RARRAY_PTR_USE): added. Use this macro if you want to access raw pointers. If you modify the contents which pointer pointed, then you need to care write barrier. * bignum.c, marshal.c, random.c: generate T_BIGNUM sunny objects. * complex.c, include/ruby/ruby.h: add write barriers for T_COMPLEX and generate sunny objects. * rational.c (nurat_s_new_internal), include/ruby/ruby.h: add write barriers for T_RATIONAL and generate sunny objects. * internal.h: add write barriers for RBasic::klass. * numeric.c (rb_float_new_in_heap): generate sunny T_FLOAT objects. * object.c (rb_class_allocate_instance), range.c: generate sunny T_OBJECT objects. * string.c: add write barriers for T_STRING and generate sunny objects. * variable.c: add write barriers for ivars. * vm_insnhelper.c (vm_setivar): ditto. * include/ruby/ruby.h, debug.c: use two flags FL_WB_PROTECTED and FL_OLDGEN. * node.h (NODE_FL_CREF_PUSHED_BY_EVAL, NODE_FL_CREF_OMOD_SHARED): move flag bits. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40703 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2013-05-13 18:07:47 +00:00
static VALUE
define_final(int argc, VALUE *argv, VALUE os)
{
VALUE obj, block;
rb_scan_args(argc, argv, "11", &obj, &block);
if (argc == 1) {
block = rb_block_proc();
}
if (rb_callable_receiver(block) == obj) {
rb_warn("finalizer references object to be finalized");
}
return rb_define_finalizer(obj, block);
}
VALUE
rb_define_finalizer(VALUE obj, VALUE block)
{
should_be_finalizable(obj);
should_be_callable(block);
block = rb_gc_impl_define_finalizer(rb_gc_get_objspace(), obj, block);
block = rb_ary_new3(2, INT2FIX(0), block);
OBJ_FREEZE(block);
return block;
}
void
rb_objspace_call_finalizer(void)
{
rb_gc_impl_shutdown_call_finalizer(rb_gc_get_objspace());
}
void
rb_objspace_free_objects(void *objspace)
{
rb_gc_impl_shutdown_free_objects(objspace);
}
int
rb_objspace_garbage_object_p(VALUE obj)
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
{
return rb_gc_impl_garbage_object_p(rb_gc_get_objspace(), obj);
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
}
bool
rb_gc_pointer_to_heap_p(VALUE obj)
{
return rb_gc_impl_pointer_to_heap_p(rb_gc_get_objspace(), (void *)obj);
}
#define OBJ_ID_INCREMENT (RUBY_IMMEDIATE_MASK + 1)
#define LAST_OBJECT_ID() (object_id_counter * OBJ_ID_INCREMENT)
static VALUE id2ref_value = 0;
static st_table *id2ref_tbl = NULL;
static bool id2ref_tbl_built = false;
#if SIZEOF_SIZE_T == SIZEOF_LONG_LONG
static size_t object_id_counter = 1;
#else
static unsigned long long object_id_counter = 1;
#endif
static inline VALUE
generate_next_object_id(void)
{
#if SIZEOF_SIZE_T == SIZEOF_LONG_LONG
// 64bit atomics are available
return SIZET2NUM(RUBY_ATOMIC_SIZE_FETCH_ADD(object_id_counter, 1) * OBJ_ID_INCREMENT);
#else
unsigned int lock_lev = RB_GC_VM_LOCK();
VALUE id = ULL2NUM(++object_id_counter * OBJ_ID_INCREMENT);
RB_GC_VM_UNLOCK(lock_lev);
return id;
#endif
}
void
rb_gc_obj_id_moved(VALUE obj)
{
if (UNLIKELY(id2ref_tbl)) {
st_insert(id2ref_tbl, (st_data_t)rb_obj_id(obj), (st_data_t)obj);
}
}
static int
object_id_cmp(st_data_t x, st_data_t y)
{
if (RB_TYPE_P(x, T_BIGNUM)) {
return !rb_big_eql(x, y);
}
else {
return x != y;
}
}
static st_index_t
object_id_hash(st_data_t n)
{
return FIX2LONG(rb_hash((VALUE)n));
}
static const struct st_hash_type object_id_hash_type = {
object_id_cmp,
object_id_hash,
};
static void gc_mark_tbl_no_pin(st_table *table);
static void
id2ref_tbl_mark(void *data)
{
st_table *table = (st_table *)data;
if (UNLIKELY(!RB_POSFIXABLE(LAST_OBJECT_ID()))) {
// It's very unlikely, but if enough object ids were generated, keys may be T_BIGNUM
rb_mark_set(table);
}
// We purposedly don't mark values, as they are weak references.
// rb_gc_obj_free_vm_weak_references takes care of cleaning them up.
}
static size_t
id2ref_tbl_memsize(const void *data)
{
return rb_st_memsize(data);
}
static void
id2ref_tbl_free(void *data)
{
id2ref_tbl = NULL; // clear global ref
st_table *table = (st_table *)data;
st_free_table(table);
}
static const rb_data_type_t id2ref_tbl_type = {
.wrap_struct_name = "VM/_id2ref_table",
.function = {
.dmark = id2ref_tbl_mark,
.dfree = id2ref_tbl_free,
.dsize = id2ref_tbl_memsize,
// dcompact function not required because the table is reference updated
// in rb_gc_vm_weak_table_foreach
},
.flags = RUBY_TYPED_WB_PROTECTED | RUBY_TYPED_FREE_IMMEDIATELY
};
#define RUBY_ATOMIC_VALUE_LOAD(x) (VALUE)(RUBY_ATOMIC_PTR_LOAD(x))
static VALUE
class_object_id(VALUE klass)
{
VALUE id = RUBY_ATOMIC_VALUE_LOAD(RCLASS(klass)->object_id);
if (!id) {
unsigned int lock_lev = RB_GC_VM_LOCK();
id = generate_next_object_id();
VALUE existing_id = RUBY_ATOMIC_VALUE_CAS(RCLASS(klass)->object_id, 0, id);
if (existing_id) {
id = existing_id;
}
else if (RB_UNLIKELY(id2ref_tbl)) {
st_insert(id2ref_tbl, id, klass);
}
RB_GC_VM_UNLOCK(lock_lev);
}
return id;
}
static inline VALUE
object_id_get(VALUE obj, shape_id_t shape_id)
{
VALUE id;
if (rb_shape_too_complex_p(shape_id)) {
id = rb_obj_field_get(obj, ROOT_TOO_COMPLEX_WITH_OBJ_ID);
}
else {
id = rb_obj_field_get(obj, rb_shape_object_id(shape_id));
}
#if RUBY_DEBUG
if (!(FIXNUM_P(id) || RB_TYPE_P(id, T_BIGNUM))) {
rb_p(obj);
rb_bug("Object's shape includes object_id, but it's missing %s", rb_obj_info(obj));
}
#endif
return id;
}
static VALUE
object_id0(VALUE obj)
{
VALUE id = Qfalse;
shape_id_t shape_id = RBASIC_SHAPE_ID(obj);
if (rb_shape_has_object_id(shape_id)) {
return object_id_get(obj, shape_id);
}
// rb_shape_object_id_shape may lock if the current shape has
// multiple children.
shape_id_t object_id_shape_id = rb_shape_transition_object_id(obj);
id = generate_next_object_id();
rb_obj_field_set(obj, object_id_shape_id, id);
RUBY_ASSERT(RBASIC_SHAPE_ID(obj) == object_id_shape_id);
RUBY_ASSERT(rb_shape_obj_has_id(obj));
if (RB_UNLIKELY(id2ref_tbl)) {
st_insert(id2ref_tbl, (st_data_t)id, (st_data_t)obj);
}
return id;
}
static VALUE
object_id(VALUE obj)
{
switch (BUILTIN_TYPE(obj)) {
case T_CLASS:
case T_MODULE:
// With namespaces, classes and modules have different fields
// in different namespaces, so we cannot store the object id
// in fields.
return class_object_id(obj);
case T_IMEMO:
rb_bug("T_IMEMO can't have an object_id");
break;
default:
break;
}
if (UNLIKELY(rb_gc_multi_ractor_p() && rb_ractor_shareable_p(obj))) {
unsigned int lock_lev = RB_GC_VM_LOCK();
VALUE id = object_id0(obj);
RB_GC_VM_UNLOCK(lock_lev);
return id;
}
return object_id0(obj);
}
static void
build_id2ref_i(VALUE obj, void *data)
{
st_table *id2ref_tbl = (st_table *)data;
switch (BUILTIN_TYPE(obj)) {
case T_CLASS:
case T_MODULE:
if (RCLASS(obj)->object_id) {
st_insert(id2ref_tbl, RCLASS(obj)->object_id, obj);
}
break;
case T_IMEMO:
case T_NONE:
break;
default:
if (rb_shape_obj_has_id(obj)) {
st_insert(id2ref_tbl, rb_obj_id(obj), obj);
}
break;
}
}
static VALUE
object_id_to_ref(void *objspace_ptr, VALUE object_id)
{
rb_objspace_t *objspace = objspace_ptr;
unsigned int lev = RB_GC_VM_LOCK();
if (!id2ref_tbl) {
rb_gc_vm_barrier(); // stop other ractors
// GC Must not trigger while we build the table, otherwise if we end
// up freeing an object that had an ID, we might try to delete it from
// the table even though it wasn't inserted yet.
id2ref_tbl = st_init_table(&object_id_hash_type);
id2ref_value = TypedData_Wrap_Struct(0, &id2ref_tbl_type, id2ref_tbl);
// build_id2ref_i will most certainly malloc, which could trigger GC and sweep
// objects we just added to the table.
bool gc_disabled = RTEST(rb_gc_disable_no_rest());
{
rb_gc_impl_each_object(objspace, build_id2ref_i, (void *)id2ref_tbl);
}
if (!gc_disabled) rb_gc_enable();
id2ref_tbl_built = true;
}
VALUE obj;
bool found = st_lookup(id2ref_tbl, object_id, &obj) && !rb_gc_impl_garbage_object_p(objspace, obj);
RB_GC_VM_UNLOCK(lev);
if (found) {
return obj;
}
if (rb_funcall(object_id, rb_intern(">="), 1, ULL2NUM(LAST_OBJECT_ID()))) {
rb_raise(rb_eRangeError, "%+"PRIsVALUE" is not an id value", rb_funcall(object_id, rb_intern("to_s"), 1, INT2FIX(10)));
}
else {
rb_raise(rb_eRangeError, "%+"PRIsVALUE" is a recycled object", rb_funcall(object_id, rb_intern("to_s"), 1, INT2FIX(10)));
}
}
static inline void
obj_free_object_id(VALUE obj)
{
if (RB_BUILTIN_TYPE(obj) == T_IMEMO) {
return;
}
#if RUBY_DEBUG
switch (BUILTIN_TYPE(obj)) {
case T_CLASS:
case T_MODULE:
break;
default:
if (rb_shape_obj_has_id(obj)) {
VALUE id = object_id_get(obj, RBASIC_SHAPE_ID(obj)); // Crash if missing
if (!(FIXNUM_P(id) || RB_TYPE_P(id, T_BIGNUM))) {
rb_p(obj);
rb_bug("Corrupted object_id");
}
}
break;
}
#endif
VALUE obj_id = 0;
if (RB_UNLIKELY(id2ref_tbl)) {
switch (BUILTIN_TYPE(obj)) {
case T_CLASS:
case T_MODULE:
obj_id = RCLASS(obj)->object_id;
break;
default: {
shape_id_t shape_id = RBASIC_SHAPE_ID(obj);
if (rb_shape_has_object_id(shape_id)) {
obj_id = object_id_get(obj, shape_id);
}
break;
}
}
if (RB_UNLIKELY(obj_id)) {
RUBY_ASSERT(FIXNUM_P(obj_id) || RB_TYPE_P(obj_id, T_BIGNUM));
if (!st_delete(id2ref_tbl, (st_data_t *)&obj_id, NULL)) {
// If we're currently building the table then it's not a bug
if (id2ref_tbl_built) {
rb_bug("Object ID seen, but not in _id2ref table: object_id=%llu object=%s", NUM2ULL(obj_id), rb_obj_info(obj));
}
}
}
}
}
void
rb_gc_obj_free_vm_weak_references(VALUE obj)
{
obj_free_object_id(obj);
if (FL_TEST_RAW(obj, FL_EXIVAR)) {
rb_free_generic_ivar((VALUE)obj);
FL_UNSET_RAW(obj, FL_EXIVAR);
}
switch (BUILTIN_TYPE(obj)) {
case T_STRING:
if (FL_TEST_RAW(obj, RSTRING_FSTR)) {
rb_gc_free_fstring(obj);
}
break;
case T_SYMBOL:
rb_gc_free_dsymbol(obj);
break;
case T_IMEMO:
switch (imemo_type(obj)) {
case imemo_callcache: {
const struct rb_callcache *cc = (const struct rb_callcache *)obj;
if (vm_cc_refinement_p(cc)) {
rb_vm_delete_cc_refinement(cc);
}
break;
}
case imemo_callinfo:
rb_vm_ci_free((const struct rb_callinfo *)obj);
break;
case imemo_ment:
rb_free_method_entry_vm_weak_references((const rb_method_entry_t *)obj);
break;
default:
break;
}
break;
default:
break;
}
}
/*
* call-seq:
* ObjectSpace._id2ref(object_id) -> an_object
*
* Converts an object id to a reference to the object. May not be
* called on an object id passed as a parameter to a finalizer.
*
* s = "I am a string" #=> "I am a string"
* r = ObjectSpace._id2ref(s.object_id) #=> "I am a string"
* r == s #=> true
*
* On multi-ractor mode, if the object is not shareable, it raises
* RangeError.
*
* This method is deprecated and should no longer be used.
*/
static VALUE
id2ref(VALUE objid)
{
#if SIZEOF_LONG == SIZEOF_VOIDP
#define NUM2PTR(x) NUM2ULONG(x)
#elif SIZEOF_LONG_LONG == SIZEOF_VOIDP
#define NUM2PTR(x) NUM2ULL(x)
#endif
objid = rb_to_int(objid);
if (FIXNUM_P(objid) || rb_big_size(objid) <= SIZEOF_VOIDP) {
VALUE ptr = NUM2PTR(objid);
if (SPECIAL_CONST_P(ptr)) {
if (ptr == Qtrue) return Qtrue;
if (ptr == Qfalse) return Qfalse;
if (NIL_P(ptr)) return Qnil;
if (FIXNUM_P(ptr)) return ptr;
if (FLONUM_P(ptr)) return ptr;
if (SYMBOL_P(ptr)) {
// Check that the symbol is valid
if (rb_static_id_valid_p(SYM2ID(ptr))) {
return ptr;
}
else {
rb_raise(rb_eRangeError, "%p is not a symbol id value", (void *)ptr);
}
}
rb_raise(rb_eRangeError, "%+"PRIsVALUE" is not an id value", rb_int2str(objid, 10));
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
}
VALUE obj = object_id_to_ref(rb_gc_get_objspace(), objid);
if (!rb_multi_ractor_p() || rb_ractor_shareable_p(obj)) {
return obj;
}
else {
rb_raise(rb_eRangeError, "%+"PRIsVALUE" is the id of an unshareable object on multi-ractor", rb_int2str(objid, 10));
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
}
/* :nodoc: */
static VALUE
os_id2ref(VALUE os, VALUE objid)
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
{
rb_category_warn(RB_WARN_CATEGORY_DEPRECATED, "ObjectSpace._id2ref is deprecated");
return id2ref(objid);
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
}
static VALUE
rb_find_object_id(void *objspace, VALUE obj, VALUE (*get_heap_object_id)(VALUE))
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
{
if (SPECIAL_CONST_P(obj)) {
#if SIZEOF_LONG == SIZEOF_VOIDP
return LONG2NUM((SIGNED_VALUE)obj);
#else
return LL2NUM((SIGNED_VALUE)obj);
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
#endif
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
return get_heap_object_id(obj);
* gc.c: add 3gen GC patch, but disabled as default. RGenGC is designed as 2 generational GC, young and old generation. Young objects will be promoted to old objects after one GC. Old objects are not collect until major (full) GC. The issue of this approach is some objects can promoted as old objects accidentally and not freed until major GC. Major GC is not frequently so short-lived but accidentally becoming old objects are not freed. For example, the program "loop{Array.new(1_000_000)}" consumes huge memories because short lived objects (an array which has 1M elements) are promoted while GC and they are not freed before major GC. To solve this problem, generational GC with more generations technique is known. This patch implements three generations gen GC. At first, newly created objects are "Infant" objects. After surviving one GC, "Infant" objects are promoted to "Young" objects. "Young" objects are promoted to "Old" objects after surviving next GC. "Infant" and "Young" objects are collected if it is not marked while minor GC. So that this technique solves this problem. Representation of generations: * Infant: !FL_PROMOTED and !oldgen_bitmap [00] * Young : FL_PROMOTED and !oldgen_bitmap [10] * Old : FL_PROMOTED and oldgen_bitmap [11] The macro "RGENGC_THREEGEN" enables/disables this feature, and turned off as default because there are several problems. (1) Failed sometimes (Heisenbugs). (2) Performance down. Especially on write barriers. We need to detect Young or Old object by oldgen_bitmap. It is slower than checking flags. To evaluate this feature on more applications, I commit this patch. Reports are very welcome. This patch includes some refactoring (renaming names, etc). * include/ruby/ruby.h: catch up 3gen GC. * .gdbinit: fix to show a prompt "[PROMOTED]" for promoted objects. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@43530 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2013-11-04 18:59:33 +00:00
}
static VALUE
nonspecial_obj_id(VALUE obj)
{
#if SIZEOF_LONG == SIZEOF_VOIDP
return (VALUE)((SIGNED_VALUE)(obj)|FIXNUM_FLAG);
#elif SIZEOF_LONG_LONG == SIZEOF_VOIDP
return LL2NUM((SIGNED_VALUE)(obj) / 2);
#else
# error not supported
#endif
}
VALUE
rb_memory_id(VALUE obj)
{
return rb_find_object_id(NULL, obj, nonspecial_obj_id);
}
/*
* Document-method: __id__
* Document-method: object_id
*
* call-seq:
* obj.__id__ -> integer
* obj.object_id -> integer
*
* Returns an integer identifier for +obj+.
*
* The same number will be returned on all calls to +object_id+ for a given
* object, and no two active objects will share an id.
*
* Note: that some objects of builtin classes are reused for optimization.
* This is the case for immediate values and frozen string literals.
*
* BasicObject implements +__id__+, Kernel implements +object_id+.
*
* Immediate values are not passed by reference but are passed by value:
* +nil+, +true+, +false+, Fixnums, Symbols, and some Floats.
*
* Object.new.object_id == Object.new.object_id # => false
* (21 * 2).object_id == (21 * 2).object_id # => true
* "hello".object_id == "hello".object_id # => false
* "hi".freeze.object_id == "hi".freeze.object_id # => true
*/
VALUE
rb_obj_id(VALUE obj)
{
/* If obj is an immediate, the object ID is obj directly converted to a Numeric.
* Otherwise, the object ID is a Numeric that is a non-zero multiple of
* (RUBY_IMMEDIATE_MASK + 1) which guarantees that it does not collide with
* any immediates. */
return rb_find_object_id(rb_gc_get_objspace(), obj, object_id);
}
bool
rb_obj_id_p(VALUE obj)
{
return !RB_TYPE_P(obj, T_IMEMO) && rb_shape_obj_has_id(obj);
}
static enum rb_id_table_iterator_result
cc_table_memsize_i(VALUE ccs_ptr, void *data_ptr)
{
size_t *total_size = data_ptr;
struct rb_class_cc_entries *ccs = (struct rb_class_cc_entries *)ccs_ptr;
*total_size += sizeof(*ccs);
*total_size += sizeof(ccs->entries[0]) * ccs->capa;
return ID_TABLE_CONTINUE;
}
static size_t
cc_table_memsize(struct rb_id_table *cc_table)
{
size_t total = rb_id_table_memsize(cc_table);
rb_id_table_foreach_values(cc_table, cc_table_memsize_i, &total);
return total;
}
2025-04-30 13:48:02 +09:00
static void
classext_memsize(rb_classext_t *ext, bool prime, VALUE namespace, void *arg)
{
size_t *size = (size_t *)arg;
size_t s = 0;
if (RCLASSEXT_M_TBL(ext)) {
s += rb_id_table_memsize(RCLASSEXT_M_TBL(ext));
}
if (RCLASSEXT_CVC_TBL(ext)) {
s += rb_id_table_memsize(RCLASSEXT_CVC_TBL(ext));
}
if (RCLASSEXT_CONST_TBL(ext)) {
s += rb_id_table_memsize(RCLASSEXT_CONST_TBL(ext));
}
if (RCLASSEXT_CC_TBL(ext)) {
s += cc_table_memsize(RCLASSEXT_CC_TBL(ext));
}
if (RCLASSEXT_SUPERCLASSES_WITH_SELF(ext)) {
s += (RCLASSEXT_SUPERCLASS_DEPTH(ext) + 1) * sizeof(VALUE);
}
if (!prime) {
s += sizeof(rb_classext_t);
}
*size += s;
}
static void
classext_superclasses_memsize(rb_classext_t *ext, bool prime, VALUE namespace, void *arg)
{
size_t *size = (size_t *)arg;
size_t array_size;
if (RCLASSEXT_SUPERCLASSES_WITH_SELF(ext)) {
RUBY_ASSERT(prime);
array_size = RCLASSEXT_SUPERCLASS_DEPTH(ext) + 1;
2025-04-30 13:48:02 +09:00
*size += array_size * sizeof(VALUE);
}
}
size_t
rb_obj_memsize_of(VALUE obj)
{
size_t size = 0;
if (SPECIAL_CONST_P(obj)) {
return 0;
}
if (FL_TEST(obj, FL_EXIVAR)) {
size += rb_generic_ivar_memsize(obj);
}
switch (BUILTIN_TYPE(obj)) {
case T_OBJECT:
if (rb_shape_obj_too_complex_p(obj)) {
size += rb_st_memsize(ROBJECT_FIELDS_HASH(obj));
}
else if (!(RBASIC(obj)->flags & ROBJECT_EMBED)) {
size += ROBJECT_FIELDS_CAPACITY(obj) * sizeof(VALUE);
}
break;
case T_MODULE:
case T_CLASS:
2025-04-30 13:48:02 +09:00
rb_class_classext_foreach(obj, classext_memsize, (void *)&size);
rb_class_classext_foreach(obj, classext_superclasses_memsize, (void *)&size);
break;
case T_ICLASS:
if (RICLASS_OWNS_M_TBL_P(obj)) {
if (RCLASS_M_TBL(obj)) {
size += rb_id_table_memsize(RCLASS_M_TBL(obj));
}
}
2025-04-30 13:48:02 +09:00
if (RCLASS_WRITABLE_CC_TBL(obj)) {
size += cc_table_memsize(RCLASS_WRITABLE_CC_TBL(obj));
}
break;
case T_STRING:
size += rb_str_memsize(obj);
break;
case T_ARRAY:
size += rb_ary_memsize(obj);
break;
case T_HASH:
if (RHASH_ST_TABLE_P(obj)) {
VM_ASSERT(RHASH_ST_TABLE(obj) != NULL);
/* st_table is in the slot */
size += st_memsize(RHASH_ST_TABLE(obj)) - sizeof(st_table);
}
break;
case T_REGEXP:
if (RREGEXP_PTR(obj)) {
size += onig_memsize(RREGEXP_PTR(obj));
}
break;
case T_DATA:
size += rb_objspace_data_type_memsize(obj);
break;
case T_MATCH:
{
rb_matchext_t *rm = RMATCH_EXT(obj);
size += onig_region_memsize(&rm->regs);
size += sizeof(struct rmatch_offset) * rm->char_offset_num_allocated;
}
break;
case T_FILE:
if (RFILE(obj)->fptr) {
size += rb_io_memsize(RFILE(obj)->fptr);
}
break;
case T_RATIONAL:
case T_COMPLEX:
break;
case T_IMEMO:
size += rb_imemo_memsize(obj);
break;
case T_FLOAT:
case T_SYMBOL:
break;
case T_BIGNUM:
if (!(RBASIC(obj)->flags & BIGNUM_EMBED_FLAG) && BIGNUM_DIGITS(obj)) {
size += BIGNUM_LEN(obj) * sizeof(BDIGIT);
}
break;
case T_NODE:
UNEXPECTED_NODE(obj_memsize_of);
break;
case T_STRUCT:
if ((RBASIC(obj)->flags & RSTRUCT_EMBED_LEN_MASK) == 0 &&
RSTRUCT(obj)->as.heap.ptr) {
size += sizeof(VALUE) * RSTRUCT_LEN(obj);
}
break;
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
case T_ZOMBIE:
case T_MOVED:
break;
default:
rb_bug("objspace/memsize_of(): unknown data type 0x%x(%p)",
BUILTIN_TYPE(obj), (void*)obj);
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
}
return size + rb_gc_obj_slot_size(obj);
}
static int
set_zero(st_data_t key, st_data_t val, st_data_t arg)
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
{
VALUE k = (VALUE)key;
VALUE hash = (VALUE)arg;
rb_hash_aset(hash, k, INT2FIX(0));
return ST_CONTINUE;
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
}
struct count_objects_data {
size_t counts[T_MASK+1];
size_t freed;
size_t total;
};
static void
count_objects_i(VALUE obj, void *d)
{
struct count_objects_data *data = (struct count_objects_data *)d;
if (RBASIC(obj)->flags) {
data->counts[BUILTIN_TYPE(obj)]++;
}
else {
data->freed++;
}
data->total++;
}
/*
* call-seq:
* ObjectSpace.count_objects([result_hash]) -> hash
*
* Counts all objects grouped by type.
*
* It returns a hash, such as:
* {
* :TOTAL=>10000,
* :FREE=>3011,
* :T_OBJECT=>6,
* :T_CLASS=>404,
* # ...
* }
*
* The contents of the returned hash are implementation specific.
* It may be changed in future.
*
* The keys starting with +:T_+ means live objects.
* For example, +:T_ARRAY+ is the number of arrays.
* +:FREE+ means object slots which is not used now.
* +:TOTAL+ means sum of above.
*
* If the optional argument +result_hash+ is given,
* it is overwritten and returned. This is intended to avoid probe effect.
*
* h = {}
* ObjectSpace.count_objects(h)
* puts h
* # => { :TOTAL=>10000, :T_CLASS=>158280, :T_MODULE=>20672, :T_STRING=>527249 }
*
* This method is only expected to work on C Ruby.
*
*/
static VALUE
count_objects(int argc, VALUE *argv, VALUE os)
{
struct count_objects_data data = { 0 };
VALUE hash = Qnil;
if (rb_check_arity(argc, 0, 1) == 1) {
hash = argv[0];
if (!RB_TYPE_P(hash, T_HASH))
rb_raise(rb_eTypeError, "non-hash given");
}
rb_gc_impl_each_object(rb_gc_get_objspace(), count_objects_i, &data);
if (NIL_P(hash)) {
hash = rb_hash_new();
}
else if (!RHASH_EMPTY_P(hash)) {
rb_hash_stlike_foreach(hash, set_zero, hash);
}
rb_hash_aset(hash, ID2SYM(rb_intern("TOTAL")), SIZET2NUM(data.total));
rb_hash_aset(hash, ID2SYM(rb_intern("FREE")), SIZET2NUM(data.freed));
for (size_t i = 0; i <= T_MASK; i++) {
VALUE type = type_sym(i);
if (data.counts[i])
rb_hash_aset(hash, type, SIZET2NUM(data.counts[i]));
}
return hash;
}
#define SET_STACK_END SET_MACHINE_STACK_END(&ec->machine.stack_end)
#define STACK_START (ec->machine.stack_start)
#define STACK_END (ec->machine.stack_end)
#define STACK_LEVEL_MAX (ec->machine.stack_maxsize/sizeof(VALUE))
#if STACK_GROW_DIRECTION < 0
# define STACK_LENGTH (size_t)(STACK_START - STACK_END)
#elif STACK_GROW_DIRECTION > 0
# define STACK_LENGTH (size_t)(STACK_END - STACK_START + 1)
#else
# define STACK_LENGTH ((STACK_END < STACK_START) ? (size_t)(STACK_START - STACK_END) \
: (size_t)(STACK_END - STACK_START + 1))
#endif
#if !STACK_GROW_DIRECTION
int ruby_stack_grow_direction;
int
ruby_get_stack_grow_direction(volatile VALUE *addr)
{
VALUE *end;
SET_MACHINE_STACK_END(&end);
if (end > addr) return ruby_stack_grow_direction = 1;
return ruby_stack_grow_direction = -1;
}
#endif
size_t
ruby_stack_length(VALUE **p)
{
rb_execution_context_t *ec = GET_EC();
SET_STACK_END;
if (p) *p = STACK_UPPER(STACK_END, STACK_START, STACK_END);
return STACK_LENGTH;
}
#define PREVENT_STACK_OVERFLOW 1
#ifndef PREVENT_STACK_OVERFLOW
#if !(defined(POSIX_SIGNAL) && defined(SIGSEGV) && defined(HAVE_SIGALTSTACK))
# define PREVENT_STACK_OVERFLOW 1
#else
# define PREVENT_STACK_OVERFLOW 0
#endif
#endif
#if PREVENT_STACK_OVERFLOW && !defined(__EMSCRIPTEN__)
static int
stack_check(rb_execution_context_t *ec, int water_mark)
{
SET_STACK_END;
size_t length = STACK_LENGTH;
size_t maximum_length = STACK_LEVEL_MAX - water_mark;
return length > maximum_length;
}
#else
#define stack_check(ec, water_mark) FALSE
#endif
#define STACKFRAME_FOR_CALL_CFUNC 2048
int
rb_ec_stack_check(rb_execution_context_t *ec)
{
return stack_check(ec, STACKFRAME_FOR_CALL_CFUNC);
}
int
ruby_stack_check(void)
{
return stack_check(GET_EC(), STACKFRAME_FOR_CALL_CFUNC);
}
2024-08-28 11:18:43 -04:00
/* ==================== Marking ==================== */
#define RB_GC_MARK_OR_TRAVERSE(func, obj_or_ptr, obj, check_obj) do { \
if (!RB_SPECIAL_CONST_P(obj)) { \
rb_vm_t *vm = GET_VM(); \
void *objspace = vm->gc.objspace; \
if (LIKELY(vm->gc.mark_func_data == NULL)) { \
GC_ASSERT(rb_gc_impl_during_gc_p(objspace)); \
(func)(objspace, (obj_or_ptr)); \
} \
else if (check_obj ? \
rb_gc_impl_pointer_to_heap_p(objspace, (const void *)obj) && \
!rb_gc_impl_garbage_object_p(objspace, obj) : \
true) { \
GC_ASSERT(!rb_gc_impl_during_gc_p(objspace)); \
struct gc_mark_func_data_struct *mark_func_data = vm->gc.mark_func_data; \
vm->gc.mark_func_data = NULL; \
mark_func_data->mark_func((obj), mark_func_data->data); \
vm->gc.mark_func_data = mark_func_data; \
} \
} \
} while (0)
static inline void
gc_mark_internal(VALUE obj)
{
RB_GC_MARK_OR_TRAVERSE(rb_gc_impl_mark, obj, obj, false);
}
2024-08-28 11:18:43 -04:00
void
rb_gc_mark_movable(VALUE obj)
{
gc_mark_internal(obj);
2024-08-28 11:18:43 -04:00
}
void
rb_gc_mark_and_move(VALUE *ptr)
{
RB_GC_MARK_OR_TRAVERSE(rb_gc_impl_mark_and_move, ptr, *ptr, false);
2024-08-28 11:18:43 -04:00
}
static inline void
gc_mark_and_pin_internal(VALUE obj)
{
RB_GC_MARK_OR_TRAVERSE(rb_gc_impl_mark_and_pin, obj, obj, false);
}
2024-08-28 11:18:43 -04:00
void
rb_gc_mark(VALUE obj)
{
gc_mark_and_pin_internal(obj);
}
static inline void
gc_mark_maybe_internal(VALUE obj)
{
RB_GC_MARK_OR_TRAVERSE(rb_gc_impl_mark_maybe, obj, obj, true);
2024-08-28 11:18:43 -04:00
}
void
rb_gc_mark_maybe(VALUE obj)
{
gc_mark_maybe_internal(obj);
2024-08-28 11:18:43 -04:00
}
void
rb_gc_mark_weak(VALUE *ptr)
{
if (RB_SPECIAL_CONST_P(*ptr)) return;
rb_vm_t *vm = GET_VM();
void *objspace = vm->gc.objspace;
if (LIKELY(vm->gc.mark_func_data == NULL)) {
GC_ASSERT(rb_gc_impl_during_gc_p(objspace));
rb_gc_impl_mark_weak(objspace, ptr);
}
else {
GC_ASSERT(!rb_gc_impl_during_gc_p(objspace));
}
2024-08-28 11:18:43 -04:00
}
void
rb_gc_remove_weak(VALUE parent_obj, VALUE *ptr)
{
rb_gc_impl_remove_weak(rb_gc_get_objspace(), parent_obj, ptr);
}
ATTRIBUTE_NO_ADDRESS_SAFETY_ANALYSIS(static void each_location(register const VALUE *x, register long n, void (*cb)(VALUE, void *), void *data));
static void
each_location(register const VALUE *x, register long n, void (*cb)(VALUE, void *), void *data)
{
VALUE v;
while (n--) {
v = *x;
cb(v, data);
x++;
}
}
static void
each_location_ptr(const VALUE *start, const VALUE *end, void (*cb)(VALUE, void *), void *data)
{
if (end <= start) return;
each_location(start, end - start, cb, data);
}
static void
gc_mark_maybe_each_location(VALUE obj, void *data)
{
gc_mark_maybe_internal(obj);
}
void
rb_gc_mark_locations(const VALUE *start, const VALUE *end)
{
each_location_ptr(start, end, gc_mark_maybe_each_location, NULL);
}
void
rb_gc_mark_values(long n, const VALUE *values)
{
for (long i = 0; i < n; i++) {
gc_mark_internal(values[i]);
}
}
void
rb_gc_mark_vm_stack_values(long n, const VALUE *values)
{
for (long i = 0; i < n; i++) {
gc_mark_and_pin_internal(values[i]);
}
}
static int
mark_key(st_data_t key, st_data_t value, st_data_t data)
{
gc_mark_and_pin_internal((VALUE)key);
return ST_CONTINUE;
}
void
rb_mark_set(st_table *tbl)
{
if (!tbl) return;
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
st_foreach(tbl, mark_key, (st_data_t)rb_gc_get_objspace());
}
static int
mark_keyvalue(st_data_t key, st_data_t value, st_data_t data)
{
gc_mark_internal((VALUE)key);
gc_mark_internal((VALUE)value);
return ST_CONTINUE;
}
static int
pin_key_pin_value(st_data_t key, st_data_t value, st_data_t data)
{
gc_mark_and_pin_internal((VALUE)key);
gc_mark_and_pin_internal((VALUE)value);
return ST_CONTINUE;
}
static int
pin_key_mark_value(st_data_t key, st_data_t value, st_data_t data)
{
gc_mark_and_pin_internal((VALUE)key);
gc_mark_internal((VALUE)value);
return ST_CONTINUE;
}
static void
mark_hash(VALUE hash)
{
if (rb_hash_compare_by_id_p(hash)) {
rb_hash_stlike_foreach(hash, pin_key_mark_value, 0);
}
else {
rb_hash_stlike_foreach(hash, mark_keyvalue, 0);
}
gc_mark_internal(RHASH(hash)->ifnone);
}
void
rb_mark_hash(st_table *tbl)
{
if (!tbl) return;
st_foreach(tbl, pin_key_pin_value, 0);
}
static enum rb_id_table_iterator_result
mark_method_entry_i(VALUE me, void *objspace)
{
gc_mark_internal(me);
return ID_TABLE_CONTINUE;
}
static void
mark_m_tbl(void *objspace, struct rb_id_table *tbl)
{
if (tbl) {
rb_id_table_foreach_values(tbl, mark_method_entry_i, objspace);
}
}
2025-04-30 13:48:02 +09:00
static enum rb_id_table_iterator_result
mark_const_entry_i(VALUE value, void *objspace)
{
const rb_const_entry_t *ce = (const rb_const_entry_t *)value;
gc_mark_internal(ce->value);
gc_mark_internal(ce->file);
return ID_TABLE_CONTINUE;
}
static void
mark_const_tbl(rb_objspace_t *objspace, struct rb_id_table *tbl)
{
if (!tbl) return;
rb_id_table_foreach_values(tbl, mark_const_entry_i, objspace);
}
struct mark_cc_entry_args {
rb_objspace_t *objspace;
VALUE klass;
};
static enum rb_id_table_iterator_result
mark_cc_entry_i(VALUE ccs_ptr, void *data)
2025-04-30 13:48:02 +09:00
{
struct rb_class_cc_entries *ccs = (struct rb_class_cc_entries *)ccs_ptr;
VM_ASSERT(vm_ccs_p(ccs));
if (METHOD_ENTRY_INVALIDATED(ccs->cme)) {
rb_vm_ccs_free(ccs);
return ID_TABLE_DELETE;
}
else {
gc_mark_internal((VALUE)ccs->cme);
for (int i=0; i<ccs->len; i++) {
VM_ASSERT(((struct mark_cc_entry_args *)data)->klass == ccs->entries[i].cc->klass);
2025-04-30 13:48:02 +09:00
VM_ASSERT(vm_cc_check_cme(ccs->entries[i].cc, ccs->cme));
gc_mark_internal((VALUE)ccs->entries[i].cc);
}
return ID_TABLE_CONTINUE;
}
}
static void
mark_cc_tbl(rb_objspace_t *objspace, struct rb_id_table *tbl, VALUE klass)
{
struct mark_cc_entry_args args;
if (!tbl) return;
args.objspace = objspace;
args.klass = klass;
rb_id_table_foreach_values(tbl, mark_cc_entry_i, (void *)&args);
2025-04-30 13:48:02 +09:00
}
static enum rb_id_table_iterator_result
mark_cvc_tbl_i(VALUE cvc_entry, void *objspace)
{
struct rb_cvar_class_tbl_entry *entry;
entry = (struct rb_cvar_class_tbl_entry *)cvc_entry;
RUBY_ASSERT(entry->cref == 0 || (BUILTIN_TYPE((VALUE)entry->cref) == T_IMEMO && IMEMO_TYPE_P(entry->cref, imemo_cref)));
gc_mark_internal((VALUE)entry->cref);
return ID_TABLE_CONTINUE;
}
static void
mark_cvc_tbl(rb_objspace_t *objspace, struct rb_id_table *tbl)
{
if (!tbl) return;
rb_id_table_foreach_values(tbl, mark_cvc_tbl_i, objspace);
}
#if STACK_GROW_DIRECTION < 0
#define GET_STACK_BOUNDS(start, end, appendix) ((start) = STACK_END, (end) = STACK_START)
#elif STACK_GROW_DIRECTION > 0
#define GET_STACK_BOUNDS(start, end, appendix) ((start) = STACK_START, (end) = STACK_END+(appendix))
#else
#define GET_STACK_BOUNDS(start, end, appendix) \
((STACK_END < STACK_START) ? \
((start) = STACK_END, (end) = STACK_START) : ((start) = STACK_START, (end) = STACK_END+(appendix)))
#endif
static void
gc_mark_machine_stack_location_maybe(VALUE obj, void *data)
{
gc_mark_maybe_internal(obj);
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
#ifdef RUBY_ASAN_ENABLED
const rb_execution_context_t *ec = (const rb_execution_context_t *)data;
void *fake_frame_start;
void *fake_frame_end;
bool is_fake_frame = asan_get_fake_stack_extents(
ec->machine.asan_fake_stack_handle, obj,
ec->machine.stack_start, ec->machine.stack_end,
&fake_frame_start, &fake_frame_end
);
if (is_fake_frame) {
each_location_ptr(fake_frame_start, fake_frame_end, gc_mark_maybe_each_location, NULL);
}
* gc.c: support RGENGC. [ruby-trunk - Feature #8339] See this ticet about RGENGC. * gc.c: Add several flags: * RGENGC_DEBUG: if >0, then prints debug information. * RGENGC_CHECK_MODE: if >0, add assertions. * RGENGC_PROFILE: if >0, add profiling features. check GC.stat and GC::Profiler. * include/ruby/ruby.h: disable RGENGC by default (USE_RGENGC == 0). * array.c: add write barriers for T_ARRAY and generate sunny objects. * include/ruby/ruby.h (RARRAY_PTR_USE): added. Use this macro if you want to access raw pointers. If you modify the contents which pointer pointed, then you need to care write barrier. * bignum.c, marshal.c, random.c: generate T_BIGNUM sunny objects. * complex.c, include/ruby/ruby.h: add write barriers for T_COMPLEX and generate sunny objects. * rational.c (nurat_s_new_internal), include/ruby/ruby.h: add write barriers for T_RATIONAL and generate sunny objects. * internal.h: add write barriers for RBasic::klass. * numeric.c (rb_float_new_in_heap): generate sunny T_FLOAT objects. * object.c (rb_class_allocate_instance), range.c: generate sunny T_OBJECT objects. * string.c: add write barriers for T_STRING and generate sunny objects. * variable.c: add write barriers for ivars. * vm_insnhelper.c (vm_setivar): ditto. * include/ruby/ruby.h, debug.c: use two flags FL_WB_PROTECTED and FL_OLDGEN. * node.h (NODE_FL_CREF_PUSHED_BY_EVAL, NODE_FL_CREF_OMOD_SHARED): move flag bits. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40703 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2013-05-13 18:07:47 +00:00
#endif
}
* gc.c: support RGENGC. [ruby-trunk - Feature #8339] See this ticet about RGENGC. * gc.c: Add several flags: * RGENGC_DEBUG: if >0, then prints debug information. * RGENGC_CHECK_MODE: if >0, add assertions. * RGENGC_PROFILE: if >0, add profiling features. check GC.stat and GC::Profiler. * include/ruby/ruby.h: disable RGENGC by default (USE_RGENGC == 0). * array.c: add write barriers for T_ARRAY and generate sunny objects. * include/ruby/ruby.h (RARRAY_PTR_USE): added. Use this macro if you want to access raw pointers. If you modify the contents which pointer pointed, then you need to care write barrier. * bignum.c, marshal.c, random.c: generate T_BIGNUM sunny objects. * complex.c, include/ruby/ruby.h: add write barriers for T_COMPLEX and generate sunny objects. * rational.c (nurat_s_new_internal), include/ruby/ruby.h: add write barriers for T_RATIONAL and generate sunny objects. * internal.h: add write barriers for RBasic::klass. * numeric.c (rb_float_new_in_heap): generate sunny T_FLOAT objects. * object.c (rb_class_allocate_instance), range.c: generate sunny T_OBJECT objects. * string.c: add write barriers for T_STRING and generate sunny objects. * variable.c: add write barriers for ivars. * vm_insnhelper.c (vm_setivar): ditto. * include/ruby/ruby.h, debug.c: use two flags FL_WB_PROTECTED and FL_OLDGEN. * node.h (NODE_FL_CREF_PUSHED_BY_EVAL, NODE_FL_CREF_OMOD_SHARED): move flag bits. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40703 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2013-05-13 18:07:47 +00:00
static VALUE
gc_location_internal(void *objspace, VALUE value)
{
if (SPECIAL_CONST_P(value)) {
return value;
}
GC_ASSERT(rb_gc_impl_pointer_to_heap_p(objspace, (void *)value));
return rb_gc_impl_location(objspace, value);
}
VALUE
rb_gc_location(VALUE value)
{
return gc_location_internal(rb_gc_get_objspace(), value);
}
#if defined(__wasm__)
* gc.c: support RGENGC. [ruby-trunk - Feature #8339] See this ticet about RGENGC. * gc.c: Add several flags: * RGENGC_DEBUG: if >0, then prints debug information. * RGENGC_CHECK_MODE: if >0, add assertions. * RGENGC_PROFILE: if >0, add profiling features. check GC.stat and GC::Profiler. * include/ruby/ruby.h: disable RGENGC by default (USE_RGENGC == 0). * array.c: add write barriers for T_ARRAY and generate sunny objects. * include/ruby/ruby.h (RARRAY_PTR_USE): added. Use this macro if you want to access raw pointers. If you modify the contents which pointer pointed, then you need to care write barrier. * bignum.c, marshal.c, random.c: generate T_BIGNUM sunny objects. * complex.c, include/ruby/ruby.h: add write barriers for T_COMPLEX and generate sunny objects. * rational.c (nurat_s_new_internal), include/ruby/ruby.h: add write barriers for T_RATIONAL and generate sunny objects. * internal.h: add write barriers for RBasic::klass. * numeric.c (rb_float_new_in_heap): generate sunny T_FLOAT objects. * object.c (rb_class_allocate_instance), range.c: generate sunny T_OBJECT objects. * string.c: add write barriers for T_STRING and generate sunny objects. * variable.c: add write barriers for ivars. * vm_insnhelper.c (vm_setivar): ditto. * include/ruby/ruby.h, debug.c: use two flags FL_WB_PROTECTED and FL_OLDGEN. * node.h (NODE_FL_CREF_PUSHED_BY_EVAL, NODE_FL_CREF_OMOD_SHARED): move flag bits. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40703 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2013-05-13 18:07:47 +00:00
static VALUE *rb_stack_range_tmp[2];
static void
rb_mark_locations(void *begin, void *end)
{
rb_stack_range_tmp[0] = begin;
rb_stack_range_tmp[1] = end;
}
void
rb_gc_save_machine_context(void)
{
// no-op
}
# if defined(__EMSCRIPTEN__)
static void
mark_current_machine_context(const rb_execution_context_t *ec)
{
emscripten_scan_stack(rb_mark_locations);
each_location_ptr(rb_stack_range_tmp[0], rb_stack_range_tmp[1], gc_mark_maybe_each_location, NULL);
emscripten_scan_registers(rb_mark_locations);
each_location_ptr(rb_stack_range_tmp[0], rb_stack_range_tmp[1], gc_mark_maybe_each_location, NULL);
}
# else // use Asyncify version
static void
mark_current_machine_context(rb_execution_context_t *ec)
{
VALUE *stack_start, *stack_end;
SET_STACK_END;
GET_STACK_BOUNDS(stack_start, stack_end, 1);
each_location_ptr(stack_start, stack_end, gc_mark_maybe_each_location, NULL);
rb_wasm_scan_locals(rb_mark_locations);
each_location_ptr(rb_stack_range_tmp[0], rb_stack_range_tmp[1], gc_mark_maybe_each_location, NULL);
}
# endif
#else // !defined(__wasm__)
void
rb_gc_save_machine_context(void)
{
rb_thread_t *thread = GET_THREAD();
RB_VM_SAVE_MACHINE_CONTEXT(thread);
}
static void
mark_current_machine_context(const rb_execution_context_t *ec)
{
rb_gc_mark_machine_context(ec);
}
#endif
void
rb_gc_mark_machine_context(const rb_execution_context_t *ec)
{
VALUE *stack_start, *stack_end;
GET_STACK_BOUNDS(stack_start, stack_end, 0);
RUBY_DEBUG_LOG("ec->th:%u stack_start:%p stack_end:%p", rb_ec_thread_ptr(ec)->serial, stack_start, stack_end);
void *data =
#ifdef RUBY_ASAN_ENABLED
/* gc_mark_machine_stack_location_maybe() uses data as const */
(rb_execution_context_t *)ec;
#else
NULL;
#endif
2024-07-03 14:40:05 -04:00
each_location_ptr(stack_start, stack_end, gc_mark_machine_stack_location_maybe, data);
2024-07-03 14:40:05 -04:00
int num_regs = sizeof(ec->machine.regs)/(sizeof(VALUE));
each_location((VALUE*)&ec->machine.regs, num_regs, gc_mark_machine_stack_location_maybe, data);
}
static int
rb_mark_tbl_i(st_data_t key, st_data_t value, st_data_t data)
{
gc_mark_and_pin_internal((VALUE)value);
return ST_CONTINUE;
}
void
rb_mark_tbl(st_table *tbl)
{
if (!tbl || tbl->num_entries == 0) return;
st_foreach(tbl, rb_mark_tbl_i, 0);
}
static void
gc_mark_tbl_no_pin(st_table *tbl)
{
if (!tbl || tbl->num_entries == 0) return;
st_foreach(tbl, gc_mark_tbl_no_pin_i, 0);
}
void
rb_mark_tbl_no_pin(st_table *tbl)
{
gc_mark_tbl_no_pin(tbl);
}
static bool
gc_declarative_marking_p(const rb_data_type_t *type)
{
return (type->flags & RUBY_TYPED_DECL_MARKING) != 0;
}
void
rb_gc_mark_roots(void *objspace, const char **categoryp)
{
rb_execution_context_t *ec = GET_EC();
rb_vm_t *vm = rb_ec_vm_ptr(ec);
#define MARK_CHECKPOINT(category) do { \
if (categoryp) *categoryp = category; \
} while (0)
MARK_CHECKPOINT("vm");
rb_vm_mark(vm);
if (vm->self) gc_mark_internal(vm->self);
MARK_CHECKPOINT("end_proc");
rb_mark_end_proc();
MARK_CHECKPOINT("global_tbl");
rb_gc_mark_global_tbl();
#if USE_YJIT
void rb_yjit_root_mark(void); // in Rust
if (rb_yjit_enabled_p) {
MARK_CHECKPOINT("YJIT");
rb_yjit_root_mark();
}
#endif
MARK_CHECKPOINT("machine_context");
mark_current_machine_context(ec);
MARK_CHECKPOINT("global_symbols");
rb_sym_global_symbols_mark();
MARK_CHECKPOINT("finish");
#undef MARK_CHECKPOINT
}
2025-04-30 13:48:02 +09:00
struct gc_mark_classext_foreach_arg {
rb_objspace_t *objspace;
VALUE obj;
};
static void
gc_mark_classext_module(rb_classext_t *ext, bool prime, VALUE namespace, void *arg)
{
struct gc_mark_classext_foreach_arg *foreach_arg = (struct gc_mark_classext_foreach_arg *)arg;
rb_objspace_t *objspace = foreach_arg->objspace;
VALUE obj = foreach_arg->obj;
if (RCLASSEXT_SUPER(ext)) {
gc_mark_internal(RCLASSEXT_SUPER(ext));
}
mark_m_tbl(objspace, RCLASSEXT_M_TBL(ext));
gc_mark_internal(RCLASSEXT_FIELDS_OBJ(ext));
2025-04-30 13:48:02 +09:00
if (!RCLASSEXT_SHARED_CONST_TBL(ext) && RCLASSEXT_CONST_TBL(ext)) {
mark_const_tbl(objspace, RCLASSEXT_CONST_TBL(ext));
}
mark_m_tbl(objspace, RCLASSEXT_CALLABLE_M_TBL(ext));
mark_cc_tbl(objspace, RCLASSEXT_CC_TBL(ext), obj);
mark_cvc_tbl(objspace, RCLASSEXT_CVC_TBL(ext));
gc_mark_internal(RCLASSEXT_CLASSPATH(ext));
}
static void
gc_mark_classext_iclass(rb_classext_t *ext, bool prime, VALUE namespace, void *arg)
{
struct gc_mark_classext_foreach_arg *foreach_arg = (struct gc_mark_classext_foreach_arg *)arg;
rb_objspace_t *objspace = foreach_arg->objspace;
VALUE iclass = foreach_arg->obj;
if (RCLASSEXT_SUPER(ext)) {
gc_mark_internal(RCLASSEXT_SUPER(ext));
}
if (RCLASSEXT_ICLASS_IS_ORIGIN(ext) && !RCLASSEXT_ICLASS_ORIGIN_SHARED_MTBL(ext)) {
mark_m_tbl(objspace, RCLASSEXT_M_TBL(ext));
}
if (RCLASSEXT_INCLUDER(ext)) {
gc_mark_internal(RCLASSEXT_INCLUDER(ext));
}
mark_m_tbl(objspace, RCLASSEXT_CALLABLE_M_TBL(ext));
mark_cc_tbl(objspace, RCLASSEXT_CC_TBL(ext), iclass);
}
Save one VALUE per embedded RTypedData This halves the amount of memory used for embedded RTypedData if they are one VALUE (8 bytes on 64-bit platforms) over the slot size limit. For Set, on 64-bit it uses an embedded 56-byte struct. With the previous implementation, the embedded structs starts at offset 32, resulting in a total size of 88. Since that is over the 80 byte limit, it goes to the next highest bucket, 160 bytes, wasting 72 bytes. This allows it to fit in a 80 byte bucket, which reduces the total size for small sets of from 224 bytes (160 bytes embedded, 64 bytes malloc, 72 bytes wasted in embedding) to 144 bytes (80 bytes embedded, 64 bytes malloc, 0 bytes wasted in embedding). Any other embedded RTypedData will see similar advantages if they are currently one VALUE over the limit. To implement this, remove the typed_flag from struct RTypedData. Embed the typed_flag information in the type member, which is now a tagged pointer using VALUE type, using the bottom low 2 bits as flags (1 bit for typed flag, the other for the embedded flag). To get the actual pointer, RTYPEDDATA_TYPE masks out the low 2 bits and then casts. That moves the RTypedData data pointer from offset 32 to offset 24 (on 64-bit). Vast amount of code in the internals (and probably external C extensions) expects the following code to work for both RData and non-embedded RTypedData: ```c DATA_PTR(obj) = some_pointer; ``` Allow this to work by moving the data pointer in RData between the dmark and dfree pointers, so it is at the same offset (24 on 64-bit). Other than these changes to the include files, the only changes needed were to gc.c, to account for the new struct layouts, handle setting the low bits in the type member, and to use RTYPEDDATA_TYPE(obj) instead of RTYPEDDATA(obj)->type.
2025-04-27 17:23:04 -07:00
#define TYPED_DATA_REFS_OFFSET_LIST(d) (size_t *)(uintptr_t)RTYPEDDATA_TYPE(d)->function.dmark
void
rb_gc_mark_children(void *objspace, VALUE obj)
{
2025-04-30 13:48:02 +09:00
struct gc_mark_classext_foreach_arg foreach_args;
if (FL_TEST_RAW(obj, FL_EXIVAR)) {
rb_mark_generic_ivar(obj);
}
switch (BUILTIN_TYPE(obj)) {
case T_FLOAT:
case T_BIGNUM:
case T_SYMBOL:
/* Not immediates, but does not have references and singleton class.
*
* RSYMBOL(obj)->fstr intentionally not marked. See log for 96815f1e
* ("symbol.c: remove rb_gc_mark_symbols()") */
return;
case T_NIL:
case T_FIXNUM:
rb_bug("rb_gc_mark() called for broken object");
break;
case T_NODE:
UNEXPECTED_NODE(rb_gc_mark);
break;
case T_IMEMO:
rb_imemo_mark_and_move(obj, false);
return;
default:
break;
}
gc_mark_internal(RBASIC(obj)->klass);
switch (BUILTIN_TYPE(obj)) {
case T_CLASS:
if (FL_TEST_RAW(obj, FL_SINGLETON)) {
gc_mark_internal(RCLASS_ATTACHED_OBJECT(obj));
}
// Continue to the shared T_CLASS/T_MODULE
case T_MODULE:
2025-04-30 13:48:02 +09:00
foreach_args.objspace = objspace;
foreach_args.obj = obj;
rb_class_classext_foreach(obj, gc_mark_classext_module, (void *)&foreach_args);
break;
Add a cache for class variables Redo of 34a2acdac788602c14bf05fb616215187badd504 and 931138b00696419945dc03e10f033b1f53cd50f3 which were reverted. GitHub PR #4340. This change implements a cache for class variables. Previously there was no cache for cvars. Cvar access is slow due to needing to travel all the way up th ancestor tree before returning the cvar value. The deeper the ancestor tree the slower cvar access will be. The benefits of the cache are more visible with a higher number of included modules due to the way Ruby looks up class variables. The benchmark here includes 26 modules and shows with the cache, this branch is 6.5x faster when accessing class variables. ``` compare-ruby: ruby 3.1.0dev (2021-03-15T06:22:34Z master 9e5105c) [x86_64-darwin19] built-ruby: ruby 3.1.0dev (2021-03-15T12:12:44Z add-cache-for-clas.. c6be009) [x86_64-darwin19] | |compare-ruby|built-ruby| |:--------|-----------:|---------:| |vm_cvar | 5.681M| 36.980M| | | -| 6.51x| ``` Benchmark.ips calling `ActiveRecord::Base.logger` from within a Rails application. ActiveRecord::Base.logger has 71 ancestors. The more ancestors a tree has, the more clear the speed increase. IE if Base had only one ancestor we'd see no improvement. This benchmark is run on a vanilla Rails application. Benchmark code: ```ruby require "benchmark/ips" require_relative "config/environment" Benchmark.ips do |x| x.report "logger" do ActiveRecord::Base.logger end end ``` Ruby 3.0 master / Rails 6.1: ``` Warming up -------------------------------------- logger 155.251k i/100ms Calculating ------------------------------------- ``` Ruby 3.0 with cvar cache / Rails 6.1: ``` Warming up -------------------------------------- logger 1.546M i/100ms Calculating ------------------------------------- logger 14.857M (± 4.8%) i/s - 74.198M in 5.006202s ``` Lastly we ran a benchmark to demonstate the difference between master and our cache when the number of modules increases. This benchmark measures 1 ancestor, 30 ancestors, and 100 ancestors. Ruby 3.0 master: ``` Warming up -------------------------------------- 1 module 1.231M i/100ms 30 modules 432.020k i/100ms 100 modules 145.399k i/100ms Calculating ------------------------------------- 1 module 12.210M (± 2.1%) i/s - 61.553M in 5.043400s 30 modules 4.354M (± 2.7%) i/s - 22.033M in 5.063839s 100 modules 1.434M (± 2.9%) i/s - 7.270M in 5.072531s Comparison: 1 module: 12209958.3 i/s 30 modules: 4354217.8 i/s - 2.80x (± 0.00) slower 100 modules: 1434447.3 i/s - 8.51x (± 0.00) slower ``` Ruby 3.0 with cvar cache: ``` Warming up -------------------------------------- 1 module 1.641M i/100ms 30 modules 1.655M i/100ms 100 modules 1.620M i/100ms Calculating ------------------------------------- 1 module 16.279M (± 3.8%) i/s - 82.038M in 5.046923s 30 modules 15.891M (± 3.9%) i/s - 79.459M in 5.007958s 100 modules 16.087M (± 3.6%) i/s - 81.005M in 5.041931s Comparison: 1 module: 16279458.0 i/s 100 modules: 16087484.6 i/s - same-ish: difference falls within error 30 modules: 15891406.2 i/s - same-ish: difference falls within error ``` Co-authored-by: Aaron Patterson <tenderlove@ruby-lang.org>
2021-06-01 13:34:06 -04:00
case T_ICLASS:
2025-04-30 13:48:02 +09:00
foreach_args.objspace = objspace;
foreach_args.obj = obj;
rb_class_classext_foreach(obj, gc_mark_classext_iclass, (void *)&foreach_args);
break;
case T_ARRAY:
if (ARY_SHARED_P(obj)) {
VALUE root = ARY_SHARED_ROOT(obj);
gc_mark_internal(root);
}
else {
long len = RARRAY_LEN(obj);
const VALUE *ptr = RARRAY_CONST_PTR(obj);
for (long i = 0; i < len; i++) {
gc_mark_internal(ptr[i]);
}
}
break;
case T_HASH:
mark_hash(obj);
break;
case T_STRING:
if (STR_SHARED_P(obj)) {
if (STR_EMBED_P(RSTRING(obj)->as.heap.aux.shared)) {
/* Embedded shared strings cannot be moved because this string
* points into the slot of the shared string. There may be code
* using the RSTRING_PTR on the stack, which would pin this
* string but not pin the shared string, causing it to move. */
gc_mark_and_pin_internal(RSTRING(obj)->as.heap.aux.shared);
}
else {
gc_mark_internal(RSTRING(obj)->as.heap.aux.shared);
}
}
break;
case T_DATA: {
void *const ptr = RTYPEDDATA_P(obj) ? RTYPEDDATA_GET_DATA(obj) : DATA_PTR(obj);
if (ptr) {
Save one VALUE per embedded RTypedData This halves the amount of memory used for embedded RTypedData if they are one VALUE (8 bytes on 64-bit platforms) over the slot size limit. For Set, on 64-bit it uses an embedded 56-byte struct. With the previous implementation, the embedded structs starts at offset 32, resulting in a total size of 88. Since that is over the 80 byte limit, it goes to the next highest bucket, 160 bytes, wasting 72 bytes. This allows it to fit in a 80 byte bucket, which reduces the total size for small sets of from 224 bytes (160 bytes embedded, 64 bytes malloc, 72 bytes wasted in embedding) to 144 bytes (80 bytes embedded, 64 bytes malloc, 0 bytes wasted in embedding). Any other embedded RTypedData will see similar advantages if they are currently one VALUE over the limit. To implement this, remove the typed_flag from struct RTypedData. Embed the typed_flag information in the type member, which is now a tagged pointer using VALUE type, using the bottom low 2 bits as flags (1 bit for typed flag, the other for the embedded flag). To get the actual pointer, RTYPEDDATA_TYPE masks out the low 2 bits and then casts. That moves the RTypedData data pointer from offset 32 to offset 24 (on 64-bit). Vast amount of code in the internals (and probably external C extensions) expects the following code to work for both RData and non-embedded RTypedData: ```c DATA_PTR(obj) = some_pointer; ``` Allow this to work by moving the data pointer in RData between the dmark and dfree pointers, so it is at the same offset (24 on 64-bit). Other than these changes to the include files, the only changes needed were to gc.c, to account for the new struct layouts, handle setting the low bits in the type member, and to use RTYPEDDATA_TYPE(obj) instead of RTYPEDDATA(obj)->type.
2025-04-27 17:23:04 -07:00
if (RTYPEDDATA_P(obj) && gc_declarative_marking_p(RTYPEDDATA_TYPE(obj))) {
size_t *offset_list = TYPED_DATA_REFS_OFFSET_LIST(obj);
for (size_t offset = *offset_list; offset != RUBY_REF_END; offset = *offset_list++) {
gc_mark_internal(*(VALUE *)((char *)ptr + offset));
}
}
else {
RUBY_DATA_FUNC mark_func = RTYPEDDATA_P(obj) ?
Save one VALUE per embedded RTypedData This halves the amount of memory used for embedded RTypedData if they are one VALUE (8 bytes on 64-bit platforms) over the slot size limit. For Set, on 64-bit it uses an embedded 56-byte struct. With the previous implementation, the embedded structs starts at offset 32, resulting in a total size of 88. Since that is over the 80 byte limit, it goes to the next highest bucket, 160 bytes, wasting 72 bytes. This allows it to fit in a 80 byte bucket, which reduces the total size for small sets of from 224 bytes (160 bytes embedded, 64 bytes malloc, 72 bytes wasted in embedding) to 144 bytes (80 bytes embedded, 64 bytes malloc, 0 bytes wasted in embedding). Any other embedded RTypedData will see similar advantages if they are currently one VALUE over the limit. To implement this, remove the typed_flag from struct RTypedData. Embed the typed_flag information in the type member, which is now a tagged pointer using VALUE type, using the bottom low 2 bits as flags (1 bit for typed flag, the other for the embedded flag). To get the actual pointer, RTYPEDDATA_TYPE masks out the low 2 bits and then casts. That moves the RTypedData data pointer from offset 32 to offset 24 (on 64-bit). Vast amount of code in the internals (and probably external C extensions) expects the following code to work for both RData and non-embedded RTypedData: ```c DATA_PTR(obj) = some_pointer; ``` Allow this to work by moving the data pointer in RData between the dmark and dfree pointers, so it is at the same offset (24 on 64-bit). Other than these changes to the include files, the only changes needed were to gc.c, to account for the new struct layouts, handle setting the low bits in the type member, and to use RTYPEDDATA_TYPE(obj) instead of RTYPEDDATA(obj)->type.
2025-04-27 17:23:04 -07:00
RTYPEDDATA_TYPE(obj)->function.dmark :
RDATA(obj)->dmark;
if (mark_func) (*mark_func)(ptr);
}
}
break;
}
case T_OBJECT: {
if (rb_shape_obj_too_complex_p(obj)) {
gc_mark_tbl_no_pin(ROBJECT_FIELDS_HASH(obj));
}
else {
const VALUE * const ptr = ROBJECT_FIELDS(obj);
uint32_t len = ROBJECT_FIELDS_COUNT(obj);
for (uint32_t i = 0; i < len; i++) {
gc_mark_internal(ptr[i]);
}
}
attr_index_t fields_count = ROBJECT_FIELDS_COUNT(obj);
if (fields_count) {
VALUE klass = RBASIC_CLASS(obj);
// Increment max_iv_count if applicable, used to determine size pool allocation
if (RCLASS_MAX_IV_COUNT(klass) < fields_count) {
RCLASS_SET_MAX_IV_COUNT(klass, fields_count);
}
}
break;
}
case T_FILE:
if (RFILE(obj)->fptr) {
gc_mark_internal(RFILE(obj)->fptr->self);
gc_mark_internal(RFILE(obj)->fptr->pathv);
gc_mark_internal(RFILE(obj)->fptr->tied_io_for_writing);
gc_mark_internal(RFILE(obj)->fptr->writeconv_asciicompat);
gc_mark_internal(RFILE(obj)->fptr->writeconv_pre_ecopts);
gc_mark_internal(RFILE(obj)->fptr->encs.ecopts);
gc_mark_internal(RFILE(obj)->fptr->write_lock);
gc_mark_internal(RFILE(obj)->fptr->timeout);
gc_mark_internal(RFILE(obj)->fptr->wakeup_mutex);
}
break;
case T_REGEXP:
gc_mark_internal(RREGEXP(obj)->src);
break;
case T_MATCH:
gc_mark_internal(RMATCH(obj)->regexp);
if (RMATCH(obj)->str) {
gc_mark_internal(RMATCH(obj)->str);
}
break;
case T_RATIONAL:
gc_mark_internal(RRATIONAL(obj)->num);
gc_mark_internal(RRATIONAL(obj)->den);
2020-04-08 15:13:37 +09:00
break;
case T_COMPLEX:
gc_mark_internal(RCOMPLEX(obj)->real);
gc_mark_internal(RCOMPLEX(obj)->imag);
break;
case T_STRUCT: {
const long len = RSTRUCT_LEN(obj);
const VALUE * const ptr = RSTRUCT_CONST_PTR(obj);
for (long i = 0; i < len; i++) {
gc_mark_internal(ptr[i]);
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
break;
}
default:
if (BUILTIN_TYPE(obj) == T_MOVED) rb_bug("rb_gc_mark(): %p is T_MOVED", (void *)obj);
if (BUILTIN_TYPE(obj) == T_NONE) rb_bug("rb_gc_mark(): %p is T_NONE", (void *)obj);
if (BUILTIN_TYPE(obj) == T_ZOMBIE) rb_bug("rb_gc_mark(): %p is T_ZOMBIE", (void *)obj);
rb_bug("rb_gc_mark(): unknown data type 0x%x(%p) %s",
BUILTIN_TYPE(obj), (void *)obj,
rb_gc_impl_pointer_to_heap_p(objspace, (void *)obj) ? "corrupted object" : "non object");
}
}
size_t
rb_gc_obj_optimal_size(VALUE obj)
{
switch (BUILTIN_TYPE(obj)) {
case T_ARRAY:
return rb_ary_size_as_embedded(obj);
case T_OBJECT:
if (rb_shape_obj_too_complex_p(obj)) {
return sizeof(struct RObject);
2023-10-19 16:01:35 -07:00
}
else {
return rb_obj_embedded_size(ROBJECT_FIELDS_CAPACITY(obj));
}
case T_STRING:
return rb_str_size_as_embedded(obj);
case T_HASH:
return sizeof(struct RHash) + (RHASH_ST_TABLE_P(obj) ? sizeof(st_table) : sizeof(ar_table));
default:
return 0;
}
}
void
rb_gc_writebarrier(VALUE a, VALUE b)
{
rb_gc_impl_writebarrier(rb_gc_get_objspace(), a, b);
}
void
rb_gc_writebarrier_unprotect(VALUE obj)
{
rb_gc_impl_writebarrier_unprotect(rb_gc_get_objspace(), obj);
}
/*
* remember `obj' if needed.
*/
void
rb_gc_writebarrier_remember(VALUE obj)
{
rb_gc_impl_writebarrier_remember(rb_gc_get_objspace(), obj);
}
void
rb_gc_copy_attributes(VALUE dest, VALUE obj)
{
rb_gc_impl_copy_attributes(rb_gc_get_objspace(), dest, obj);
}
int
rb_gc_modular_gc_loaded_p(void)
{
#if USE_MODULAR_GC
return rb_gc_functions.modular_gc_loaded_p;
#else
return false;
#endif
}
const char *
rb_gc_active_gc_name(void)
{
const char *gc_name = rb_gc_impl_active_gc_name();
const size_t len = strlen(gc_name);
if (len > RB_GC_MAX_NAME_LEN) {
rb_bug("GC should have a name no more than %d chars long. Currently: %zu (%s)",
RB_GC_MAX_NAME_LEN, len, gc_name);
}
return gc_name;
}
struct rb_gc_object_metadata_entry *
rb_gc_object_metadata(VALUE obj)
{
return rb_gc_impl_object_metadata(rb_gc_get_objspace(), obj);
}
/* GC */
void *
rb_gc_ractor_cache_alloc(rb_ractor_t *ractor)
{
return rb_gc_impl_ractor_cache_alloc(rb_gc_get_objspace(), ractor);
}
void
rb_gc_ractor_cache_free(void *cache)
{
rb_gc_impl_ractor_cache_free(rb_gc_get_objspace(), cache);
}
void
rb_gc_register_mark_object(VALUE obj)
{
if (!rb_gc_impl_pointer_to_heap_p(rb_gc_get_objspace(), (void *)obj))
return;
rb_vm_register_global_object(obj);
}
void
rb_gc_register_address(VALUE *addr)
{
rb_vm_t *vm = GET_VM();
VALUE obj = *addr;
struct global_object_list *tmp = ALLOC(struct global_object_list);
tmp->next = vm->global_object_list;
tmp->varptr = addr;
vm->global_object_list = tmp;
/*
* Because some C extensions have assignment-then-register bugs,
* we guard `obj` here so that it would not get swept defensively.
*/
RB_GC_GUARD(obj);
if (0 && !SPECIAL_CONST_P(obj)) {
rb_warn("Object is assigned to registering address already: %"PRIsVALUE,
rb_obj_class(obj));
rb_print_backtrace(stderr);
}
}
void
rb_gc_unregister_address(VALUE *addr)
{
rb_vm_t *vm = GET_VM();
struct global_object_list *tmp = vm->global_object_list;
if (tmp->varptr == addr) {
vm->global_object_list = tmp->next;
xfree(tmp);
return;
}
while (tmp->next) {
if (tmp->next->varptr == addr) {
struct global_object_list *t = tmp->next;
tmp->next = tmp->next->next;
xfree(t);
break;
}
tmp = tmp->next;
}
}
void
rb_global_variable(VALUE *var)
{
rb_gc_register_address(var);
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
static VALUE
gc_start_internal(rb_execution_context_t *ec, VALUE self, VALUE full_mark, VALUE immediate_mark, VALUE immediate_sweep, VALUE compact)
{
rb_gc_impl_start(rb_gc_get_objspace(), RTEST(full_mark), RTEST(immediate_mark), RTEST(immediate_sweep), RTEST(compact));
return Qnil;
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
/*
* rb_objspace_each_objects() is special C API to walk through
* Ruby object space. This C API is too difficult to use it.
* To be frank, you should not use it. Or you need to read the
* source code of this function and understand what this function does.
*
* 'callback' will be called several times (the number of heap page,
* at current implementation) with:
* vstart: a pointer to the first living object of the heap_page.
* vend: a pointer to next to the valid heap_page area.
* stride: a distance to next VALUE.
*
* If callback() returns non-zero, the iteration will be stopped.
*
* This is a sample callback code to iterate liveness objects:
*
* static int
* sample_callback(void *vstart, void *vend, int stride, void *data)
* {
* VALUE v = (VALUE)vstart;
* for (; v != (VALUE)vend; v += stride) {
* if (!rb_objspace_internal_object_p(v)) { // liveness check
* // do something with live object 'v'
* }
* }
* return 0; // continue to iteration
* }
*
* Note: 'vstart' is not a top of heap_page. This point the first
* living object to grasp at least one object to avoid GC issue.
* This means that you can not walk through all Ruby object page
* including freed object page.
*
* Note: On this implementation, 'stride' is the same as sizeof(RVALUE).
* However, there are possibilities to pass variable values with
* 'stride' with some reasons. You must use stride instead of
* use some constant value in the iteration.
*/
void
rb_objspace_each_objects(int (*callback)(void *, void *, size_t, void *), void *data)
{
rb_gc_impl_each_objects(rb_gc_get_objspace(), callback, data);
}
static void
gc_ref_update_array(void *objspace, VALUE v)
{
if (ARY_SHARED_P(v)) {
VALUE old_root = RARRAY(v)->as.heap.aux.shared_root;
UPDATE_IF_MOVED(objspace, RARRAY(v)->as.heap.aux.shared_root);
VALUE new_root = RARRAY(v)->as.heap.aux.shared_root;
// If the root is embedded and its location has changed
if (ARY_EMBED_P(new_root) && new_root != old_root) {
size_t offset = (size_t)(RARRAY(v)->as.heap.ptr - RARRAY(old_root)->as.ary);
GC_ASSERT(RARRAY(v)->as.heap.ptr >= RARRAY(old_root)->as.ary);
RARRAY(v)->as.heap.ptr = RARRAY(new_root)->as.ary + offset;
}
}
else {
long len = RARRAY_LEN(v);
if (len > 0) {
VALUE *ptr = (VALUE *)RARRAY_CONST_PTR(v);
for (long i = 0; i < len; i++) {
UPDATE_IF_MOVED(objspace, ptr[i]);
}
}
if (rb_gc_obj_slot_size(v) >= rb_ary_size_as_embedded(v)) {
if (rb_ary_embeddable_p(v)) {
rb_ary_make_embedded(v);
}
}
}
}
static void
gc_ref_update_object(void *objspace, VALUE v)
{
VALUE *ptr = ROBJECT_FIELDS(v);
if (rb_shape_obj_too_complex_p(v)) {
gc_ref_update_table_values_only(ROBJECT_FIELDS_HASH(v));
return;
}
size_t slot_size = rb_gc_obj_slot_size(v);
size_t embed_size = rb_obj_embedded_size(ROBJECT_FIELDS_CAPACITY(v));
if (slot_size >= embed_size && !RB_FL_TEST_RAW(v, ROBJECT_EMBED)) {
// Object can be re-embedded
memcpy(ROBJECT(v)->as.ary, ptr, sizeof(VALUE) * ROBJECT_FIELDS_COUNT(v));
RB_FL_SET_RAW(v, ROBJECT_EMBED);
xfree(ptr);
ptr = ROBJECT(v)->as.ary;
}
for (uint32_t i = 0; i < ROBJECT_FIELDS_COUNT(v); i++) {
UPDATE_IF_MOVED(objspace, ptr[i]);
}
}
void
rb_gc_ref_update_table_values_only(st_table *tbl)
{
gc_ref_update_table_values_only(tbl);
}
/* Update MOVED references in a VALUE=>VALUE st_table */
void
rb_gc_update_tbl_refs(st_table *ptr)
{
gc_update_table_refs(ptr);
}
static void
gc_ref_update_hash(void *objspace, VALUE v)
{
rb_hash_stlike_foreach_with_replace(v, hash_foreach_replace, hash_replace_ref, (st_data_t)objspace);
}
static void
gc_update_values(void *objspace, long n, VALUE *values)
{
for (long i = 0; i < n; i++) {
UPDATE_IF_MOVED(objspace, values[i]);
}
}
void
rb_gc_update_values(long n, VALUE *values)
{
gc_update_values(rb_gc_get_objspace(), n, values);
}
static enum rb_id_table_iterator_result
check_id_table_move(VALUE value, void *data)
{
void *objspace = (void *)data;
if (rb_gc_impl_object_moved_p(objspace, (VALUE)value)) {
return ID_TABLE_REPLACE;
}
return ID_TABLE_CONTINUE;
}
void
rb_gc_prepare_heap_process_object(VALUE obj)
{
switch (BUILTIN_TYPE(obj)) {
case T_STRING:
// Precompute the string coderange. This both save time for when it will be
// eventually needed, and avoid mutating heap pages after a potential fork.
rb_enc_str_coderange(obj);
break;
default:
break;
}
}
void
rb_gc_prepare_heap(void)
{
rb_gc_impl_prepare_heap(rb_gc_get_objspace());
}
size_t
rb_gc_heap_id_for_size(size_t size)
{
return rb_gc_impl_heap_id_for_size(rb_gc_get_objspace(), size);
}
bool
rb_gc_size_allocatable_p(size_t size)
{
return rb_gc_impl_size_allocatable_p(size);
}
static enum rb_id_table_iterator_result
update_id_table(VALUE *value, void *data, int existing)
{
void *objspace = (void *)data;
if (rb_gc_impl_object_moved_p(objspace, (VALUE)*value)) {
*value = gc_location_internal(objspace, (VALUE)*value);
}
return ID_TABLE_CONTINUE;
}
static void
update_m_tbl(void *objspace, struct rb_id_table *tbl)
{
if (tbl) {
rb_id_table_foreach_values_with_replace(tbl, check_id_table_move, update_id_table, objspace);
}
}
static enum rb_id_table_iterator_result
update_cc_tbl_i(VALUE ccs_ptr, void *objspace)
{
struct rb_class_cc_entries *ccs = (struct rb_class_cc_entries *)ccs_ptr;
VM_ASSERT(vm_ccs_p(ccs));
if (rb_gc_impl_object_moved_p(objspace, (VALUE)ccs->cme)) {
ccs->cme = (const rb_callable_method_entry_t *)gc_location_internal(objspace, (VALUE)ccs->cme);
}
for (int i=0; i<ccs->len; i++) {
if (rb_gc_impl_object_moved_p(objspace, (VALUE)ccs->entries[i].cc)) {
ccs->entries[i].cc = (struct rb_callcache *)gc_location_internal(objspace, (VALUE)ccs->entries[i].cc);
}
}
// do not replace
return ID_TABLE_CONTINUE;
}
static void
2025-04-30 13:48:02 +09:00
update_cc_tbl(void *objspace, struct rb_id_table *tbl)
{
2025-04-30 13:48:02 +09:00
if (!tbl) return;
rb_id_table_foreach_values(tbl, update_cc_tbl_i, objspace);
}
static enum rb_id_table_iterator_result
update_cvc_tbl_i(VALUE cvc_entry, void *objspace)
{
struct rb_cvar_class_tbl_entry *entry;
entry = (struct rb_cvar_class_tbl_entry *)cvc_entry;
if (entry->cref) {
TYPED_UPDATE_IF_MOVED(objspace, rb_cref_t *, entry->cref);
}
entry->class_value = gc_location_internal(objspace, entry->class_value);
return ID_TABLE_CONTINUE;
}
static void
2025-04-30 13:48:02 +09:00
update_cvc_tbl(void *objspace, struct rb_id_table *tbl)
{
2025-04-30 13:48:02 +09:00
if (!tbl) return;
rb_id_table_foreach_values(tbl, update_cvc_tbl_i, objspace);
}
static enum rb_id_table_iterator_result
2025-04-30 13:48:02 +09:00
update_const_tbl_i(VALUE value, void *objspace)
{
rb_const_entry_t *ce = (rb_const_entry_t *)value;
if (rb_gc_impl_object_moved_p(objspace, ce->value)) {
ce->value = gc_location_internal(objspace, ce->value);
}
if (rb_gc_impl_object_moved_p(objspace, ce->file)) {
ce->file = gc_location_internal(objspace, ce->file);
}
return ID_TABLE_CONTINUE;
}
static void
update_const_tbl(void *objspace, struct rb_id_table *tbl)
{
if (!tbl) return;
2025-04-30 13:48:02 +09:00
rb_id_table_foreach_values(tbl, update_const_tbl_i, objspace);
}
static void
2025-04-30 13:48:02 +09:00
update_subclasses(void *objspace, rb_classext_t *ext)
{
2025-04-30 13:48:02 +09:00
rb_subclass_entry_t *entry;
rb_subclass_anchor_t *anchor = RCLASSEXT_SUBCLASSES(ext);
if (!anchor) return;
entry = anchor->head;
while (entry) {
2025-04-30 13:48:02 +09:00
if (entry->klass)
UPDATE_IF_MOVED(objspace, entry->klass);
entry = entry->next;
}
}
static void
2025-04-30 13:48:02 +09:00
update_superclasses(rb_objspace_t *objspace, rb_classext_t *ext)
{
if (RCLASSEXT_SUPERCLASSES_WITH_SELF(ext)) {
size_t array_size = RCLASSEXT_SUPERCLASS_DEPTH(ext) + 1;
2025-04-30 13:48:02 +09:00
for (size_t i = 0; i < array_size; i++) {
UPDATE_IF_MOVED(objspace, RCLASSEXT_SUPERCLASSES(ext)[i]);
}
}
}
static void
update_classext_values(rb_objspace_t *objspace, rb_classext_t *ext, bool is_iclass)
{
2025-04-30 13:48:02 +09:00
UPDATE_IF_MOVED(objspace, RCLASSEXT_ORIGIN(ext));
UPDATE_IF_MOVED(objspace, RCLASSEXT_REFINED_CLASS(ext));
UPDATE_IF_MOVED(objspace, RCLASSEXT_CLASSPATH(ext));
if (is_iclass) {
UPDATE_IF_MOVED(objspace, RCLASSEXT_INCLUDER(ext));
}
}
static void
2025-04-30 13:48:02 +09:00
update_classext(rb_classext_t *ext, bool is_prime, VALUE namespace, void *arg)
{
2025-04-30 13:48:02 +09:00
struct classext_foreach_args *args = (struct classext_foreach_args *)arg;
rb_objspace_t *objspace = args->objspace;
if (RCLASSEXT_SUPER(ext)) {
UPDATE_IF_MOVED(objspace, RCLASSEXT_SUPER(ext));
}
update_m_tbl(objspace, RCLASSEXT_M_TBL(ext));
UPDATE_IF_MOVED(objspace, ext->fields_obj);
2025-04-30 13:48:02 +09:00
if (!RCLASSEXT_SHARED_CONST_TBL(ext)) {
update_const_tbl(objspace, RCLASSEXT_CONST_TBL(ext));
}
update_cc_tbl(objspace, RCLASSEXT_CC_TBL(ext));
update_cvc_tbl(objspace, RCLASSEXT_CVC_TBL(ext));
update_superclasses(objspace, ext);
update_subclasses(objspace, ext);
update_classext_values(objspace, ext, false);
2025-04-30 13:48:02 +09:00
}
static void
update_iclass_classext(rb_classext_t *ext, bool is_prime, VALUE namespace, void *arg)
{
struct classext_foreach_args *args = (struct classext_foreach_args *)arg;
rb_objspace_t *objspace = args->objspace;
if (RCLASSEXT_SUPER(ext)) {
UPDATE_IF_MOVED(objspace, RCLASSEXT_SUPER(ext));
}
update_m_tbl(objspace, RCLASSEXT_M_TBL(ext));
update_m_tbl(objspace, RCLASSEXT_CALLABLE_M_TBL(ext));
update_cc_tbl(objspace, RCLASSEXT_CC_TBL(ext));
update_subclasses(objspace, ext);
update_classext_values(objspace, ext, true);
}
extern rb_symbols_t ruby_global_symbols;
#define global_symbols ruby_global_symbols
struct global_vm_table_foreach_data {
vm_table_foreach_callback_func callback;
vm_table_update_callback_func update_callback;
void *data;
bool weak_only;
};
static int
vm_weak_table_foreach_weak_key(st_data_t key, st_data_t value, st_data_t data, int error)
{
struct global_vm_table_foreach_data *iter_data = (struct global_vm_table_foreach_data *)data;
int ret = iter_data->callback((VALUE)key, iter_data->data);
if (!iter_data->weak_only) {
if (ret != ST_CONTINUE) return ret;
ret = iter_data->callback((VALUE)value, iter_data->data);
}
return ret;
}
static int
vm_weak_table_foreach_update_weak_key(st_data_t *key, st_data_t *value, st_data_t data, int existing)
{
struct global_vm_table_foreach_data *iter_data = (struct global_vm_table_foreach_data *)data;
int ret = iter_data->update_callback((VALUE *)key, iter_data->data);
if (!iter_data->weak_only) {
if (ret != ST_CONTINUE) return ret;
ret = iter_data->update_callback((VALUE *)value, iter_data->data);
}
return ret;
}
static int
vm_weak_table_cc_refinement_foreach(st_data_t key, st_data_t data, int error)
{
struct global_vm_table_foreach_data *iter_data = (struct global_vm_table_foreach_data *)data;
return iter_data->callback((VALUE)key, iter_data->data);
}
static int
vm_weak_table_cc_refinement_foreach_update_update(st_data_t *key, st_data_t data, int existing)
{
struct global_vm_table_foreach_data *iter_data = (struct global_vm_table_foreach_data *)data;
return iter_data->update_callback((VALUE *)key, iter_data->data);
}
static int
vm_weak_table_str_sym_foreach(st_data_t key, st_data_t value, st_data_t data, int error)
{
struct global_vm_table_foreach_data *iter_data = (struct global_vm_table_foreach_data *)data;
if (!iter_data->weak_only) {
int ret = iter_data->callback((VALUE)key, iter_data->data);
if (ret != ST_CONTINUE) return ret;
}
if (STATIC_SYM_P(value)) {
return ST_CONTINUE;
}
else {
return iter_data->callback((VALUE)value, iter_data->data);
}
}
static int
vm_weak_table_foreach_update_weak_value(st_data_t *key, st_data_t *value, st_data_t data, int existing)
{
struct global_vm_table_foreach_data *iter_data = (struct global_vm_table_foreach_data *)data;
if (!iter_data->weak_only) {
int ret = iter_data->update_callback((VALUE *)key, iter_data->data);
if (ret != ST_CONTINUE) return ret;
}
return iter_data->update_callback((VALUE *)value, iter_data->data);
}
static void
free_gen_fields_tbl(VALUE obj, struct gen_fields_tbl *fields_tbl)
{
if (UNLIKELY(rb_shape_obj_too_complex_p(obj))) {
st_free_table(fields_tbl->as.complex.table);
}
xfree(fields_tbl);
}
static int
vm_weak_table_gen_fields_foreach_too_complex_i(st_data_t _key, st_data_t value, st_data_t data, int error)
{
struct global_vm_table_foreach_data *iter_data = (struct global_vm_table_foreach_data *)data;
GC_ASSERT(!iter_data->weak_only);
if (SPECIAL_CONST_P((VALUE)value)) return ST_CONTINUE;
return iter_data->callback((VALUE)value, iter_data->data);
}
static int
vm_weak_table_gen_fields_foreach_too_complex_replace_i(st_data_t *_key, st_data_t *value, st_data_t data, int existing)
{
struct global_vm_table_foreach_data *iter_data = (struct global_vm_table_foreach_data *)data;
GC_ASSERT(!iter_data->weak_only);
return iter_data->update_callback((VALUE *)value, iter_data->data);
}
struct st_table *rb_generic_fields_tbl_get(void);
static int
vm_weak_table_id2ref_foreach(st_data_t key, st_data_t value, st_data_t data, int error)
{
struct global_vm_table_foreach_data *iter_data = (struct global_vm_table_foreach_data *)data;
if (!iter_data->weak_only && !FIXNUM_P((VALUE)key)) {
int ret = iter_data->callback((VALUE)key, iter_data->data);
if (ret != ST_CONTINUE) return ret;
}
return iter_data->callback((VALUE)value, iter_data->data);
}
static int
vm_weak_table_id2ref_foreach_update(st_data_t *key, st_data_t *value, st_data_t data, int existing)
{
struct global_vm_table_foreach_data *iter_data = (struct global_vm_table_foreach_data *)data;
iter_data->update_callback((VALUE *)value, iter_data->data);
if (!iter_data->weak_only && !FIXNUM_P((VALUE)*key)) {
iter_data->update_callback((VALUE *)key, iter_data->data);
}
return ST_CONTINUE;
}
static int
vm_weak_table_gen_fields_foreach(st_data_t key, st_data_t value, st_data_t data)
{
struct global_vm_table_foreach_data *iter_data = (struct global_vm_table_foreach_data *)data;
int ret = iter_data->callback((VALUE)key, iter_data->data);
switch (ret) {
case ST_CONTINUE:
break;
case ST_DELETE:
free_gen_fields_tbl((VALUE)key, (struct gen_fields_tbl *)value);
FL_UNSET((VALUE)key, FL_EXIVAR);
return ST_DELETE;
case ST_REPLACE: {
VALUE new_key = (VALUE)key;
ret = iter_data->update_callback(&new_key, iter_data->data);
if (key != new_key) ret = ST_DELETE;
DURING_GC_COULD_MALLOC_REGION_START();
{
st_insert(rb_generic_fields_tbl_get(), (st_data_t)new_key, value);
}
DURING_GC_COULD_MALLOC_REGION_END();
key = (st_data_t)new_key;
break;
}
default:
return ret;
}
if (!iter_data->weak_only) {
struct gen_fields_tbl *fields_tbl = (struct gen_fields_tbl *)value;
if (rb_shape_obj_too_complex_p((VALUE)key)) {
st_foreach_with_replace(
fields_tbl->as.complex.table,
vm_weak_table_gen_fields_foreach_too_complex_i,
vm_weak_table_gen_fields_foreach_too_complex_replace_i,
data
);
}
else {
uint32_t fields_count = RSHAPE_LEN(RBASIC_SHAPE_ID((VALUE)key));
for (uint32_t i = 0; i < fields_count; i++) {
if (SPECIAL_CONST_P(fields_tbl->as.shape.fields[i])) continue;
int ivar_ret = iter_data->callback(fields_tbl->as.shape.fields[i], iter_data->data);
switch (ivar_ret) {
case ST_CONTINUE:
break;
case ST_REPLACE:
iter_data->update_callback(&fields_tbl->as.shape.fields[i], iter_data->data);
break;
default:
rb_bug("vm_weak_table_gen_fields_foreach: return value %d not supported", ivar_ret);
}
}
}
}
return ret;
}
static int
vm_weak_table_frozen_strings_foreach(st_data_t key, st_data_t value, st_data_t data, int error)
{
int retval = vm_weak_table_foreach_weak_key(key, value, data, error);
if (retval == ST_DELETE) {
FL_UNSET((VALUE)key, RSTRING_FSTR);
}
return retval;
}
Lock-free hash set for fstrings [Feature #21268] This implements a hash set which is wait-free for lookup and lock-free for insert (unless resizing) to use for fstring de-duplication. As highlighted in https://bugs.ruby-lang.org/issues/19288, heavy use of fstrings (frozen interned strings) can significantly reduce the parallelism of Ractors. I tried a few other approaches first: using an RWLock, striping a series of RWlocks (partitioning the hash N-ways to reduce lock contention), and putting a cache in front of it. All of these improved the situation, but were unsatisfying as all still required locks for writes (and granular locks are awkward, since we run the risk of needing to reach a vm barrier) and this table is somewhat write-heavy. My main reference for this was Cliff Click's talk on a lock free hash-table for java https://www.youtube.com/watch?v=HJ-719EGIts. It turns out this lock-free hash set is made easier to implement by a few properties: * We only need a hash set rather than a hash table (we only need keys, not values), and so the full entry can be written as a single VALUE * As a set we only need lookup/insert/delete, no update * Delete is only run inside GC so does not need to be atomic (It could be made concurrent) * I use rb_vm_barrier for the (rare) table rebuilds (It could be made concurrent) We VM lock (but don't require other threads to stop) for table rebuilds, as those are rare * The conservative garbage collector makes deferred replication easy, using a T_DATA object Another benefits of having a table specific to fstrings is that we compare by value on lookup/insert, but by identity on delete, as we only want to remove the exact string which is being freed. This is faster and provides a second way to avoid the race condition in https://bugs.ruby-lang.org/issues/21172. This is a pretty standard open-addressing hash table with quadratic probing. Similar to our existing st_table or id_table. Deletes (which happen on GC) replace existing keys with a tombstone, which is the only type of update which can occur. Tombstones are only cleared out on resize. Unlike st_table, the VALUEs are stored in the hash table itself (st_table's bins) rather than as a compact index. This avoids an extra pointer dereference and is possible because we don't need to preserve insertion order. The table targets a load factor of 2 (it is enlarged once it is half full).
2025-03-06 16:16:35 -08:00
void rb_fstring_foreach_with_replace(st_foreach_check_callback_func *func, st_update_callback_func *replace, st_data_t arg);
void
rb_gc_vm_weak_table_foreach(vm_table_foreach_callback_func callback,
vm_table_update_callback_func update_callback,
void *data,
bool weak_only,
enum rb_gc_vm_weak_tables table)
{
rb_vm_t *vm = GET_VM();
struct global_vm_table_foreach_data foreach_data = {
.callback = callback,
.update_callback = update_callback,
.data = data,
.weak_only = weak_only,
};
switch (table) {
case RB_GC_VM_CI_TABLE: {
if (vm->ci_table) {
st_foreach_with_replace(
vm->ci_table,
vm_weak_table_foreach_weak_key,
vm_weak_table_foreach_update_weak_key,
(st_data_t)&foreach_data
);
}
break;
}
case RB_GC_VM_OVERLOADED_CME_TABLE: {
if (vm->overloaded_cme_table) {
st_foreach_with_replace(
vm->overloaded_cme_table,
vm_weak_table_foreach_weak_key,
vm_weak_table_foreach_update_weak_key,
(st_data_t)&foreach_data
);
}
break;
}
case RB_GC_VM_GLOBAL_SYMBOLS_TABLE: {
if (global_symbols.str_sym) {
st_foreach_with_replace(
global_symbols.str_sym,
vm_weak_table_str_sym_foreach,
vm_weak_table_foreach_update_weak_value,
(st_data_t)&foreach_data
);
}
break;
}
case RB_GC_VM_ID2REF_TABLE: {
if (id2ref_tbl) {
st_foreach_with_replace(
id2ref_tbl,
vm_weak_table_id2ref_foreach,
vm_weak_table_id2ref_foreach_update,
(st_data_t)&foreach_data
);
}
break;
}
case RB_GC_VM_GENERIC_FIELDS_TABLE: {
st_table *generic_fields_tbl = rb_generic_fields_tbl_get();
if (generic_fields_tbl) {
st_foreach(
generic_fields_tbl,
vm_weak_table_gen_fields_foreach,
(st_data_t)&foreach_data
);
}
break;
}
case RB_GC_VM_FROZEN_STRINGS_TABLE: {
Lock-free hash set for fstrings [Feature #21268] This implements a hash set which is wait-free for lookup and lock-free for insert (unless resizing) to use for fstring de-duplication. As highlighted in https://bugs.ruby-lang.org/issues/19288, heavy use of fstrings (frozen interned strings) can significantly reduce the parallelism of Ractors. I tried a few other approaches first: using an RWLock, striping a series of RWlocks (partitioning the hash N-ways to reduce lock contention), and putting a cache in front of it. All of these improved the situation, but were unsatisfying as all still required locks for writes (and granular locks are awkward, since we run the risk of needing to reach a vm barrier) and this table is somewhat write-heavy. My main reference for this was Cliff Click's talk on a lock free hash-table for java https://www.youtube.com/watch?v=HJ-719EGIts. It turns out this lock-free hash set is made easier to implement by a few properties: * We only need a hash set rather than a hash table (we only need keys, not values), and so the full entry can be written as a single VALUE * As a set we only need lookup/insert/delete, no update * Delete is only run inside GC so does not need to be atomic (It could be made concurrent) * I use rb_vm_barrier for the (rare) table rebuilds (It could be made concurrent) We VM lock (but don't require other threads to stop) for table rebuilds, as those are rare * The conservative garbage collector makes deferred replication easy, using a T_DATA object Another benefits of having a table specific to fstrings is that we compare by value on lookup/insert, but by identity on delete, as we only want to remove the exact string which is being freed. This is faster and provides a second way to avoid the race condition in https://bugs.ruby-lang.org/issues/21172. This is a pretty standard open-addressing hash table with quadratic probing. Similar to our existing st_table or id_table. Deletes (which happen on GC) replace existing keys with a tombstone, which is the only type of update which can occur. Tombstones are only cleared out on resize. Unlike st_table, the VALUEs are stored in the hash table itself (st_table's bins) rather than as a compact index. This avoids an extra pointer dereference and is possible because we don't need to preserve insertion order. The table targets a load factor of 2 (it is enlarged once it is half full).
2025-03-06 16:16:35 -08:00
rb_fstring_foreach_with_replace(
vm_weak_table_frozen_strings_foreach,
vm_weak_table_foreach_update_weak_key,
(st_data_t)&foreach_data
);
break;
}
case RB_GC_VM_CC_REFINEMENT_TABLE: {
if (vm->cc_refinement_table) {
set_foreach_with_replace(
vm->cc_refinement_table,
vm_weak_table_cc_refinement_foreach,
vm_weak_table_cc_refinement_foreach_update_update,
(st_data_t)&foreach_data
);
}
break;
}
case RB_GC_VM_WEAK_TABLE_COUNT:
rb_bug("Unreacheable");
default:
rb_bug("rb_gc_vm_weak_table_foreach: unknown table %d", table);
}
}
void
rb_gc_update_vm_references(void *objspace)
{
rb_execution_context_t *ec = GET_EC();
rb_vm_t *vm = rb_ec_vm_ptr(ec);
rb_vm_update_references(vm);
rb_gc_update_global_tbl();
rb_sym_global_symbols_update_references();
#if USE_YJIT
void rb_yjit_root_update_references(void); // in Rust
if (rb_yjit_enabled_p) {
rb_yjit_root_update_references();
}
#endif
}
void
rb_gc_update_object_references(void *objspace, VALUE obj)
{
2025-04-30 13:48:02 +09:00
struct classext_foreach_args args;
switch (BUILTIN_TYPE(obj)) {
case T_CLASS:
if (FL_TEST_RAW(obj, FL_SINGLETON)) {
UPDATE_IF_MOVED(objspace, RCLASS_ATTACHED_OBJECT(obj));
}
// Continue to the shared T_CLASS/T_MODULE
case T_MODULE:
2025-04-30 13:48:02 +09:00
args.klass = obj;
args.objspace = objspace;
rb_class_classext_foreach(obj, update_classext, (void *)&args);
break;
case T_ICLASS:
2025-04-30 13:48:02 +09:00
args.objspace = objspace;
rb_class_classext_foreach(obj, update_iclass_classext, (void *)&args);
2020-04-08 15:13:37 +09:00
break;
case T_IMEMO:
rb_imemo_mark_and_move(obj, true);
return;
case T_NIL:
case T_FIXNUM:
case T_NODE:
case T_MOVED:
case T_NONE:
/* These can't move */
return;
case T_ARRAY:
gc_ref_update_array(objspace, obj);
break;
case T_HASH:
gc_ref_update_hash(objspace, obj);
UPDATE_IF_MOVED(objspace, RHASH(obj)->ifnone);
break;
case T_STRING:
{
if (STR_SHARED_P(obj)) {
UPDATE_IF_MOVED(objspace, RSTRING(obj)->as.heap.aux.shared);
}
/* If, after move the string is not embedded, and can fit in the
* slot it's been placed in, then re-embed it. */
if (rb_gc_obj_slot_size(obj) >= rb_str_size_as_embedded(obj)) {
if (!STR_EMBED_P(obj) && rb_str_reembeddable_p(obj)) {
rb_str_make_embedded(obj);
}
}
break;
}
case T_DATA:
/* Call the compaction callback, if it exists */
{
void *const ptr = RTYPEDDATA_P(obj) ? RTYPEDDATA_GET_DATA(obj) : DATA_PTR(obj);
if (ptr) {
Save one VALUE per embedded RTypedData This halves the amount of memory used for embedded RTypedData if they are one VALUE (8 bytes on 64-bit platforms) over the slot size limit. For Set, on 64-bit it uses an embedded 56-byte struct. With the previous implementation, the embedded structs starts at offset 32, resulting in a total size of 88. Since that is over the 80 byte limit, it goes to the next highest bucket, 160 bytes, wasting 72 bytes. This allows it to fit in a 80 byte bucket, which reduces the total size for small sets of from 224 bytes (160 bytes embedded, 64 bytes malloc, 72 bytes wasted in embedding) to 144 bytes (80 bytes embedded, 64 bytes malloc, 0 bytes wasted in embedding). Any other embedded RTypedData will see similar advantages if they are currently one VALUE over the limit. To implement this, remove the typed_flag from struct RTypedData. Embed the typed_flag information in the type member, which is now a tagged pointer using VALUE type, using the bottom low 2 bits as flags (1 bit for typed flag, the other for the embedded flag). To get the actual pointer, RTYPEDDATA_TYPE masks out the low 2 bits and then casts. That moves the RTypedData data pointer from offset 32 to offset 24 (on 64-bit). Vast amount of code in the internals (and probably external C extensions) expects the following code to work for both RData and non-embedded RTypedData: ```c DATA_PTR(obj) = some_pointer; ``` Allow this to work by moving the data pointer in RData between the dmark and dfree pointers, so it is at the same offset (24 on 64-bit). Other than these changes to the include files, the only changes needed were to gc.c, to account for the new struct layouts, handle setting the low bits in the type member, and to use RTYPEDDATA_TYPE(obj) instead of RTYPEDDATA(obj)->type.
2025-04-27 17:23:04 -07:00
if (RTYPEDDATA_P(obj) && gc_declarative_marking_p(RTYPEDDATA_TYPE(obj))) {
size_t *offset_list = TYPED_DATA_REFS_OFFSET_LIST(obj);
for (size_t offset = *offset_list; offset != RUBY_REF_END; offset = *offset_list++) {
VALUE *ref = (VALUE *)((char *)ptr + offset);
*ref = gc_location_internal(objspace, *ref);
}
}
else if (RTYPEDDATA_P(obj)) {
Save one VALUE per embedded RTypedData This halves the amount of memory used for embedded RTypedData if they are one VALUE (8 bytes on 64-bit platforms) over the slot size limit. For Set, on 64-bit it uses an embedded 56-byte struct. With the previous implementation, the embedded structs starts at offset 32, resulting in a total size of 88. Since that is over the 80 byte limit, it goes to the next highest bucket, 160 bytes, wasting 72 bytes. This allows it to fit in a 80 byte bucket, which reduces the total size for small sets of from 224 bytes (160 bytes embedded, 64 bytes malloc, 72 bytes wasted in embedding) to 144 bytes (80 bytes embedded, 64 bytes malloc, 0 bytes wasted in embedding). Any other embedded RTypedData will see similar advantages if they are currently one VALUE over the limit. To implement this, remove the typed_flag from struct RTypedData. Embed the typed_flag information in the type member, which is now a tagged pointer using VALUE type, using the bottom low 2 bits as flags (1 bit for typed flag, the other for the embedded flag). To get the actual pointer, RTYPEDDATA_TYPE masks out the low 2 bits and then casts. That moves the RTypedData data pointer from offset 32 to offset 24 (on 64-bit). Vast amount of code in the internals (and probably external C extensions) expects the following code to work for both RData and non-embedded RTypedData: ```c DATA_PTR(obj) = some_pointer; ``` Allow this to work by moving the data pointer in RData between the dmark and dfree pointers, so it is at the same offset (24 on 64-bit). Other than these changes to the include files, the only changes needed were to gc.c, to account for the new struct layouts, handle setting the low bits in the type member, and to use RTYPEDDATA_TYPE(obj) instead of RTYPEDDATA(obj)->type.
2025-04-27 17:23:04 -07:00
RUBY_DATA_FUNC compact_func = RTYPEDDATA_TYPE(obj)->function.dcompact;
if (compact_func) (*compact_func)(ptr);
}
}
}
break;
case T_OBJECT:
gc_ref_update_object(objspace, obj);
break;
case T_FILE:
if (RFILE(obj)->fptr) {
UPDATE_IF_MOVED(objspace, RFILE(obj)->fptr->self);
UPDATE_IF_MOVED(objspace, RFILE(obj)->fptr->pathv);
UPDATE_IF_MOVED(objspace, RFILE(obj)->fptr->tied_io_for_writing);
UPDATE_IF_MOVED(objspace, RFILE(obj)->fptr->writeconv_asciicompat);
UPDATE_IF_MOVED(objspace, RFILE(obj)->fptr->writeconv_pre_ecopts);
UPDATE_IF_MOVED(objspace, RFILE(obj)->fptr->encs.ecopts);
UPDATE_IF_MOVED(objspace, RFILE(obj)->fptr->write_lock);
UPDATE_IF_MOVED(objspace, RFILE(obj)->fptr->timeout);
UPDATE_IF_MOVED(objspace, RFILE(obj)->fptr->wakeup_mutex);
}
break;
case T_REGEXP:
UPDATE_IF_MOVED(objspace, RREGEXP(obj)->src);
break;
case T_SYMBOL:
UPDATE_IF_MOVED(objspace, RSYMBOL(obj)->fstr);
break;
case T_FLOAT:
case T_BIGNUM:
break;
case T_MATCH:
UPDATE_IF_MOVED(objspace, RMATCH(obj)->regexp);
if (RMATCH(obj)->str) {
UPDATE_IF_MOVED(objspace, RMATCH(obj)->str);
}
break;
case T_RATIONAL:
UPDATE_IF_MOVED(objspace, RRATIONAL(obj)->num);
UPDATE_IF_MOVED(objspace, RRATIONAL(obj)->den);
break;
case T_COMPLEX:
UPDATE_IF_MOVED(objspace, RCOMPLEX(obj)->real);
UPDATE_IF_MOVED(objspace, RCOMPLEX(obj)->imag);
break;
case T_STRUCT:
{
long i, len = RSTRUCT_LEN(obj);
VALUE *ptr = (VALUE *)RSTRUCT_CONST_PTR(obj);
for (i = 0; i < len; i++) {
UPDATE_IF_MOVED(objspace, ptr[i]);
}
}
break;
default:
rb_bug("unreachable");
break;
}
UPDATE_IF_MOVED(objspace, RBASIC(obj)->klass);
}
VALUE
rb_gc_start(void)
{
rb_gc();
return Qnil;
}
void
rb_gc(void)
{
unless_objspace(objspace) { return; }
rb_gc_impl_start(objspace, true, true, true, false);
}
int
rb_during_gc(void)
{
unless_objspace(objspace) { return FALSE; }
return rb_gc_impl_during_gc_p(objspace);
}
size_t
rb_gc_count(void)
{
return rb_gc_impl_gc_count(rb_gc_get_objspace());
}
static VALUE
gc_count(rb_execution_context_t *ec, VALUE self)
{
return SIZET2NUM(rb_gc_count());
}
VALUE
rb_gc_latest_gc_info(VALUE key)
{
if (!SYMBOL_P(key) && !RB_TYPE_P(key, T_HASH)) {
rb_raise(rb_eTypeError, "non-hash or symbol given");
}
VALUE val = rb_gc_impl_latest_gc_info(rb_gc_get_objspace(), key);
if (val == Qundef) {
rb_raise(rb_eArgError, "unknown key: %"PRIsVALUE, rb_sym2str(key));
}
return val;
}
static VALUE
gc_stat(rb_execution_context_t *ec, VALUE self, VALUE arg) // arg is (nil || hash || symbol)
{
if (NIL_P(arg)) {
arg = rb_hash_new();
}
else if (!RB_TYPE_P(arg, T_HASH) && !SYMBOL_P(arg)) {
rb_raise(rb_eTypeError, "non-hash or symbol given");
}
VALUE ret = rb_gc_impl_stat(rb_gc_get_objspace(), arg);
if (ret == Qundef) {
GC_ASSERT(SYMBOL_P(arg));
2022-07-21 09:23:58 -07:00
rb_raise(rb_eArgError, "unknown key: %"PRIsVALUE, rb_sym2str(arg));
}
return ret;
}
size_t
rb_gc_stat(VALUE arg)
{
if (!RB_TYPE_P(arg, T_HASH) && !SYMBOL_P(arg)) {
rb_raise(rb_eTypeError, "non-hash or symbol given");
}
VALUE ret = rb_gc_impl_stat(rb_gc_get_objspace(), arg);
if (ret == Qundef) {
GC_ASSERT(SYMBOL_P(arg));
rb_raise(rb_eArgError, "unknown key: %"PRIsVALUE, rb_sym2str(arg));
}
if (SYMBOL_P(arg)) {
return NUM2SIZET(ret);
}
else {
return 0;
}
}
static VALUE
gc_stat_heap(rb_execution_context_t *ec, VALUE self, VALUE heap_name, VALUE arg)
{
if (NIL_P(arg)) {
arg = rb_hash_new();
}
if (NIL_P(heap_name)) {
if (!RB_TYPE_P(arg, T_HASH)) {
rb_raise(rb_eTypeError, "non-hash given");
}
}
else if (FIXNUM_P(heap_name)) {
if (!SYMBOL_P(arg) && !RB_TYPE_P(arg, T_HASH)) {
rb_raise(rb_eTypeError, "non-hash or symbol given");
}
}
else {
rb_raise(rb_eTypeError, "heap_name must be nil or an Integer");
}
VALUE ret = rb_gc_impl_stat_heap(rb_gc_get_objspace(), heap_name, arg);
if (ret == Qundef) {
GC_ASSERT(SYMBOL_P(arg));
rb_raise(rb_eArgError, "unknown key: %"PRIsVALUE, rb_sym2str(arg));
}
return ret;
}
Provide GC.config to disable major GC collections This feature provides a new method `GC.config` that configures internal GC configuration variables provided by an individual GC implementation. Implemented in this PR is the option `full_mark`: a boolean value that will determine whether the Ruby GC is allowed to run a major collection while the process is running. It has the following semantics This feature configures Ruby's GC to only run minor GC's. It's designed to give users relying on Out of Band GC complete control over when a major GC is run. Configuring `full_mark: false` does two main things: * Never runs a Major GC. When the heap runs out of space during a minor and when a major would traditionally be run, instead we allocate more heap pages, and mark objspace as needing a major GC. * Don't increment object ages. We don't promote objects during GC, this will cause every object to be scanned on every minor. This is an intentional trade-off between minor GC's doing more work every time, and potentially promoting objects that will then never be GC'd. The intention behind not aging objects is that users of this feature should use a preforking web server, or some other method of pre-warming the oldgen (like Nakayoshi fork)before disabling Majors. That way most objects that are going to be old will have already been promoted. This will interleave major and minor GC collections in exactly the same what that the Ruby GC runs in versions previously to this. This is the default behaviour. * This new method has the following extra semantics: - `GC.config` with no arguments returns a hash of the keys of the currently configured GC - `GC.config` with a key pair (eg. `GC.config(full_mark: true)` sets the matching config key to the corresponding value and returns the entire known config hash, including the new values. If the key does not exist, `nil` is returned * When a minor GC is run, Ruby sets an internal status flag to determine whether the next GC will be a major or a minor. When `full_mark: false` this flag is ignored and every GC will be a minor. This status flag can be accessed at `GC.latest_gc_info(:needs_major_by)`. Any value other than `nil` means that the next collection would have been a major. Thus it's possible to use this feature to check at a predetermined time, whether a major GC is necessary and run one if it is. eg. After a request has finished processing. ```ruby if GC.latest_gc_info(:needs_major_by) GC.start(full_mark: true) end ``` [Feature #20443]
2024-07-04 15:21:09 +01:00
static VALUE
gc_config_get(rb_execution_context_t *ec, VALUE self)
{
VALUE cfg_hash = rb_gc_impl_config_get(rb_gc_get_objspace());
rb_hash_aset(cfg_hash, sym("implementation"), rb_fstring_cstr(rb_gc_impl_active_gc_name()));
return cfg_hash;
Provide GC.config to disable major GC collections This feature provides a new method `GC.config` that configures internal GC configuration variables provided by an individual GC implementation. Implemented in this PR is the option `full_mark`: a boolean value that will determine whether the Ruby GC is allowed to run a major collection while the process is running. It has the following semantics This feature configures Ruby's GC to only run minor GC's. It's designed to give users relying on Out of Band GC complete control over when a major GC is run. Configuring `full_mark: false` does two main things: * Never runs a Major GC. When the heap runs out of space during a minor and when a major would traditionally be run, instead we allocate more heap pages, and mark objspace as needing a major GC. * Don't increment object ages. We don't promote objects during GC, this will cause every object to be scanned on every minor. This is an intentional trade-off between minor GC's doing more work every time, and potentially promoting objects that will then never be GC'd. The intention behind not aging objects is that users of this feature should use a preforking web server, or some other method of pre-warming the oldgen (like Nakayoshi fork)before disabling Majors. That way most objects that are going to be old will have already been promoted. This will interleave major and minor GC collections in exactly the same what that the Ruby GC runs in versions previously to this. This is the default behaviour. * This new method has the following extra semantics: - `GC.config` with no arguments returns a hash of the keys of the currently configured GC - `GC.config` with a key pair (eg. `GC.config(full_mark: true)` sets the matching config key to the corresponding value and returns the entire known config hash, including the new values. If the key does not exist, `nil` is returned * When a minor GC is run, Ruby sets an internal status flag to determine whether the next GC will be a major or a minor. When `full_mark: false` this flag is ignored and every GC will be a minor. This status flag can be accessed at `GC.latest_gc_info(:needs_major_by)`. Any value other than `nil` means that the next collection would have been a major. Thus it's possible to use this feature to check at a predetermined time, whether a major GC is necessary and run one if it is. eg. After a request has finished processing. ```ruby if GC.latest_gc_info(:needs_major_by) GC.start(full_mark: true) end ``` [Feature #20443]
2024-07-04 15:21:09 +01:00
}
static VALUE
gc_config_set(rb_execution_context_t *ec, VALUE self, VALUE hash)
{
void *objspace = rb_gc_get_objspace();
rb_gc_impl_config_set(objspace, hash);
return rb_gc_impl_config_get(objspace);
Provide GC.config to disable major GC collections This feature provides a new method `GC.config` that configures internal GC configuration variables provided by an individual GC implementation. Implemented in this PR is the option `full_mark`: a boolean value that will determine whether the Ruby GC is allowed to run a major collection while the process is running. It has the following semantics This feature configures Ruby's GC to only run minor GC's. It's designed to give users relying on Out of Band GC complete control over when a major GC is run. Configuring `full_mark: false` does two main things: * Never runs a Major GC. When the heap runs out of space during a minor and when a major would traditionally be run, instead we allocate more heap pages, and mark objspace as needing a major GC. * Don't increment object ages. We don't promote objects during GC, this will cause every object to be scanned on every minor. This is an intentional trade-off between minor GC's doing more work every time, and potentially promoting objects that will then never be GC'd. The intention behind not aging objects is that users of this feature should use a preforking web server, or some other method of pre-warming the oldgen (like Nakayoshi fork)before disabling Majors. That way most objects that are going to be old will have already been promoted. This will interleave major and minor GC collections in exactly the same what that the Ruby GC runs in versions previously to this. This is the default behaviour. * This new method has the following extra semantics: - `GC.config` with no arguments returns a hash of the keys of the currently configured GC - `GC.config` with a key pair (eg. `GC.config(full_mark: true)` sets the matching config key to the corresponding value and returns the entire known config hash, including the new values. If the key does not exist, `nil` is returned * When a minor GC is run, Ruby sets an internal status flag to determine whether the next GC will be a major or a minor. When `full_mark: false` this flag is ignored and every GC will be a minor. This status flag can be accessed at `GC.latest_gc_info(:needs_major_by)`. Any value other than `nil` means that the next collection would have been a major. Thus it's possible to use this feature to check at a predetermined time, whether a major GC is necessary and run one if it is. eg. After a request has finished processing. ```ruby if GC.latest_gc_info(:needs_major_by) GC.start(full_mark: true) end ``` [Feature #20443]
2024-07-04 15:21:09 +01:00
}
static VALUE
gc_stress_get(rb_execution_context_t *ec, VALUE self)
{
return rb_gc_impl_stress_get(rb_gc_get_objspace());
}
static VALUE
gc_stress_set_m(rb_execution_context_t *ec, VALUE self, VALUE flag)
{
rb_gc_impl_stress_set(rb_gc_get_objspace(), flag);
return flag;
}
void
rb_gc_initial_stress_set(VALUE flag)
{
2024-09-18 14:54:52 -04:00
initial_stress = flag;
}
size_t *
rb_gc_heap_sizes(void)
{
return rb_gc_impl_heap_sizes(rb_gc_get_objspace());
}
VALUE
rb_gc_enable(void)
{
return rb_objspace_gc_enable(rb_gc_get_objspace());
}
VALUE
rb_objspace_gc_enable(void *objspace)
{
bool disabled = !rb_gc_impl_gc_enabled_p(objspace);
rb_gc_impl_gc_enable(objspace);
return RBOOL(disabled);
}
static VALUE
gc_enable(rb_execution_context_t *ec, VALUE _)
{
return rb_gc_enable();
}
static VALUE
gc_disable_no_rest(void *objspace)
{
bool disabled = !rb_gc_impl_gc_enabled_p(objspace);
rb_gc_impl_gc_disable(objspace, false);
return RBOOL(disabled);
}
VALUE
rb_gc_disable_no_rest(void)
{
return gc_disable_no_rest(rb_gc_get_objspace());
}
VALUE
rb_gc_disable(void)
{
return rb_objspace_gc_disable(rb_gc_get_objspace());
}
VALUE
rb_objspace_gc_disable(void *objspace)
{
bool disabled = !rb_gc_impl_gc_enabled_p(objspace);
rb_gc_impl_gc_disable(objspace, true);
return RBOOL(disabled);
}
static VALUE
gc_disable(rb_execution_context_t *ec, VALUE _)
{
return rb_gc_disable();
}
// TODO: think about moving ruby_gc_set_params into Init_heap or Init_gc
void
ruby_gc_set_params(void)
{
rb_gc_impl_set_params(rb_gc_get_objspace());
}
void
rb_objspace_reachable_objects_from(VALUE obj, void (func)(VALUE, void *), void *data)
{
2025-05-23 20:14:20 +09:00
RB_VM_LOCKING() {
if (rb_gc_impl_during_gc_p(rb_gc_get_objspace())) rb_bug("rb_objspace_reachable_objects_from() is not supported while during GC");
if (!RB_SPECIAL_CONST_P(obj)) {
rb_vm_t *vm = GET_VM();
struct gc_mark_func_data_struct *prev_mfd = vm->gc.mark_func_data;
struct gc_mark_func_data_struct mfd = {
.mark_func = func,
.data = data,
};
vm->gc.mark_func_data = &mfd;
rb_gc_mark_children(rb_gc_get_objspace(), obj);
vm->gc.mark_func_data = prev_mfd;
}
}
}
struct root_objects_data {
const char *category;
void (*func)(const char *category, VALUE, void *);
void *data;
};
static void
root_objects_from(VALUE obj, void *ptr)
{
const struct root_objects_data *data = (struct root_objects_data *)ptr;
(*data->func)(data->category, obj, data->data);
}
void
rb_objspace_reachable_objects_from_root(void (func)(const char *category, VALUE, void *), void *passing_data)
{
if (rb_gc_impl_during_gc_p(rb_gc_get_objspace())) rb_bug("rb_gc_impl_objspace_reachable_objects_from_root() is not supported while during GC");
rb_vm_t *vm = GET_VM();
struct root_objects_data data = {
.func = func,
.data = passing_data,
};
struct gc_mark_func_data_struct *prev_mfd = vm->gc.mark_func_data;
struct gc_mark_func_data_struct mfd = {
.mark_func = root_objects_from,
.data = &data,
};
vm->gc.mark_func_data = &mfd;
rb_gc_save_machine_context();
rb_gc_mark_roots(vm->gc.objspace, &data.category);
vm->gc.mark_func_data = prev_mfd;
}
/*
------------------------------ DEBUG ------------------------------
*/
static const char *
type_name(int type, VALUE obj)
{
switch (type) {
#define TYPE_NAME(t) case (t): return #t;
TYPE_NAME(T_NONE);
TYPE_NAME(T_OBJECT);
TYPE_NAME(T_CLASS);
TYPE_NAME(T_MODULE);
TYPE_NAME(T_FLOAT);
TYPE_NAME(T_STRING);
TYPE_NAME(T_REGEXP);
TYPE_NAME(T_ARRAY);
TYPE_NAME(T_HASH);
TYPE_NAME(T_STRUCT);
TYPE_NAME(T_BIGNUM);
TYPE_NAME(T_FILE);
TYPE_NAME(T_MATCH);
TYPE_NAME(T_COMPLEX);
TYPE_NAME(T_RATIONAL);
TYPE_NAME(T_NIL);
TYPE_NAME(T_TRUE);
TYPE_NAME(T_FALSE);
TYPE_NAME(T_SYMBOL);
TYPE_NAME(T_FIXNUM);
TYPE_NAME(T_UNDEF);
TYPE_NAME(T_IMEMO);
TYPE_NAME(T_ICLASS);
TYPE_NAME(T_MOVED);
TYPE_NAME(T_ZOMBIE);
case T_DATA:
if (obj && rb_objspace_data_type_name(obj)) {
return rb_objspace_data_type_name(obj);
}
return "T_DATA";
#undef TYPE_NAME
}
return "unknown";
}
static const char *
obj_type_name(VALUE obj)
{
return type_name(TYPE(obj), obj);
}
const char *
rb_method_type_name(rb_method_type_t type)
{
switch (type) {
case VM_METHOD_TYPE_ISEQ: return "iseq";
case VM_METHOD_TYPE_ATTRSET: return "attrest";
case VM_METHOD_TYPE_IVAR: return "ivar";
case VM_METHOD_TYPE_BMETHOD: return "bmethod";
case VM_METHOD_TYPE_ALIAS: return "alias";
case VM_METHOD_TYPE_REFINED: return "refined";
case VM_METHOD_TYPE_CFUNC: return "cfunc";
case VM_METHOD_TYPE_ZSUPER: return "zsuper";
case VM_METHOD_TYPE_MISSING: return "missing";
case VM_METHOD_TYPE_OPTIMIZED: return "optimized";
case VM_METHOD_TYPE_UNDEF: return "undef";
case VM_METHOD_TYPE_NOTIMPLEMENTED: return "notimplemented";
}
rb_bug("rb_method_type_name: unreachable (type: %d)", type);
}
static void
2022-07-05 22:39:08 +09:00
rb_raw_iseq_info(char *const buff, const size_t buff_size, const rb_iseq_t *iseq)
{
if (buff_size > 0 && ISEQ_BODY(iseq) && ISEQ_BODY(iseq)->location.label && !RB_TYPE_P(ISEQ_BODY(iseq)->location.pathobj, T_MOVED)) {
VALUE path = rb_iseq_path(iseq);
int n = ISEQ_BODY(iseq)->location.first_lineno;
2019-05-20 22:08:27 +09:00
snprintf(buff, buff_size, " %s@%s:%d",
RSTRING_PTR(ISEQ_BODY(iseq)->location.label),
RSTRING_PTR(path), n);
}
}
static int
str_len_no_raise(VALUE str)
{
long len = RSTRING_LEN(str);
if (len < 0) return 0;
if (len > INT_MAX) return INT_MAX;
return (int)len;
}
#define BUFF_ARGS buff + pos, buff_size - pos
#define APPEND_F(...) if ((pos += snprintf(BUFF_ARGS, "" __VA_ARGS__)) >= buff_size) goto end
2022-07-06 00:59:23 +09:00
#define APPEND_S(s) do { \
if ((pos + (int)rb_strlen_lit(s)) >= buff_size) { \
goto end; \
} \
else { \
memcpy(buff + pos, (s), rb_strlen_lit(s) + 1); \
} \
} while (0)
2022-07-06 00:23:49 +09:00
#define C(c, s) ((c) != 0 ? (s) : " ")
static size_t
rb_raw_obj_info_common(char *const buff, const size_t buff_size, const VALUE obj)
{
size_t pos = 0;
2022-07-06 00:59:23 +09:00
if (SPECIAL_CONST_P(obj)) {
APPEND_F("%s", obj_type_name(obj));
if (FIXNUM_P(obj)) {
APPEND_F(" %ld", FIX2LONG(obj));
}
else if (SYMBOL_P(obj)) {
APPEND_F(" %s", rb_id2name(SYM2ID(obj)));
}
}
else {
// const int age = RVALUE_AGE_GET(obj);
if (rb_gc_impl_pointer_to_heap_p(rb_gc_get_objspace(), (void *)obj)) {
APPEND_F("%p %s/", (void *)obj, obj_type_name(obj));
// TODO: fixme
// APPEND_F("%p [%d%s%s%s%s%s%s] %s ",
// (void *)obj, age,
// C(RVALUE_UNCOLLECTIBLE_BITMAP(obj), "L"),
// C(RVALUE_MARK_BITMAP(obj), "M"),
// C(RVALUE_PIN_BITMAP(obj), "P"),
// C(RVALUE_MARKING_BITMAP(obj), "R"),
// C(RVALUE_WB_UNPROTECTED_BITMAP(obj), "U"),
// C(rb_objspace_garbage_object_p(obj), "G"),
// obj_type_name(obj));
}
else {
/* fake */
// APPEND_F("%p [%dXXXX] %s",
// (void *)obj, age,
// obj_type_name(obj));
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
if (internal_object_p(obj)) {
/* ignore */
}
else if (RBASIC(obj)->klass == 0) {
2022-07-06 00:59:23 +09:00
APPEND_S("(temporary internal)");
}
2022-06-02 14:34:48 +09:00
else if (RTEST(RBASIC(obj)->klass)) {
2019-05-22 16:54:47 +09:00
VALUE class_path = rb_class_path_cached(RBASIC(obj)->klass);
if (!NIL_P(class_path)) {
APPEND_F("%s ", RSTRING_PTR(class_path));
}
}
2022-07-06 00:23:49 +09:00
}
end:
return pos;
}
const char *rb_raw_obj_info(char *const buff, const size_t buff_size, VALUE obj);
2022-07-06 00:23:49 +09:00
static size_t
rb_raw_obj_info_buitin_type(char *const buff, const size_t buff_size, const VALUE obj, size_t pos)
{
if (LIKELY(pos < buff_size) && !SPECIAL_CONST_P(obj)) {
const enum ruby_value_type type = BUILTIN_TYPE(obj);
switch (type) {
case T_NODE:
UNEXPECTED_NODE(rb_raw_obj_info);
break;
case T_ARRAY:
if (ARY_SHARED_P(obj)) {
2022-07-06 00:59:23 +09:00
APPEND_S("shared -> ");
rb_raw_obj_info(BUFF_ARGS, ARY_SHARED_ROOT(obj));
}
else if (ARY_EMBED_P(obj)) {
APPEND_F("[%s%s] len: %ld (embed)",
C(ARY_EMBED_P(obj), "E"),
C(ARY_SHARED_P(obj), "S"),
RARRAY_LEN(obj));
}
else {
2023-06-13 16:19:33 -04:00
APPEND_F("[%s%s] len: %ld, capa:%ld ptr:%p",
C(ARY_EMBED_P(obj), "E"),
C(ARY_SHARED_P(obj), "S"),
RARRAY_LEN(obj),
ARY_EMBED_P(obj) ? -1L : RARRAY(obj)->as.heap.aux.capa,
(void *)RARRAY_CONST_PTR(obj));
}
break;
case T_STRING: {
if (STR_SHARED_P(obj)) {
APPEND_F(" [shared] len: %ld", RSTRING_LEN(obj));
}
else {
2022-07-06 00:59:23 +09:00
if (STR_EMBED_P(obj)) APPEND_S(" [embed]");
APPEND_F(" len: %ld, capa: %" PRIdSIZE, RSTRING_LEN(obj), rb_str_capacity(obj));
}
APPEND_F(" \"%.*s\"", str_len_no_raise(obj), RSTRING_PTR(obj));
break;
}
case T_SYMBOL: {
2022-07-06 00:45:06 +09:00
VALUE fstr = RSYMBOL(obj)->fstr;
ID id = RSYMBOL(obj)->id;
if (RB_TYPE_P(fstr, T_STRING)) {
APPEND_F(":%s id:%d", RSTRING_PTR(fstr), (unsigned int)id);
2022-07-06 00:45:06 +09:00
}
else {
APPEND_F("(%p) id:%d", (void *)fstr, (unsigned int)id);
2022-07-06 00:45:06 +09:00
}
break;
}
case T_MOVED: {
APPEND_F("-> %p", (void*)gc_location_internal(rb_gc_get_objspace(), obj));
break;
}
case T_HASH: {
2023-04-19 15:59:25 -04:00
APPEND_F("[%c] %"PRIdSIZE,
2022-07-06 00:45:06 +09:00
RHASH_AR_TABLE_P(obj) ? 'A' : 'S',
RHASH_SIZE(obj));
2022-07-06 00:45:06 +09:00
break;
}
case T_CLASS:
case T_MODULE:
{
VALUE class_path = rb_class_path_cached(obj);
if (!NIL_P(class_path)) {
APPEND_F("%s", RSTRING_PTR(class_path));
}
else {
APPEND_S("(anon)");
}
break;
}
case T_ICLASS:
{
VALUE class_path = rb_class_path_cached(RBASIC_CLASS(obj));
if (!NIL_P(class_path)) {
APPEND_F("src:%s", RSTRING_PTR(class_path));
}
break;
}
case T_OBJECT:
{
if (rb_shape_obj_too_complex_p(obj)) {
size_t hash_len = rb_st_table_size(ROBJECT_FIELDS_HASH(obj));
APPEND_F("(too_complex) len:%zu", hash_len);
}
else {
uint32_t len = ROBJECT_FIELDS_CAPACITY(obj);
if (RBASIC(obj)->flags & ROBJECT_EMBED) {
APPEND_F("(embed) len:%d", len);
}
else {
VALUE *ptr = ROBJECT_FIELDS(obj);
APPEND_F("len:%d ptr:%p", len, (void *)ptr);
}
}
}
break;
case T_DATA: {
const struct rb_block *block;
const rb_iseq_t *iseq;
if (rb_obj_is_proc(obj) &&
(block = vm_proc_block(obj)) != NULL &&
(vm_block_type(block) == block_type_iseq) &&
(iseq = vm_block_iseq(block)) != NULL) {
2019-05-20 22:08:27 +09:00
rb_raw_iseq_info(BUFF_ARGS, iseq);
}
else if (rb_ractor_p(obj)) {
rb_ractor_t *r = (void *)DATA_PTR(obj);
if (r) {
APPEND_F("r:%d", r->pub.id);
}
}
else {
const char * const type_name = rb_objspace_data_type_name(obj);
if (type_name) {
APPEND_F("%s", type_name);
}
}
break;
}
case T_IMEMO: {
APPEND_F("<%s> ", rb_imemo_name(imemo_type(obj)));
switch (imemo_type(obj)) {
case imemo_ment:
{
const rb_method_entry_t *me = (const rb_method_entry_t *)obj;
APPEND_F(":%s (%s%s%s%s) type:%s aliased:%d owner:%p defined_class:%p",
rb_id2name(me->called_id),
METHOD_ENTRY_VISI(me) == METHOD_VISI_PUBLIC ? "pub" :
METHOD_ENTRY_VISI(me) == METHOD_VISI_PRIVATE ? "pri" : "pro",
METHOD_ENTRY_COMPLEMENTED(me) ? ",cmp" : "",
METHOD_ENTRY_CACHED(me) ? ",cc" : "",
METHOD_ENTRY_INVALIDATED(me) ? ",inv" : "",
me->def ? rb_method_type_name(me->def->type) : "NULL",
me->def ? me->def->aliased : -1,
(void *)me->owner, // obj_info(me->owner),
(void *)me->defined_class); //obj_info(me->defined_class)));
if (me->def) {
switch (me->def->type) {
case VM_METHOD_TYPE_ISEQ:
2022-07-06 00:59:23 +09:00
APPEND_S(" (iseq:");
2022-07-06 00:57:59 +09:00
rb_raw_obj_info(BUFF_ARGS, (VALUE)me->def->body.iseq.iseqptr);
2022-07-06 00:59:23 +09:00
APPEND_S(")");
break;
default:
break;
}
}
break;
}
case imemo_iseq: {
const rb_iseq_t *iseq = (const rb_iseq_t *)obj;
2019-05-20 22:08:27 +09:00
rb_raw_iseq_info(BUFF_ARGS, iseq);
break;
}
case imemo_callinfo:
{
const struct rb_callinfo *ci = (const struct rb_callinfo *)obj;
APPEND_F("(mid:%s, flag:%x argc:%d, kwarg:%s)",
rb_id2name(vm_ci_mid(ci)),
vm_ci_flag(ci),
vm_ci_argc(ci),
vm_ci_kwarg(ci) ? "available" : "NULL");
break;
}
case imemo_callcache:
{
const struct rb_callcache *cc = (const struct rb_callcache *)obj;
VALUE class_path = cc->klass ? rb_class_path_cached(cc->klass) : Qnil;
2021-01-06 13:05:30 +09:00
const rb_callable_method_entry_t *cme = vm_cc_cme(cc);
APPEND_F("(klass:%s cme:%s%s (%p) call:%p",
2021-01-06 13:05:30 +09:00
NIL_P(class_path) ? (cc->klass ? "??" : "<NULL>") : RSTRING_PTR(class_path),
cme ? rb_id2name(cme->called_id) : "<NULL>",
cme ? (METHOD_ENTRY_INVALIDATED(cme) ? " [inv]" : "") : "",
(void *)cme,
(void *)(uintptr_t)vm_cc_call(cc));
break;
}
default:
break;
}
}
default:
break;
}
}
end:
2022-07-06 00:23:49 +09:00
return pos;
}
#undef C
void
2024-12-18 12:01:09 -05:00
rb_asan_poison_object(VALUE obj)
{
MAYBE_UNUSED(struct RVALUE *) ptr = (void *)obj;
asan_poison_memory_region(ptr, rb_gc_obj_slot_size(obj));
}
void
2024-12-18 12:01:09 -05:00
rb_asan_unpoison_object(VALUE obj, bool newobj_p)
{
MAYBE_UNUSED(struct RVALUE *) ptr = (void *)obj;
asan_unpoison_memory_region(ptr, rb_gc_obj_slot_size(obj), newobj_p);
}
void *
2024-12-18 12:01:09 -05:00
rb_asan_poisoned_object_p(VALUE obj)
{
MAYBE_UNUSED(struct RVALUE *) ptr = (void *)obj;
return __asan_region_is_poisoned(ptr, rb_gc_obj_slot_size(obj));
}
static void
raw_obj_info(char *const buff, const size_t buff_size, VALUE obj)
{
size_t pos = rb_raw_obj_info_common(buff, buff_size, obj);
pos = rb_raw_obj_info_buitin_type(buff, buff_size, obj, pos);
if (pos >= buff_size) {} // truncated
}
2022-07-06 00:23:49 +09:00
const char *
rb_raw_obj_info(char *const buff, const size_t buff_size, VALUE obj)
{
void *objspace = rb_gc_get_objspace();
2020-06-10 10:17:56 -04:00
if (SPECIAL_CONST_P(obj)) {
raw_obj_info(buff, buff_size, obj);
}
else if (!rb_gc_impl_pointer_to_heap_p(objspace, (const void *)obj)) {
snprintf(buff, buff_size, "out-of-heap:%p", (void *)obj);
}
#if 0 // maybe no need to check it?
else if (0 && rb_gc_impl_garbage_object_p(objspace, obj)) {
snprintf(buff, buff_size, "garbage:%p", (void *)obj);
}
#endif
else {
asan_unpoisoning_object(obj) {
raw_obj_info(buff, buff_size, obj);
}
}
return buff;
}
#undef APPEND_S
#undef APPEND_F
#undef BUFF_ARGS
#if RGENGC_OBJ_INFO
#define OBJ_INFO_BUFFERS_NUM 10
#define OBJ_INFO_BUFFERS_SIZE 0x100
static rb_atomic_t obj_info_buffers_index = 0;
static char obj_info_buffers[OBJ_INFO_BUFFERS_NUM][OBJ_INFO_BUFFERS_SIZE];
/* Increments *var atomically and resets *var to 0 when maxval is
* reached. Returns the wraparound old *var value (0...maxval). */
static rb_atomic_t
atomic_inc_wraparound(rb_atomic_t *var, const rb_atomic_t maxval)
{
rb_atomic_t oldval = RUBY_ATOMIC_FETCH_ADD(*var, 1);
if (RB_UNLIKELY(oldval >= maxval - 1)) { // wraparound *var
const rb_atomic_t newval = oldval + 1;
RUBY_ATOMIC_CAS(*var, newval, newval % maxval);
oldval %= maxval;
}
return oldval;
}
static const char *
obj_info(VALUE obj)
{
rb_atomic_t index = atomic_inc_wraparound(&obj_info_buffers_index, OBJ_INFO_BUFFERS_NUM);
char *const buff = obj_info_buffers[index];
return rb_raw_obj_info(buff, OBJ_INFO_BUFFERS_SIZE, obj);
}
#else
static const char *
obj_info(VALUE obj)
{
return obj_type_name(obj);
}
#endif
/*
------------------------ Extended allocator ------------------------
*/
struct gc_raise_tag {
VALUE exc;
const char *fmt;
va_list *ap;
};
static void *
gc_vraise(void *ptr)
{
struct gc_raise_tag *argv = ptr;
rb_vraise(argv->exc, argv->fmt, *argv->ap);
UNREACHABLE_RETURN(NULL);
}
static void
gc_raise(VALUE exc, const char *fmt, ...)
{
va_list ap;
va_start(ap, fmt);
struct gc_raise_tag argv = {
exc, fmt, &ap,
};
if (ruby_thread_has_gvl_p()) {
gc_vraise(&argv);
UNREACHABLE;
}
else if (ruby_native_thread_p()) {
rb_thread_call_with_gvl(gc_vraise, &argv);
UNREACHABLE;
}
else {
/* Not in a ruby thread */
fprintf(stderr, "%s", "[FATAL] ");
vfprintf(stderr, fmt, ap);
}
va_end(ap);
abort();
}
NORETURN(static void negative_size_allocation_error(const char *));
static void
negative_size_allocation_error(const char *msg)
{
gc_raise(rb_eNoMemError, "%s", msg);
}
static void *
ruby_memerror_body(void *dummy)
{
rb_memerror();
return 0;
}
NORETURN(static void ruby_memerror(void));
RBIMPL_ATTR_MAYBE_UNUSED()
static void
ruby_memerror(void)
{
if (ruby_thread_has_gvl_p()) {
rb_memerror();
}
else {
if (ruby_native_thread_p()) {
rb_thread_call_with_gvl(ruby_memerror_body, 0);
}
else {
/* no ruby thread */
fprintf(stderr, "[FATAL] failed to allocate memory\n");
}
}
/* We have discussions whether we should die here; */
/* We might rethink about it later. */
exit(EXIT_FAILURE);
}
void
rb_memerror(void)
{
/* the `GET_VM()->special_exceptions` below assumes that
* the VM is reachable from the current thread. We should
* definitely make sure of that. */
RUBY_ASSERT_ALWAYS(ruby_thread_has_gvl_p());
rb_execution_context_t *ec = GET_EC();
VALUE exc = GET_VM()->special_exceptions[ruby_error_nomemory];
if (!exc ||
rb_ec_raised_p(ec, RAISED_NOMEMORY) ||
rb_ec_vm_lock_rec(ec) != ec->tag->lock_rec) {
fprintf(stderr, "[FATAL] failed to allocate memory\n");
exit(EXIT_FAILURE);
}
if (rb_ec_raised_p(ec, RAISED_NOMEMORY)) {
rb_ec_raised_clear(ec);
}
else {
rb_ec_raised_set(ec, RAISED_NOMEMORY);
exc = ruby_vm_special_exception_copy(exc);
}
ec->errinfo = exc;
EC_JUMP_TAG(ec, TAG_RAISE);
}
bool
rb_memerror_reentered(void)
{
rb_execution_context_t *ec = GET_EC();
return (ec && rb_ec_raised_p(ec, RAISED_NOMEMORY));
}
static void *
handle_malloc_failure(void *ptr)
{
if (LIKELY(ptr)) {
return ptr;
}
else {
ruby_memerror();
UNREACHABLE_RETURN(ptr);
}
}
static void *ruby_xmalloc_body(size_t size);
void *
ruby_xmalloc(size_t size)
{
return handle_malloc_failure(ruby_xmalloc_body(size));
}
static void *
ruby_xmalloc_body(size_t size)
{
if ((ssize_t)size < 0) {
negative_size_allocation_error("too large allocation size");
}
return rb_gc_impl_malloc(rb_gc_get_objspace(), size);
2022-07-06 00:23:49 +09:00
}
2022-07-06 00:59:23 +09:00
void
ruby_malloc_size_overflow(size_t count, size_t elsize)
{
rb_raise(rb_eArgError,
"malloc: possible integer overflow (%"PRIuSIZE"*%"PRIuSIZE")",
count, elsize);
}
void
ruby_malloc_add_size_overflow(size_t x, size_t y)
{
rb_raise(rb_eArgError,
"malloc: possible integer overflow (%"PRIuSIZE"+%"PRIuSIZE")",
x, y);
}
static void *ruby_xmalloc2_body(size_t n, size_t size);
void *
ruby_xmalloc2(size_t n, size_t size)
{
return handle_malloc_failure(ruby_xmalloc2_body(n, size));
}
static void *
ruby_xmalloc2_body(size_t n, size_t size)
{
return rb_gc_impl_malloc(rb_gc_get_objspace(), xmalloc2_size(n, size));
}
static void *ruby_xcalloc_body(size_t n, size_t size);
void *
ruby_xcalloc(size_t n, size_t size)
{
return handle_malloc_failure(ruby_xcalloc_body(n, size));
}
static void *
ruby_xcalloc_body(size_t n, size_t size)
{
return rb_gc_impl_calloc(rb_gc_get_objspace(), xmalloc2_size(n, size));
}
static void *ruby_sized_xrealloc_body(void *ptr, size_t new_size, size_t old_size);
#ifdef ruby_sized_xrealloc
#undef ruby_sized_xrealloc
#endif
void *
ruby_sized_xrealloc(void *ptr, size_t new_size, size_t old_size)
{
return handle_malloc_failure(ruby_sized_xrealloc_body(ptr, new_size, old_size));
}
static void *
ruby_sized_xrealloc_body(void *ptr, size_t new_size, size_t old_size)
{
if ((ssize_t)new_size < 0) {
negative_size_allocation_error("too large allocation size");
}
return rb_gc_impl_realloc(rb_gc_get_objspace(), ptr, new_size, old_size);
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
}
void *
ruby_xrealloc(void *ptr, size_t new_size)
{
return ruby_sized_xrealloc(ptr, new_size, 0);
}
static void *ruby_sized_xrealloc2_body(void *ptr, size_t n, size_t size, size_t old_n);
#ifdef ruby_sized_xrealloc2
#undef ruby_sized_xrealloc2
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
#endif
void *
ruby_sized_xrealloc2(void *ptr, size_t n, size_t size, size_t old_n)
{
return handle_malloc_failure(ruby_sized_xrealloc2_body(ptr, n, size, old_n));
}
static void *
ruby_sized_xrealloc2_body(void *ptr, size_t n, size_t size, size_t old_n)
{
size_t len = xmalloc2_size(n, size);
return rb_gc_impl_realloc(rb_gc_get_objspace(), ptr, len, old_n * size);
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
void *
ruby_xrealloc2(void *ptr, size_t n, size_t size)
{
return ruby_sized_xrealloc2(ptr, n, size, 0);
}
#ifdef ruby_sized_xfree
#undef ruby_sized_xfree
#endif
void
ruby_sized_xfree(void *x, size_t size)
{
if (LIKELY(x)) {
/* It's possible for a C extension's pthread destructor function set by pthread_key_create
* to be called after ruby_vm_destruct and attempt to free memory. Fall back to mimfree in
* that case. */
if (LIKELY(GET_VM())) {
rb_gc_impl_free(rb_gc_get_objspace(), x, size);
}
else {
ruby_mimfree(x);
}
}
}
2023-03-06 21:34:31 -08:00
void
ruby_xfree(void *x)
{
ruby_sized_xfree(x, 0);
}
void *
rb_xmalloc_mul_add(size_t x, size_t y, size_t z) /* x * y + z */
{
size_t w = size_mul_add_or_raise(x, y, z, rb_eArgError);
return ruby_xmalloc(w);
}
void *
rb_xcalloc_mul_add(size_t x, size_t y, size_t z) /* x * y + z */
{
size_t w = size_mul_add_or_raise(x, y, z, rb_eArgError);
return ruby_xcalloc(w, 1);
}
void *
rb_xrealloc_mul_add(const void *p, size_t x, size_t y, size_t z) /* x * y + z */
{
size_t w = size_mul_add_or_raise(x, y, z, rb_eArgError);
return ruby_xrealloc((void *)p, w);
}
void *
rb_xmalloc_mul_add_mul(size_t x, size_t y, size_t z, size_t w) /* x * y + z * w */
{
size_t u = size_mul_add_mul_or_raise(x, y, z, w, rb_eArgError);
return ruby_xmalloc(u);
}
void *
rb_xcalloc_mul_add_mul(size_t x, size_t y, size_t z, size_t w) /* x * y + z * w */
{
size_t u = size_mul_add_mul_or_raise(x, y, z, w, rb_eArgError);
return ruby_xcalloc(u, 1);
}
/* Mimic ruby_xmalloc, but need not rb_objspace.
* should return pointer suitable for ruby_xfree
*/
void *
ruby_mimmalloc(size_t size)
{
void *mem;
#if CALC_EXACT_MALLOC_SIZE
size += sizeof(struct malloc_obj_info);
#endif
mem = malloc(size);
#if CALC_EXACT_MALLOC_SIZE
if (!mem) {
return NULL;
}
else
/* set 0 for consistency of allocated_size/allocations */
{
struct malloc_obj_info *info = mem;
info->size = 0;
mem = info + 1;
}
#endif
return mem;
}
void *
ruby_mimcalloc(size_t num, size_t size)
{
void *mem;
#if CALC_EXACT_MALLOC_SIZE
struct rbimpl_size_mul_overflow_tag t = rbimpl_size_mul_overflow(num, size);
if (UNLIKELY(t.left)) {
return NULL;
}
size = t.right + sizeof(struct malloc_obj_info);
mem = calloc1(size);
if (!mem) {
return NULL;
}
else
/* set 0 for consistency of allocated_size/allocations */
{
struct malloc_obj_info *info = mem;
info->size = 0;
mem = info + 1;
}
#else
mem = calloc(num, size);
#endif
return mem;
}
void
ruby_mimfree(void *ptr)
{
#if CALC_EXACT_MALLOC_SIZE
struct malloc_obj_info *info = (struct malloc_obj_info *)ptr - 1;
ptr = info;
#endif
free(ptr);
}
* gc.c: add incremental GC algorithm. [Feature #10137] Please refer this ticket for details. This change also introduces the following changes. * Remove RGENGC_AGE2_PROMOTION and introduce object age (0 to 3). Age can be count with FL_PROMOTE0 and FL_PROMOTE1 flags in RBasic::flags (2 bit). Age == 3 objects become old objects. * WB_PROTECTED flag in RBasic to WB_UNPROTECTED bitmap. * LONG_LIVED bitmap to represent living objects while minor GCs It specifies (1) Old objects and (2) remembered shady objects. * Introduce rb_objspace_t::marked_objects which counts marked objects in current marking phase. marking count is needed to introduce incremental marking. * rename mark related function and sweep related function to gc_(marks|sweep)_(start|finish|step|rest|continue). * rename rgengc_report() to gc_report(). * Add obj_info() function to get cstr of object details. * Add MEASURE_LINE() macro to measure execution time of specific line. * and many small fixes. * include/ruby/ruby.h: add flag USE_RINCGC. Now USE_RINCGC can be set only with USE_RGENGC. * include/ruby/ruby.h: introduce FL_PROMOTED0 and add FL_PROMOTED1 to count object age. * include/ruby/ruby.h: rewrite write barriers for incremental marking. * debug.c: catch up flag name changes. * internal.h: add rb_gc_writebarrier_remember() instead of rb_gc_writebarrier_remember_promoted(). * array.c (ary_memcpy0): use rb_gc_writebarrier_remember(). * array.c (rb_ary_modify): ditto. * hash.c (rb_hash_keys): ditto. * hash.c (rb_hash_values): ditto. * object.c (init_copy): use rb_copy_wb_protected_attribute() because FL_WB_PROTECTED is moved from RBasic::flags. * test/objspace/test_objspace.rb: catch up ObjectSpace.dump() changes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2014-09-08 04:11:00 +00:00
void
rb_gc_adjust_memory_usage(ssize_t diff)
{
unless_objspace(objspace) { return; }
rb_gc_impl_adjust_memory_usage(objspace, diff);
}
const char *
rb_obj_info(VALUE obj)
{
return obj_info(obj);
}
void
rb_obj_info_dump(VALUE obj)
{
char buff[0x100];
fprintf(stderr, "rb_obj_info_dump: %s\n", rb_raw_obj_info(buff, 0x100, obj));
}
void
rb_obj_info_dump_loc(VALUE obj, const char *file, int line, const char *func)
{
char buff[0x100];
fprintf(stderr, "<OBJ_INFO:%s@%s:%d> %s\n", func, file, line, rb_raw_obj_info(buff, 0x100, obj));
}
void
rb_gc_before_fork(void)
{
rb_gc_impl_before_fork(rb_gc_get_objspace());
}
void
rb_gc_after_fork(rb_pid_t pid)
{
rb_gc_impl_after_fork(rb_gc_get_objspace(), pid);
}
/*
* Document-module: ObjectSpace
*
* The ObjectSpace module contains a number of routines
* that interact with the garbage collection facility and allow you to
* traverse all living objects with an iterator.
*
* ObjectSpace also provides support for object finalizers, procs that will be
* called after a specific object was destroyed by garbage collection. See
* the documentation for +ObjectSpace.define_finalizer+ for important
* information on how to use this method correctly.
*
* a = "A"
* b = "B"
*
* ObjectSpace.define_finalizer(a, proc {|id| puts "Finalizer one on #{id}" })
* ObjectSpace.define_finalizer(b, proc {|id| puts "Finalizer two on #{id}" })
*
* a = nil
* b = nil
*
* _produces:_
*
* Finalizer two on 537763470
* Finalizer one on 537763480
*/
/* Document-class: GC::Profiler
*
* The GC profiler provides access to information on GC runs including time,
* length and object space size.
*
* Example:
*
* GC::Profiler.enable
*
* require 'rdoc/rdoc'
*
* GC::Profiler.report
*
* GC::Profiler.disable
*
* See also GC.count, GC.malloc_allocated_size and GC.malloc_allocations
*/
#include "gc.rbinc"
void
Init_GC(void)
{
#undef rb_intern
rb_gc_register_address(&id2ref_value);
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 15:32:12 +02:00
malloc_offset = gc_compute_malloc_offset();
rb_mGC = rb_define_module("GC");
VALUE rb_mObjSpace = rb_define_module("ObjectSpace");
rb_define_module_function(rb_mObjSpace, "each_object", os_each_obj, -1);
rb_define_module_function(rb_mObjSpace, "define_finalizer", define_final, -1);
rb_define_module_function(rb_mObjSpace, "undefine_finalizer", undefine_final, 1);
rb_define_module_function(rb_mObjSpace, "_id2ref", os_id2ref, 1);
rb_vm_register_special_exception(ruby_error_nomemory, rb_eNoMemError, "failed to allocate memory");
rb_define_method(rb_cBasicObject, "__id__", rb_obj_id, 0);
rb_define_method(rb_mKernel, "object_id", rb_obj_id, 0);
rb_define_module_function(rb_mObjSpace, "count_objects", count_objects, -1);
rb_gc_impl_init();
}
// Set a name for the anonymous virtual memory area. `addr` is the starting
// address of the area and `size` is its length in bytes. `name` is a
// NUL-terminated human-readable string.
//
// This function is usually called after calling `mmap()`. The human-readable
// annotation helps developers identify the call site of `mmap()` that created
// the memory mapping.
//
// This function currently only works on Linux 5.17 or higher. After calling
// this function, we can see annotations in the form of "[anon:...]" in
// `/proc/self/maps`, where `...` is the content of `name`. This function has
// no effect when called on other platforms.
void
ruby_annotate_mmap(const void *addr, unsigned long size, const char *name)
{
#if defined(HAVE_SYS_PRCTL_H) && defined(PR_SET_VMA) && defined(PR_SET_VMA_ANON_NAME)
// The name length cannot exceed 80 (including the '\0').
RUBY_ASSERT(strlen(name) < 80);
prctl(PR_SET_VMA, PR_SET_VMA_ANON_NAME, (unsigned long)addr, size, name);
// We ignore errors in prctl. prctl may set errno to EINVAL for several
// reasons.
// 1. The attr (PR_SET_VMA_ANON_NAME) is not a valid attribute.
// 2. addr is an invalid address.
// 3. The string pointed by name is too long.
// The first error indicates PR_SET_VMA_ANON_NAME is not available, and may
// happen if we run the compiled binary on an old kernel. In theory, all
// other errors should result in a failure. But since EINVAL cannot tell
// the first error from others, and this function is mainly used for
// debugging, we silently ignore the error.
errno = 0;
#endif
}