so can skip oldgen bitmap check around mark & sweep.
* gc.c (slot_sweep_body): change scan algorithm for performance:
from object's pointer base to bitmap one.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@41128 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
"Sunny" doesn't mean antonym of "Shady" (questionable, doubtful, etc).
Instead of "Suuny", use "non-shady" or "normal".
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@41121 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* gc.c (rgengc_rememberset_mark): bypass gc_mark() in order to push
sunny old object at minor gc.
* gc.c (gc_mark_children): move sunny old check to gc_mark().
* gc.c (rgengc_check_shady): remove DEMOTE that already unnecessary.
* gc.c (rb_gc_writebarrier): ditto.
change sunny old check point in order to save mark stack and
remove unnatural rest_sweep & demote.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@41097 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Profiling record is allocated at first of marking phase.
Enable at lazy sweeping may cause an error (SEGV).
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40984 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
RUBY_INTERNAL_EVENT_GC_END. This event invokes at the end of
after_sweep().
Time chart with lazy sweep is here:
(1) Kick RUBY_INTERNAL_EVENT_GC_START
(2) [gc_marks()]
(3) [lazy_sweep()]
(4) [... run Ruby program (mutator) with lazy_sweep() ...]
(5) [after_sweep()]
(6) Kick RUBY_INTERNAL_EVENT_GC_END
(7) [... run Ruby program (mutator), and go to (1) ...]
* ext/-test-/tracepoint/tracepoint.c,
test/-ext-/tracepoint/test_tracepoint.rb: modify a test.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40963 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Read the following test to know HOWTO.
This feature is a sample of RUBY_INTERNAL_EVENT.
* test/objspace/test_objspace.rb: add a test.
* ext/objspace/object_tracing.c: ditto.
* gc.c (rb_gc_count): add. THis function returns GC count.
* internal.h: add decl. of rb_gc_count(). Same as `GC.count'.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40957 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* RUBY_INTERNAL_EVENT_NEWOBJ: object created.
* RUBY_INTERNAL_EVENT_FREE: object freeed.
* RUBY_INTERNAL_EVENT_GC_START: GC started.
And rename `RUBY_EVENT_SWITCH' to `RUBY_INTERNAL_EVENT_SWITCH'.
Internal events can not invoke any Ruby program because the tracing
timing may be critical (under huge restriction).
These events can be hooked only by C-extensions.
We recommend to use rb_potponed_job_register() API to call Ruby
program safely.
This change is mostly written by Aman Gupta (tmm1).
https://bugs.ruby-lang.org/issues/8107#note-12
[Feature #8107]
* include/ruby/debug.h, vm_trace.c: added two new APIs.
* rb_tracearg_event_flag() returns rb_event_flag_t of this event.
* rb_tracearg_object() returns created/freeed object.
* ext/-test-/tracepoint/extconf.rb,
ext/-test-/tracepoint/tracepoint.c,
test/-ext-/tracepoint/test_tracepoint.rb: add a test.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40946 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Postponed jobs are registered with this API. Registered jobs
are invoked at `ruby-running-safe-point' as soon as possible.
This timing is completely same as finalizer timing.
There are two APIs:
* rb_postponed_job_register(flags, func, data): register a
postponed job with data. flags are reserved.
* rb_postponed_job_register_one(flags, func, data): same as
`rb_postponed_job_register', but only one `func' job is
registered (skip if `func' is already registered).
This change is mostly written by Aman Gupta (tmm1).
https://bugs.ruby-lang.org/issues/8107#note-15
[Feature #8107]
* gc.c: use postponed job API for finalizer.
* common.mk: add dependency from vm_trace.c to debug.h.
* ext/-test-/postponed_job/extconf.rb, postponed_job.c,
test/-ext-/postponed_job/test_postponed_job.rb: add a test.
* thread.c: implement postponed API.
* vm_core.h: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40940 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
remembered_sunny_object_count and remembered_shady_object_count
for each types when RGENGC_PROFILE >= 2.
They are informative for optimization.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40938 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* number of oldgen object is biggger than twice of
number of oldgen object at last full GC.
* number of remembered shady object is bigger than twice of
number of remembered shady object at last full GC.
* number of oldgen object and remembered shady object is bigger
than half of total object space.
(please fix my English!)
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40914 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
before extending heaps.
TODO: do major GC when there are many old (promoted) objects.
* gc.c (after_gc_sweep): remove TODO comments.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40909 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* gc.c (gc_profile_dump_on): revert r40898. ok to show the record
accumulating while lazy_sweep().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40899 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* gc.c (gc_profile_dump_on): do not dump the last record, for some
reason.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40898 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* gc.c (gc_profile_dump_on): use size_t to get rid of overflow and
show the header when next_index > 0, instead of next_index != 1.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40889 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
(or garbage_collect_body()). `garbage_collect()' accept additional
two parameters `full_mark' and `immediate_sweep'.
If `full_mark' is TRUE, then force it full gc (major gc), otherwise,
it depends on status of object space. Now, it will be minor gc.
If `immediate_sweep' is TRUE, then disable lazy sweep.
To allocate free memory, `full_mark' and `immediate_sweep' should be
TRUE. Otherwise, they should be FALSE.
* gc.c (gc_prepare_free_objects): use `garbage_collect_body()'.
* gc.c (slot_sweep, before_gc_sweep, after_gc_sweep): add logging code.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40884 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
sweeping time. At lazy GC, GC::Profiler makes new record entry
for each lazy_sweep(). In this change, accumulating all
slot_sweep() time.
And change indentation.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40876 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
can be negative number.
And use pointer for `record' (don't copy).
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40870 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* gc.c (rb_gc_guarded_ptr): unoptimize on other compilers than gcc and
msvc.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40795 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
three parameters v1, v2, v3. newobj_of() do OBJSETUP() and
fill values with v1, v2, v3.
* gc.c (rb_data_object_alloc, rb_data_typed_object_alloc):
use newobj_of().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40764 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Profiling result can be check by GC.stat.
* gc.c (type_name): separate from obj_type_name().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40761 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
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
We can use `objspace_live_num()' instead of removed `marked_num'
if it is after `after_gc_sweep()' function call.
* gc.c (after_gc_sweep): use objspace_live_num() instead of removed
rb_objspace_t::marked_num.
* gc.c (gc_mark_ptr, gc_marks): remove rb_objspace_t::marked_num code.
* gc.c (gc_prepare_free_objects): do not call set_heaps_increment()
with checking objspace->heap.marked_num. At this point, we only
need to check availability of free-cell.
* gc.c (gc_prepare_free_objects):
* gc.c (lazy_sweep): call after_gc_sweep() if there are no sweep_able entry.
* gc.c (rest_sweep, gc_prepare_free_objects): remove after_gc_sweep() call.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40702 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* gc.c (gc_prof_set_malloc_info): fix "objspace->heap.live_num" to
"objspace_live_num(objspace)". There is no such member variable.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40701 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* gc.c (gc_prof_sweep_timer_start/stop): removed because
they doesn't support lazy sweep.
* gc.c (gc_prof_sweep_slot_timer_start/stop): added.
redefine `sweeping time' to accumulated time of all of
slot_sweep().
* gc.c (rb_objspace_t::profile::count): renamed to
rb_objspace_t::profile::next_index. `counter' seems ambiguous.
increment it when next record is acquired.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40700 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
RBASIC_CLASS(obj) macro which returns a class of `obj'.
This change is a part of RGENGC branch [ruby-trunk - Feature #8339].
* object.c: add new function rb_obj_reveal().
This function reveal interal (hidden) object by rb_obj_hide().
Note that do not change class before and after hiding.
Only permitted example is:
klass = RBASIC_CLASS(obj);
rb_obj_hide(obj);
....
rb_obj_reveal(obj, klass);
TODO: API design. rb_obj_reveal() should be replaced with others.
TODO: modify constified variables using cast may be harmful for
compiler's analysis and optimizaton.
Any idea to prohibt inserting RBasic::klass directly?
If rename RBasic::klass and force to use RBASIC_CLASS(obj),
then all codes such as `RBASIC(obj)->klass' will be
compilation error. Is it acceptable? (We have similar
experience at Ruby 1.9,
for example "RARRAY(ary)->ptr" to "RARRAY_PTR(ary)".
* internal.h: add some macros.
* RBASIC_CLEAR_CLASS(obj) clear RBasic::klass to make it internal
object.
* RBASIC_SET_CLASS(obj, cls) set RBasic::klass.
* RBASIC_SET_CLASS_RAW(obj, cls) same as RBASIC_SET_CLASS
without write barrier (planned).
* RCLASS_SET_SUPER(a, b) set super class of a.
* array.c, class.c, compile.c, encoding.c, enum.c, error.c, eval.c,
file.c, gc.c, hash.c, io.c, iseq.c, marshal.c, object.c,
parse.y, proc.c, process.c, random.c, ruby.c, sprintf.c,
string.c, thread.c, transcode.c, vm.c, vm_eval.c, win32/file.c:
Use above macros and functions to access RBasic::klass.
* ext/coverage/coverage.c, ext/readline/readline.c,
ext/socket/ancdata.c, ext/socket/init.c,
* ext/zlib/zlib.c: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40691 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* gc.c (gc_prepare_free_objects): fix compile error. there should be
free objects when new slot is added.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@39902 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
restricted. So collecting is often invoked if there is not
enough free space in the heap. Try to expand heap when this is
the case.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@39900 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
