* array.c (flatten): no need to call to_ary method on elements
beyond the given level. [ruby-core:67637] [Bug #10748]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49400 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_bsearch): trivial optimizations, for Fixnum, and
by keeping the last satisfied element.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49216 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_bsearch): preserve encoding of class name in an
exception message.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49215 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (ary_recycle_hash): add RB_GC_GUARD
(rb_ary_diff): remove volatile
[Bug #10369]
The RB_GC_GUARD seems needed for most of the ary_recycle_hash
functions anyways.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47896 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* internal.h (rb_ary_new_from_args): optimization by expanding
arguments in caller to get rid of va_list if possible.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47858 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (ary_new): allocate buffer in heap after new object get
allocated successfully, to get rid of potential memory leak at
object allocation failure.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47855 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
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
This function creates internal use only array (which is completely
hided by ObjectSpace.each_object) with filling nil.
Otherwise, it can be incldues strange VALUEs.
* internal.h: added.
* node.h: use rb_ary_tmp_new_fill() for MEMO.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47189 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_any_p): the array may be modified in the yielded
block, do not access directly.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@46868 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_any_p), hash.c (rb_hash_any_p): optimized
versions. these are bit faster than optimization in
Enumerable#any?.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@46866 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rcombinate0): remove recursion, by looping with indexes
stored in `p`.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@46428 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rpermute0): remove recursion, by looping with indexes
stored in `p`.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@46427 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_combination): iterate on a shared copy, and use
array of indexes instead of array of chosen objects.
[ruby-core:63149] [Bug #9939]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@46418 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_permutation): `p` is the array of size `r`, as
commented at permute0(). since `n >= r` here, buffer overflow
never happened, just reduce unnecessary allocation though.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@46416 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
ARY_SET() is same functionality of RARRAY_ASET(), but
it has an assertion (`ary' doesn't have shared array).
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@45569 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Shared arrays were WB-unprotected object because
sharing array can modify shared array's buffer
if it occupied shared array.
[sharing array (ary)] -> [shared array (shared)] -> <buff>
| A
+---------------------------------------+
write `buff' with WB(ary, &buff[i], obj)
-> if `ary' and `shared' are old, then only `ary'
will be remembered.
-> traverse from `ary'. But `shared' is old, so
that written `obj' is not marked.
It cause WB miss so that shared arrays were WB-unprotected.
(WB-unprotected objects are marked everytime if it is living)
This patch insert WB() for `shared' if it is needed.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@45568 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (ary_reject): may be turned into a shared array during
the given block. [ruby-dev:48101] [Bug #9727]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@45562 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
array owner is promoted and a shared array is not promoted.
Now, shared array is WB-unprotected so that shared arrays are not
promoted. All objects referred from shared array should be marked
correctly.
[ruby-core:61919] [ruby-trunk - Bug #9718]
* test/ruby/test_array.rb: add a test for above.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@45553 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
add links to `Object#hash` to each #`hash` methods rdocs.
[Fixes GH-567]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@45329 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_initialize): NUM2LONG() may call size.to_int,
ary can be frozen, modified, etc, so recheck after argument
conversion is necessary. [Closes GH-526]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@44855 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_shuffle_bang, rb_ary_sample): share id_random
instead of no longer used sym_random.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@44066 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c (rb_execarg_fixup): use RHASH_TBL_RAW and insert write
barriers where appropriate
* vm.c (kwmerge_i): use RHASH_TBL_RAW
* vm.c (HASH_ASET): use rb_hash_aset instead of calling directly into
st_insert
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@44052 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_or): lhs elements are prefered, so should not
replace with rhs elements.
* test/ruby/test_array.rb (test_OR_in_order): import the test failed
by r43969 from rubyspec/core/array/union_spec.rb.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@44015 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* hash.c (rb_hash): detect recursion for all `hash' methods. each
`hash' methods no longer need to use rb_exec_recursive().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@43980 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* include/ruby/ruby.h (RB_BLOCK_CALL_FUNC_ARGLIST): for declaration
argument list of rb_block_call_func.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@43910 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
if HAVE_MALLOC_USABLE_SIZE (or _WIN32) is defined.
We don't need these function if malloc_usable_size() is available.
* gc.c: catch up this change.
* gc.c: define HAVE_MALLOC_USABLE_SIZE on _WIN32.
* array.c (ary_resize_capa): do not use ruby_sized_xfree() with
local variable to avoid "unused local variable" warning.
This change only has few impact.
* string.c (rb_str_resize): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@43839 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c: reduce use of RARRAY_LEN and ARY_CAPA when object is not
modified.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@43291 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_uniq): use rb_hash_values(), as well as the case no
block is given.
* internal.h: define rb_hash_values() as internal API.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@43197 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_uniq): set class of the return value to the receiver
class. fix failure in TestArray#test_array_subclass.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@43196 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_uniq): remove no longer used local variables since
r43194.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@43195 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
RARRAY_RAWPTR(ary) returns (const VALUE *) type pointer and
usecase of this macro is not acquire raw pointer, but acquire
read-only pointer. So we rename to better name.
RSTRUCT_RAWPTR() is also renamed to RSTRUCT_CONST_PTR()
(I expect that nobody use it).
* array.c, compile.c, cont.c, enumerator.c, gc.c, proc.c, random.c,
string.c, struct.c, thread.c, vm_eval.c, vm_insnhelper.c:
catch up this change.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@43043 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
there are not new relations.
* enum.c (enum_sort_by): ditto.
* struct.c (setup_struct): use RARRAY_RAWPTR().
* vm_eval.c (yield_under): ditto.
* ext/pathname/pathname.c (path_entries): use RARRAY_AREF().
* ext/pathname/pathname.c (path_s_glob): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42688 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
make it normal object.
* array.c (rb_ary_modify): use RARRAY_PTR_USE() without WB because
there are not new relations.
* array.c (ary_ensure_room_for_unshift): use RARRAY_RAWPTR() because
there are not new relations.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42665 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
At least on my environments, I don't see any errors
with many trials. Please tell us if you find any GC bugs.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42192 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
rb_gc_unprotect_logging() which is general version
* include/ruby/ruby.h: add USE_RGENGC_LOGGING_WB_UNPROTECT
to enable.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42151 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_initialize): use ary_memfill().
* array.c (rb_ary_fill): ditto.
* array.c (rb_ary_slice_bang): use RARRAY_RAWPTR() because
this code creates no new references.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42105 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Making non-shady shared array causes SEGV (see rubyci).
It seems a bug around shared array.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42079 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_modify, ary_make_partial, rb_ary_splice,
rb_ary_replace, rb_ary_eql, rb_ary_compact_bang):
use RARRAY_RAWPTR() instead of RARRAY_PTR().
* array.c (rb_ary_shift): use RARRAY_PTR_USE() without WB because
there are not new relations.
* array.c (ary_ensure_room_for_unshift): ditto.
* array.c (rb_ary_sort_bang): ditto.
* array.c (rb_ary_delete_at): ditto.
* array.c (rb_ary_reverse_m): use RARRAY_RAWPTR() because
there are not new relations.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42072 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
because this operation creates a reference to Qnil.
* array.c (ary_make_shared, rb_ary_store, rb_ary_shift_m,
rb_ary_splice, rb_ary_resize, rb_ary_fill): use ary_mem_clear()
instead of rb_mem_clear().
* array.c (ary_make_shared): use RARRAY_RAWPTR() instead of RARRAY_PTR().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42070 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
with write barrier. If ary is promoted, use write barrier correctly.
* array.c (rb_ary_cat, rb_ary_unshift_m, rb_ary_dup,
rb_ary_sort_bang, rb_ary_replace, rb_ary_plus): use ary_memcpy().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42065 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
RARRAY_PTR. In this code, there are no "write" operation.
* array.c (rb_ary_equal): ditto.
* array.c (recursive_equal): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42063 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
because VALUE is not defined before including ruby/ruby.h.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42061 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
while iterating. Remove other pointer loop when arg is given.
* test/ruby/test_array.rb (test_count): add test for bug.
[ruby-core:56072] [Bug #8654]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42041 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
rb_hash_to_a() to avoid array creation with rb_assoc_new().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42040 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Use RARRAY_ASET() or RARRAY_PTR_USE() to modify Array objects.
* array.c, gc.c: catch up above changes.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@41537 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_uniq_bang): must not be modified once frozen even in
a callback method.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@41247 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (rb_ary_sort_bang): must not be modified once frozen even in
a callback method.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@41246 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* array.c (ary_shrink_capa): shrink the capacity so it fits just with
the length.
* array.c (ary_make_shared): release never used elements from frozen
array to be shared. [ruby-dev:47416] [Bug #8510]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@41227 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
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
* vm_eval.c (check_funcall_respond_to): preserve passed_block, which
is modified in vm_call0_body() via vm_call0(), and caused a bug of
rb_check_funcall() by false negative result of rb_block_given_p().
re-fix [ruby-core:53650] [Bug #8153].
[ruby-core:53653] [Bug #8154]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@39881 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
nobu
hsbt
ayumin
duerst
akr
normal
glass
ko1
zzak
marcandre
tmm1
ktsj
charliesome
kazu
eregon
shugo
tarui