Age | Commit message (Collapse) | Author |
|
|
|
It's an embedded TypedData, it can much more efficiently store
the references it need without using ivars.
Notes:
Merged: https://github.com/ruby/ruby/pull/13577
|
|
enumerator's meth attribute is an ID, which is not a Ruby object, so we
should not apply the write barrier on it.
Notes:
Merged: https://github.com/ruby/ruby/pull/11671
|
|
Enumerator#zip yields a single array, not multiple arguments,
so Enumerator::Lazy#zip should do the same.
Fixes [#20623]
Notes:
Merged: https://github.com/ruby/ruby/pull/11212
|
|
While working on a separate issue we found that in some cases
`ary_heap_realloc` was being called on frozen arrays. To fix this, this
change does the following:
1) Updates `rb_ary_freeze` to assert the type is an array, return if
already frozen, and shrink the capacity if it is not embedded, shared
or a shared root.
2) Replaces `rb_obj_freeze` with `rb_ary_freeze` when the object is
always an array.
3) In `ary_heap_realloc`, ensure the new capa is set with
`ARY_SET_CAPA`. Previously the change in capa was not set.
4) Adds an assertion to `ary_heap_realloc` that the array is not frozen.
Some of this work was originally done in
https://github.com/ruby/ruby/pull/2640, referencing this issue
https://bugs.ruby-lang.org/issues/16291. There didn't appear to be any
objections to this PR, it appears to have simply lost traction.
The original PR made changes to arrays and strings at the same time,
this PR only does arrays. Also it was old enough that rather than revive
that branch I've made a new one. I added Lourens as co-author in addtion
to Aaron who helped me with this patch.
The original PR made this change for performance reasons, and while
that's still true for this PR, the goal of this PR is to avoid
calling `ary_heap_realloc` on frozen arrays. The capacity should be
shrunk _before_ the array is frozen, not after.
Co-authored-by: Aaron Patterson <[email protected]>
Co-Authored-By: methodmissing <[email protected]>
|
|
|
|
This `st_table` is used to both mark and pin classes
defined from the C API. But `vm->mark_object_ary` already
does both much more efficiently.
Currently a Ruby process starts with 252 rooted classes,
which uses `7224B` in an `st_table` or `2016B` in an `RArray`.
So a baseline of 5kB saved, but since `mark_object_ary` is
preallocated with `1024` slots but only use `405` of them,
it's a net `7kB` save.
`vm->mark_object_ary` is also being refactored.
Prior to this changes, `mark_object_ary` was a regular `RArray`, but
since this allows for references to be moved, it was marked a second
time from `rb_vm_mark()` to pin these objects.
This has the detrimental effect of marking these references on every
minors even though it's a mostly append only list.
But using a custom TypedData we can save from having to mark
all the references on minor GC runs.
Addtionally, immediate values are now ignored and not appended
to `vm->mark_object_ary` as it's just wasted space.
|
|
enumerator_block_call was not safe for compaction because the Array
backing the argv was not pinned, so it could get moved during compaction
which would make argv point to somewhere else.
The test crashes when RGENGC_CHECK_MODE is turned on:
TestEnumerator#test_with_index_under_gc_compact_stress
check_rvalue_consistency: 0x55db0b399450 is not a Ruby object.
test/ruby/test_enumerator.rb:133: [BUG] check_rvalue_consistency_force: there is 1 errors.
ruby 3.3.0dev (2023-12-23T23:00:27Z master 50bf437341) [x86_64-linux]
-- Control frame information -----------------------------------------------
c:0034 p:---- s:0192 e:000187 CFUNC :with_index
c:0033 p:---- s:0185 e:000184 CFUNC :each
c:0032 p:---- s:0182 e:000181 CFUNC :to_a
c:0031 p:0055 s:0178 e:000175 BLOCK test/ruby/test_enumerator.rb:133
c:0030 p:0024 s:0172 e:000171 METHOD tool/lib/envutil.rb:242
c:0029 p:0024 s:0167 e:000166 METHOD tool/lib/envutil.rb:251
c:0028 p:0005 s:0160 e:000159 METHOD test/ruby/test_enumerator.rb:131
...
-- C level backtrace information -------------------------------------------
build/ruby(rb_print_backtrace+0x14) [0x55db0b1deb21] vm_dump.c:820
build/ruby(rb_vm_bugreport) vm_dump.c:1151
build/ruby(bug_report_end+0x0) [0x55db0b3a53a6] error.c:1042
build/ruby(rb_bug_without_die) error.c:1042
build/ruby(die+0x0) [0x55db0afc77c2] error.c:1050
build/ruby(rb_bug) error.c:1052
build/ruby(gc_move+0x0) [0x55db0afbada0] gc.c:1714
build/ruby(check_rvalue_consistency+0xa) [0x55db0afef0c3] gc.c:1729
build/ruby(is_markable_object) gc.c:4769
build/ruby(gc_mark_stack_values) gc.c:6595
build/ruby(rb_gc_mark_vm_stack_values) gc.c:6605
build/ruby(rb_execution_context_mark+0x39) [0x55db0b1d8589] vm.c:3309
build/ruby(thread_mark+0x15) [0x55db0b1a9805] vm.c:3381
build/ruby(gc_mark_stacked_objects+0x6d) [0x55db0aff2c3d] gc.c:7564
build/ruby(gc_mark_stacked_objects_all) gc.c:7602
build/ruby(gc_marks_rest) gc.c:8797
build/ruby(gc_marks+0xd) [0x55db0aff43d5] gc.c:8855
build/ruby(gc_start) gc.c:9608
build/ruby(rb_multi_ractor_p+0x0) [0x55db0aff5463] gc.c:9489
build/ruby(rb_vm_lock_leave) vm_sync.h:92
build/ruby(garbage_collect) gc.c:9491
build/ruby(newobj_slowpath+0xcb) [0x55db0aff57ab] gc.c:2871
build/ruby(newobj_slowpath_wb_protected) gc.c:2895
build/ruby(newobj_of0+0x24) [0x55db0aff59e4] gc.c:2937
build/ruby(newobj_of) gc.c:2947
build/ruby(rb_wb_protected_newobj_of) gc.c:2962
build/ruby(ary_alloc_embed+0x10) [0x55db0b2f3e40] array.c:668
build/ruby(ary_new) array.c:709
build/ruby(rb_ary_tmp_new_from_values) array.c:759
build/ruby(rb_ary_new_from_values) array.c:771
build/ruby(args_copy+0x18) [0x55db0b1bbb88] vm_args.c:158
|
|
Instead of `RUBY_REFERENCES_START` and `RUBY_REFERENCES_END`, so that
auto-indent works well.
|
|
Also move `struct` so that `typedef`-ed names can be used.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This commit implements Enumerator objects on VWA. This speeds allocations
and decreases memory usage.
```
require "benchmark"
ary = []
puts(Benchmark.measure do
10_000_000.times do
u = ary.to_enum
end
end)
puts `ps -o rss= -p #{$$}`
```
Before:
```
1.500774 0.002717 1.503491 ( 1.506791)
18512
```
After:
```
0.892580 0.002539 0.895119 ( 0.897642)
16480
```
|
|
* Fix [Bug #19632]: Disable external iterator for frozen enumerator
Currently, methods to manipulate an external iterator like `#next`
and `#feed` can be called even if a receiver of an enumerator is
frozen. However, these methods change the state of an external
iterator in an enumerator. Therefore, it seems a BUG to me, and
these methods should raise FrozenError if the receiver is frozen.
This fixed the following methods to raise FrozenError if the receiver is
frozen.
- `Enumerator#next`
- `Enumerator#next_values`
- `Enumerator#peek`
- `Enumerator#peek_values`
- `Enumerator#feed`
- `Enumerator#rewind`
* Fix a typo in the document
Thanks @Maumagnaguagno.
|
|
Notes:
Merged: https://github.com/ruby/ruby/pull/6201
|
|
Fixes [Bug #19569]
Notes:
Merged: https://github.com/ruby/ruby/pull/7967
|
|
Notes:
Merged: https://github.com/ruby/ruby/pull/7153
|
|
|
|
|
|
Inconsistency pointed out by @mame:
```
>> Enumerator.product([1], [2], [3]).to_a
=> [[1, 2, 3]]
>> Enumerator.product([1], [2]).to_a
=> [[1, 2]]
>> Enumerator.product([1]).to_a
=> [1]
>> Enumerator.product().to_a
=> [nil]
```
Got fixed as follows:
```
>> Enumerator.product([1], [2], [3]).to_a
=> [[1, 2, 3]]
>> Enumerator.product([1], [2]).to_a
=> [[1, 2]]
>> Enumerator.product([1]).to_a
=> [[1]]
>> Enumerator.product().to_a
=> [[]]
```
This was due to the nature of the N-argument funcall in Ruby.
|
|
Notes:
Merged: https://github.com/ruby/ruby/pull/6974
|
|
* Especially around Enumerator.
Notes:
Merged: https://github.com/ruby/ruby/pull/6974
|
|
* See https://bugs.ruby-lang.org/issues/19078#note-30
Notes:
Merged: https://github.com/ruby/ruby/pull/6972
|
|
The use of keyword arguments should be reserved for future extension.
|
|
Notes:
Merged-By: ioquatix <[email protected]>
|
|
Notes:
Merged: https://github.com/ruby/ruby/pull/6809
|
|
Notes:
Merged: https://github.com/ruby/ruby/pull/6809
|
|
Notes:
Merged: https://github.com/ruby/ruby/pull/6721
|
|
* Fix Array#[] with ArithmeticSequence with negative steps
Previously, Array#[] when called with an ArithmeticSequence
with a negative step did not handle all cases correctly,
especially cases involving infinite ranges, inverted ranges,
and/or exclusive ends.
Fixes [Bug #18247]
* Add Array#slice tests for ArithmeticSequence with negative step to test_array
Add tests of rb_arithmetic_sequence_beg_len_step C-API function.
* Fix ext/-test-/arith_seq/beg_len_step/depend
* Rename local variables
* Fix a variable name
Co-authored-by: Kenta Murata <[email protected]>
Notes:
Merged-By: mrkn <[email protected]>
|
|
|
|
Notes:
Merged: https://github.com/ruby/ruby/pull/6197
|
|
[Misc #18891]
Notes:
Merged: https://github.com/ruby/ruby/pull/6094
|
|
Runtime assertion for the argument declared as non-null.
This macro does nothing if `RBIMPL_ATTR_NONNULL` is effective,
otherwise asserts that the argument is non-null.
Notes:
Merged: https://github.com/ruby/ruby/pull/4898
Merged-By: nobu <[email protected]>
|
|
|
|
Notes:
Merged: https://github.com/ruby/ruby/pull/4821
|
|
Notes:
Merged: https://github.com/ruby/ruby/pull/4791
|
|
* --braces-after-func-def-line
* --dont-cuddle-else
* --procnames-start-lines
* --space-after-for
* --space-after-if
* --space-after-while
|
|
Fixes [Bug #17889]
Notes:
Merged: https://github.com/ruby/ruby/pull/4534
|
|
If a block is given, it returns a lazy enumerator that will iterate
over the block, it doesn't iterate over the block immediately.
Fixes [Bug #17789]
|
|
Previously these methods were defined but raised TypeError, which
seems worse.
Notes:
Merged: https://github.com/ruby/ruby/pull/3811
|
|
Implements [Feature #17347]
Notes:
Merged: https://github.com/ruby/ruby/pull/3811
|
|
This previously raised a TypeError. Wrap the Enumerator::Chain in
an Enumerator to work around the problem.
Fixes [Bug #17216]
Notes:
Merged: https://github.com/ruby/ruby/pull/3811
|