Object Shapes is used for accessing instance variables and representing the
"frozenness" of objects. Object instances have a "shape" and the shape
represents some attributes of the object (currently which instance variables are
set and the "frozenness"). Shapes form a tree data structure, and when a new
instance variable is set on an object, that object "transitions" to a new shape
in the shape tree. Each shape has an ID that is used for caching. The shape
structure is independent of class, so objects of different types can have the
same shape.
For example:
```ruby
class Foo
def initialize
# Starts with shape id 0
@a = 1 # transitions to shape id 1
@b = 1 # transitions to shape id 2
end
end
class Bar
def initialize
# Starts with shape id 0
@a = 1 # transitions to shape id 1
@b = 1 # transitions to shape id 2
end
end
foo = Foo.new # `foo` has shape id 2
bar = Bar.new # `bar` has shape id 2
```
Both `foo` and `bar` instances have the same shape because they both set
instance variables of the same name in the same order.
This technique can help to improve inline cache hits as well as generate more
efficient machine code in JIT compilers.
This commit also adds some methods for debugging shapes on objects. See
`RubyVM::Shape` for more details.
For more context on Object Shapes, see [Feature: #18776]
Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
Co-Authored-By: Eileen M. Uchitelle <eileencodes@gmail.com>
Co-Authored-By: John Hawthorn <john@hawthorn.email>
Object Shapes is used for accessing instance variables and representing the
"frozenness" of objects. Object instances have a "shape" and the shape
represents some attributes of the object (currently which instance variables are
set and the "frozenness"). Shapes form a tree data structure, and when a new
instance variable is set on an object, that object "transitions" to a new shape
in the shape tree. Each shape has an ID that is used for caching. The shape
structure is independent of class, so objects of different types can have the
same shape.
For example:
```ruby
class Foo
def initialize
# Starts with shape id 0
@a = 1 # transitions to shape id 1
@b = 1 # transitions to shape id 2
end
end
class Bar
def initialize
# Starts with shape id 0
@a = 1 # transitions to shape id 1
@b = 1 # transitions to shape id 2
end
end
foo = Foo.new # `foo` has shape id 2
bar = Bar.new # `bar` has shape id 2
```
Both `foo` and `bar` instances have the same shape because they both set
instance variables of the same name in the same order.
This technique can help to improve inline cache hits as well as generate more
efficient machine code in JIT compilers.
This commit also adds some methods for debugging shapes on objects. See
`RubyVM::Shape` for more details.
For more context on Object Shapes, see [Feature: #18776]
Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
Co-Authored-By: Eileen M. Uchitelle <eileencodes@gmail.com>
Co-Authored-By: John Hawthorn <john@hawthorn.email>
Previously, newline: :lf was accepted but ignored. Where it
should have been used was commented out code that didn't work,
but unlike all other invalid values, using newline: :lf did
not raise an error.
This adds support for newline: :lf and :lf_newline, for consistency
with newline: :cr and :cr_newline. This is basically the same as
universal_newline, except that it only affects writing and not
reading due to RUBY_ECONV_NEWLINE_DECORATOR_WRITE_MASK.
Add tests for the File.open :newline option while here.
Fixes [Bug #12436]
As commented in include/ruby/internal/abi.h, since teeny versions of
Ruby should guarantee ABI compatibility, `RUBY_ABI_VERSION` has no role
in released versions of Ruby.
rb_ary_tmp_new suggests that the array is temporary in some way, but
that's not true, it just creates an array that's hidden and not on the
transient heap. This commit renames it to rb_ary_hidden_new.
This commit implements Objects on Variable Width Allocation. This allows
Objects with more ivars to be embedded (i.e. contents directly follow the
object header) which improves performance through better cache locality.
The (sole) use of memcpy in our public header is now replaced to
directly call ruby_nonempty_memcpy, and the previous definition of
memcpy is now internal-only. [Bug#18893]
Since enabling YJIT or MJIT drastically changes what could go wrong at
runtime, it's good to be front and center about whether they are enabled
when dumping a crash report. Previously, `RUBY_DESCRIPTION` and the
description printed when crashing can be different when a JIT is on.
Introduce a new internal data global, `rb_dynamic_description`, and set
it to be the same as `RUBY_DESCRIPTION` during initialization; use it
when crashing.
* version.c: Init_ruby_description(): Initialize and use
`rb_dynamic_description`.
* error.c: Change crash reports to use `rb_dynamic_description`.
* ruby.c: Call `Init_ruby_description()` earlier. Slightly more work
for when we exit right after printing the description but that
was deemed acceptable.
* include/ruby/version.h: Talk about how JIT info is not in
`ruby_description`.
* test/-ext-/bug_reporter/test_bug_reporter.rb: Remove handling for
crash description being different from `RUBY_DESCRIPTION`.
* test/ruby/test_rubyoptions.rb: ditto
Co-authored-by: Nobuyoshi Nakada <nobu@ruby-lang.org>
Co-authored-by: Alan Wu <alanwu@ruby-lang.org>
[Feature #18339]
After experimenting with the initial version of the API I figured there is a need
for an exit event to cleanup instrumentation data. e.g. if you record data in a
{thread_id -> data} table, you need to free associated data when a thread goes away.
This function was added to a public header in [1] probably
unintentionally since it's not used anywhere, exposes implementation
details, and isn't related to the goals of that pull request.
[1]: 56cc3e99b6
Previously, because opt_aref and opt_aset don't push a frame, when they
would call rb_hash to determine the hash value of the key, the initial
level of recursion would incorrectly use the method id at the top of the
stack instead of "hash".
This commit replaces rb_exec_recursive_outer with
rb_exec_recursive_outer_mid, which takes an explicit method id, so that
we can make the hash calculation behave consistently.
rb_exec_recursive_outer was documented as being internal, so I believe
this should be okay to change.
Ref: https://bugs.ruby-lang.org/issues/18339
Design:
- This tries to minimize the overhead when no hook is registered.
It should only incur an extra unsynchronized boolean check.
- The hook list is protected with a read-write lock as to cause
contention when some hooks are registered.
- The hooks MUST be thread safe, and MUST NOT call into Ruby as they
are executed outside the GVL.
- It's simply a noop on Windows.
API:
```
rb_internal_thread_event_hook_t * rb_internal_thread_add_event_hook(rb_internal_thread_event_callback callback, rb_event_flag_t internal_event, void *user_data);
bool rb_internal_thread_remove_event_hook(rb_internal_thread_event_hook_t * hook);
```
You can subscribe to 3 events:
- READY: called right before attempting to acquire the GVL
- RESUMED: called right after successfully acquiring the GVL
- SUSPENDED: called right after releasing the GVL.
The hooks MUST be threadsafe, as they are executed outside of the GVL, they also MUST NOT call any Ruby API.
Having more size pools will allow us to allocate larger objects
through Variable Width Allocation.
I have attached some benchmark results below.
Discourse:
On Discourse, we don't see much change in response times. We do see
a small reduction in RSS.
Branch RSS: 377.8 MB
Master RSS: 396.3 MB
railsbench:
On railsbench, we don't see a big change in RPS or p99 performance.
We see a small increase in RSS.
Branch RPS: 815.38
Master RPS: 811.73
Branch p99: 1.69 ms
Master p99: 1.68 ms
Branch RSS: 90.6 MB
Master RSS: 89.4 MB
liquid:
We don't see a significant change in liquid performance.
Branch parse & render: 29.041 I/s
Master parse & render: 29.211 I/s
[Feature #18683]
This allows parsers and similar libraries to create Hashes of
a certain capacity in advance. It's useful when the key and values
are streamed, hence `bulk_insert()` can't be used.
This commit reintroduces finer-grained constant cache invalidation.
After 8008fb7 got merged, it was causing issues on token-threaded
builds (such as on Windows).
The issue was that when you're iterating through instruction sequences
and using the translator functions to get back the instruction structs,
you're either using `rb_vm_insn_null_translator` or
`rb_vm_insn_addr2insn2` depending if it's a direct-threading build.
`rb_vm_insn_addr2insn2` does some normalization to always return to
you the non-trace version of whatever instruction you're looking at.
`rb_vm_insn_null_translator` does not do that normalization.
This means that when you're looping through the instructions if you're
trying to do an opcode comparison, it can change depending on the type
of threading that you're using. This can be very confusing. So, this
commit creates a new translator function
`rb_vm_insn_normalizing_translator` to always return the non-trace
version so that opcode comparisons don't have to worry about different
configurations.
[Feature #18589]
This reverts commits for [Feature #18589]:
* 8008fb7352
"Update formatting per feedback"
* 8f6eaca2e1
"Delete ID from constant cache table if it becomes empty on ISEQ free"
* 629908586b
"Finer-grained inline constant cache invalidation"
MSWin builds on AppVeyor have been crashing since the merger.
Current behavior - caches depend on a global counter. All constant mutations cause caches to be invalidated.
```ruby
class A
B = 1
end
def foo
A::B # inline cache depends on global counter
end
foo # populate inline cache
foo # hit inline cache
C = 1 # global counter increments, all caches are invalidated
foo # misses inline cache due to `C = 1`
```
Proposed behavior - caches depend on name components. Only constant mutations with corresponding names will invalidate the cache.
```ruby
class A
B = 1
end
def foo
A::B # inline cache depends constants named "A" and "B"
end
foo # populate inline cache
foo # hit inline cache
C = 1 # caches that depend on the name "C" are invalidated
foo # hits inline cache because IC only depends on "A" and "B"
```
Examples of breaking the new cache:
```ruby
module C
# Breaks `foo` cache because "A" constant is set and the cache in foo depends
# on "A" and "B"
class A; end
end
B = 1
```
We expect the new cache scheme to be invalidated less often because names aren't frequently reused. With the cache being invalidated less, we can rely on its stability more to keep our constant references fast and reduce the need to throw away generated code in YJIT.
This commit implements arrays on Variable Width Allocation. This allows
longer arrays to be embedded (i.e. contents directly follow the object
header) which improves performance through better cache locality.
Header file include/ruby/internal/abi.h contains RUBY_ABI_VERSION which
is the ABI version. This value should be bumped whenever an ABI
incompatible change is introduced.
When loading dynamic libraries, Ruby will compare its own
`ruby_abi_version` and the `ruby_abi_version` of the loaded library. If
these two values don't match it will raise a `LoadError`. This feature
can also be turned off by setting the environment variable
`RUBY_RUBY_ABI_CHECK=0`.
This feature will prevent cases where previously installed native gems
fail in unexpected ways due to incompatibility of changes in header
files. This will force the developer to recompile their gems to use the
same header files as the built Ruby.
In Ruby, the ABI version is exposed through
`RbConfig::CONFIG["ruby_abi_version"]`.
Configuration for mingw32 can't detect 'shutdown' due to wrong -l
option even though it's available (this has been going on for a while,
and it needs to be fixed).
In this situation, include/ruby/missing.h declares a stub shutdown
function since 7ee786388a, and another shutdown decl is came from
system header. They are incompatible at stdcall attribute, so it
causes compilation failure.
This change defines a HAVE_SHUTDOWN to guard a newly introduced stub
decl in include/ruby/missing.h
At least OpenBSD/sparc64 doesn't appear to define them, and possibly
some other OpenBSD GCC platforms don't (most OpenBSD platforms have
already switched to clang).
- prerequisite of supporting YJIT with VC++.
- note that now can specfily `--yjit` on mswin64, but not enabled
YJIT'ed code because of YJIT requires `OPT_DIRECT_THREADED_CODE`
or `OPT_CALL_THREADED_CODE` in `rb_yjit_compile_iseq`.
Presence of RBIMPL_ATTR_NONNULL let C compilers to eliminate
must_not_null(). Because null pointers are not allowed to exist there
are no reason to call the function. In reality null pointers are still
passed to those functions in a number of ways. Runtime check for them
are definitely nice to have. fix [Feature#18280]
This commit deprecates rb_gc_force_recycle and coverts it to a no-op
function. Also removes invalidate_mark_stack_chunk since only
rb_gc_force_recycle uses it.
Refinement#import_methods imports methods from modules.
Unlike Module#include, it copies methods and adds them into the refinement,
so the refinement is activated in the imported methods.
[Bug #17429] [ruby-core:101639]
2,291 lines are too much! include/ruby/encoding.h became the biggest
header file once it had doxygen comments. Let us split it into smaller
parts, so that we can better organise their contents.
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.
Silly bug, they write back consumed bytes through passed pointers. Must
never be pure functions.
ruby_scan_oct does not refer any static variables so it can still be
__declspec(noalias), while ruby_scan_hex is not because it reads from
ruby_digit36_to_number_table.
The only thing that remains in this file which is still worth
documenting is the RUBY macro. Everything else were split into many
files in
1ff4cee2b1
[ci skip]
For instane CHAR_BIT is an ANSI C macro which should be documented in
other places than Ruby (and the definition in this file must work
exactly identical to that of standard one). [ci skip]
Jean Boussier
Samuel Williams
Aaron Patterson
Jemma Issroff
Nobuyoshi Nakada
卜部昌平
Takashi Kokubun
spaette
Jeremy Evans
Peter Zhu
Yusuke Endoh
Kevin Backhouse
Chris Seaton
Jean Boussier
Alan Wu
John Hawthorn
Matt Valentine-House
Daisuke Fujimura (fd0)
Alex Matchneer
Kevin Newton
Lars Kanis
Yuta Saito
U.Nakamura
Yukihiro "Matz" Matsumoto
Kazuhiro NISHIYAMA
Kenta Murata
Shugo Maeda