This patch introduce M:N thread scheduler for Ractor system.
In general, M:N thread scheduler employs N native threads (OS threads)
to manage M user-level threads (Ruby threads in this case).
On the Ruby interpreter, 1 native thread is provided for 1 Ractor
and all Ruby threads are managed by the native thread.
From Ruby 1.9, the interpreter uses 1:1 thread scheduler which means
1 Ruby thread has 1 native thread. M:N scheduler change this strategy.
Because of compatibility issue (and stableness issue of the implementation)
main Ractor doesn't use M:N scheduler on default. On the other words,
threads on the main Ractor will be managed with 1:1 thread scheduler.
There are additional settings by environment variables:
`RUBY_MN_THREADS=1` enables M:N thread scheduler on the main ractor.
Note that non-main ractors use the M:N scheduler without this
configuration. With this configuration, single ractor applications
run threads on M:1 thread scheduler (green threads, user-level threads).
`RUBY_MAX_CPU=n` specifies maximum number of native threads for
M:N scheduler (default: 8).
This patch will be reverted soon if non-easy issues are found.
[Bug #19842]
This patch rewrites Ractor synchronization mechanism, send/receive
and take/yield.
* API
* Ractor::Selector is introduced for lightweight waiting
for many ractors.
* Data structure
* remove `struct rb_ractor_waiting_list` and use
`struct rb_ractor_queue takers_queue` to manage takers.
* remove `rb_ractor_t::yield_atexit` and use
`rb_ractor_t::sync::will_basket::type` to check the will.
* add `rb_ractor_basket::p.take` to represent a taking ractor.
* Synchronization protocol
* For the Ractor local GC, `take` can not make a copy object
directly so ask to generate the copy from the yielding ractor.
* The following steps shows what `r1.take` does on `r0`.
* step1: (r0) register `r0` into `r1`'s takers.
* step2: (r0) check `r1`'s status and wakeup r0 if `r1` is waiting
for yielding a value.
* step3: (r0) sleep until `r1` wakes up `r0`.
* The following steps shows what `Ractor.yield(v)` on `r1`.
* step1: (r1) check first takers of `r1` and if there is (`r0`),
make a copy object of `v` and pass it to `r0` and
wakes up `r0`.
* step2: (r1) if there is no taker ractors, sleep until
another ractor try to take.
This commit adds RVALUE_OVERHEAD for storing metadata at the end of the
slot. This commit moves the ractor_belonging_id in debug builds from the
flags to RVALUE_OVERHEAD which frees the 16 bits in the headers for
object shapes.
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>
Ractor verification requires storing the ractor id in the top 32 bits of
the object header. Unfortunately 32 bit machines only have 32 bits in
the object header. The verification code has a 32 bit left shift which
doesn't work on i686 and will clobber existing flags.
This commit disables the verification code on i686 since i686 will crash
if it's enabled.
Co-Authored-By: John Hawthorn <john@hawthorn.email>
Co-Authored-By: Jemma Issroff <jemmaissroff@gmail.com>
Now GVL is not process *Global* so this patch try to use
another words.
* `rb_global_vm_lock_t` -> `struct rb_thread_sched`
* `gvl->owner` -> `sched->running`
* `gvl->waitq` -> `sched->readyq`
* `rb_gvl_init` -> `rb_thread_sched_init`
* `gvl_destroy` -> `rb_thread_sched_destroy`
* `gvl_acquire` -> `thread_sched_to_running` # waiting -> ready -> running
* `gvl_release` -> `thread_sched_to_waiting` # running -> waiting
* `gvl_yield` -> `thread_sched_yield`
* `GVL_UNLOCK_BEGIN` -> `THREAD_BLOCKING_BEGIN`
* `GVL_UNLOCK_END` -> `THREAD_BLOCKING_END`
* removed
* `rb_ractor_gvl`
* `rb_vm_gvl_destroy` (not used)
There are GVL functions such as `rb_thread_call_without_gvl()` yet
but I don't have good name to replace them. Maybe GVL stands for
"Greate Valuable Lock" or something like that.
When a Ractor is removed, the freelist in the Ractor cache is not
returned to the GC, leaving the freelist permanently lost. This commit
recycles the freelist when the Ractor is destroyed, preventing a memory
leak from occurring.
Ractor.yield(obj, move: true) and
Ractor.select(..., yield_value: obj, move: true) tried to yield a
value with move semantices, but if the trial is faild, the obj
should not become a moved object.
To keep this rule, `wait_moving` wait status is introduced.
New yield/take process:
(1) If a ractor tried to yield (move:true), make taking racotr's
wait status `wait_moving` and make a moved object by
`ractor_move(obj)` and wakeup taking ractor.
(2) If a ractor tried to take a message from a ractor waiting fo
yielding (move:true), wakeup the ractor and wait for (1).
"experimental_everything" makes the assigned value, it means
the assignment change the state of assigned value.
"experimental_copy" tries to make a deep copy and make copyied object
sharable.
separate some fields from rb_ractor_t to rb_ractor_pub and put it
at the beggining of rb_ractor_t and declare it in vm_core.h so
vm_core.h can access rb_ractor_pub fields.
Now rb_ec_ractor_hooks() is a complete inline function and no
MJIT related issue.
Per ractor method cache (GH-#3842) only cached 1 page and this patch
caches several pages to keep at least 512 free slots if available.
If you increase the number of cached free slots, all cached slots
will be collected when the GC is invoked.
Now object allocation requires VM global lock to synchronize objspace.
However, of course, it introduces huge overhead.
This patch caches some slots (in a page) by each ractor and use cached
slots for object allocation. If there is no cached slots, acquire the global lock
and get new cached slots, or start GC (marking or lazy sweeping).
ruby_multi_ractor was a flag that indicates the interpreter doesn't
make any additional ractors (single ractor mode).
Instead of boolean flag, ruby_single_main_ractor pointer is introduced
which keeps main ractor's pointer if single ractor mode. If additional
ractors are created, ruby_single_main_ractor becomes NULL.
To manage ractor-local data for C extension, the following APIs
are defined.
* rb_ractor_local_storage_value_newkey
* rb_ractor_local_storage_value
* rb_ractor_local_storage_value_set
* rb_ractor_local_storage_ptr_newkey
* rb_ractor_local_storage_ptr
* rb_ractor_local_storage_ptr_set
At first, you need to create a key of storage by
rb_ractor_local_(value|ptr)_newkey().
For ptr storage, it accepts the type of storage,
how to mark and how to free with ractor's lifetime.
rb_ractor_local_storage_value/set are used to access a VALUE
and rb_ractor_local_storage_ptr/set are used to access a pointer.
random.c uses this API.
Random generators are not Ractor-safe, so we need to prepare
per-ractor default random genearators. This patch set
`Random::DEFAULT = Randm` (not a Random instance, but the Random
class) and singleton methods like `Random.rand()` use a per-ractor
random generator.
[Feature #17322]
To make some kind of Ractor related extensions, some functions
should be exposed.
* include/ruby/thread_native.h
* rb_native_mutex_*
* rb_native_cond_*
* include/ruby/ractor.h
* RB_OBJ_SHAREABLE_P(obj)
* rb_ractor_shareable_p(obj)
* rb_ractor_std*()
* rb_cRactor
and rm ractor_pub.h
and rename srcdir/ractor.h to srcdir/ractor_core.h
(to avoid conflict with include/ruby/ractor.h)
HParker
Koichi Sasada
Nobuyoshi Nakada
Matt Valentine-House
Peter Zhu
Aaron Patterson
Jemma Issroff
Yusuke Endoh
S.H
Benoit Daloze
Kazuhiro NISHIYAMA