vm_invoke_builtin() accesses VM stack via cfp->sp. However, MJIT
can use their own stack. To access them appropriately, we need to
use STACK_ADDR_FROM_TOP().
A method which has keyword parameters has an implicit local variable
to specify which keywords are (un)specified.
vm_call_iseq_setup_kwparm_nokwarg() is special function to invoke
a ISeq method without any keyword arguments. However, it should
also initialize the special local var. Without this initialization,
the implicit lvar can points a freed (T_NONE) object.
Support loading builtin features written in Ruby, which implement
with C builtin functions.
[Feature #16254]
Several features:
(1) Load .rb file at boottime with native binary.
Now, prelude.rb is loaded at boottime. However, this file is contained
into the interpreter as a text format and we need to compile it.
This patch contains a feature to load from binary format.
(2) __builtin_func() in Ruby call func() written in C.
In Ruby file, we can write `__builtin_func()` like method call.
However this is not a method call, but special syntax to call
a function `func()` written in C. C functions should be defined
in a file (same compile unit) which load this .rb file.
Functions (`func` in above example) should be defined with
(a) 1st parameter: rb_execution_context_t *ec
(b) rest parameters (0 to 15).
(c) VALUE return type.
This is very similar requirements for functions used by
rb_define_method(), however `rb_execution_context_t *ec`
is new requirement.
(3) automatic C code generation from .rb files.
tool/mk_builtin_loader.rb creates a C code to load .rb files
needed by miniruby and ruby command. This script is run by
BASERUBY, so *.rb should be written in BASERUBY compatbile
syntax. This script load a .rb file and find all of __builtin_
prefix method calls, and generate a part of C code to export
functions.
tool/mk_builtin_binary.rb creates a C code which contains
binary compiled Ruby files needed by ruby command.
Prior to this changeset, majority of inline cache mishits resulted
into the same method entry when rb_callable_method_entry() resolves
a method search. Let's not call the function at the first place on
such situations.
In doing so we extend the struct rb_call_cache from 44 bytes (in
case of 64 bit machine) to 64 bytes, and fill the gap with
secondary class serial(s). Call cache's class serials now behavies
as a LRU cache.
Calculating -------------------------------------
ours 2.7 2.6
vm2_poly_same_method 2.339M 1.744M 1.369M i/s - 6.000M times in 2.565086s 3.441329s 4.381386s
Comparison:
vm2_poly_same_method
ours: 2339103.0 i/s
2.7: 1743512.3 i/s - 1.34x slower
2.6: 1369429.8 i/s - 1.71x slower
Noticed that rb_method_basic_definition_p is frequently called.
Its callers include vm_caller_setup_args_block(),
rb_hash_default_value(), rb_num_neative_int_p(), and a lot more.
It seems worth caching the method resolution part. Majority of
rb_method_basic_definion_p() usages take fixed class and fixed
method id combinations.
Calculating -------------------------------------
ours trunk
so_matrix 2.379 2.115 i/s - 1.000 times in 0.420409s 0.472879s
Comparison:
so_matrix
ours: 2.4 i/s
trunk: 2.1 i/s - 1.12x slower
This mirrors the behavior when manually splatting a hash. This
mirrors the changes made in setup_parameters_complex in
6081ddd6e6, so that splatting to a
non-iseq method works the same as splatting to an iseq method.
To perform a regular method call, the VM needs two structs,
`rb_call_info` and `rb_call_cache`. At the moment, we allocate these two
structures in separate buffers. In the worst case, the CPU needs to read
4 cache lines to complete a method call. Putting the two structures
together reduces the maximum number of cache line reads to 2.
Combining the structures also saves 8 bytes per call site as the current
layout uses separate two pointers for the call info and the call cache.
This saves about 2 MiB on Discourse.
This change improves the Optcarrot benchmark at least 3%. For more
details, see attached bugs.ruby-lang.org ticket.
Complications:
- A new instruction attribute `comptime_sp_inc` is introduced to
calculate SP increase at compile time without using call caches. At
compile time, a `TS_CALLDATA` operand points to a call info struct, but
at runtime, the same operand points to a call data struct. Instruction
that explicitly define `sp_inc` also need to define `comptime_sp_inc`.
- MJIT code for copying call cache becomes slightly more complicated.
- This changes the bytecode format, which might break existing tools.
[Misc #16258]
This reverts commits: 10d6a3aca78ba48c1b85fba8627dc1dd883de5ba6c6a25feca167e6b48f17cb96d41a53207979278595b3c4fdd1521f7cf89c11c5e69accf336082033632a812c0f56506be0d86427a3219 .
The reason for the revert is that we observe ABA problem around
inline method cache. When a cache misshits, we search for a
method entry. And if the entry is identical to what was cached
before, we reuse the cache. But the commits we are reverting here
introduced situations where a method entry is freed, then the
identical memory region is used for another method entry. An
inline method cache cannot detect that ABA.
Here is a code that reproduce such situation:
```ruby
require 'prime'
class << Integer
alias org_sqrt sqrt
def sqrt(n)
raise
end
GC.stress = true
Prime.each(7*37){} rescue nil # <- Here we populate CC
class << Object.new; end
# These adjacent remove-then-alias maneuver
# frees a method entry, then immediately
# reuses it for another.
remove_method :sqrt
alias sqrt org_sqrt
end
Prime.each(7*37).to_a # <- SEGV
```
return directly in class/module is an error, so return in
proc in class/module should also be an error. I believe the
previous behavior was an unintentional oversight during the
addition of top-level return in 2.4.
At last, not only myself but also your compiler are fully confident
that the method entries pointed from call caches are immutable. We
don't have to worry about silent updates. Just delete the branch
that is now always false.
Calculating -------------------------------------
ours trunk
vm2_poly_same_method 2.142M 2.070M i/s - 6.000M times in 2.801148s 2.898994s
Comparison:
vm2_poly_same_method
ours: 2141979.2 i/s
trunk: 2069683.8 i/s - 1.03x slower
Now that we have eliminated most destructive operations over the
rb_method_entry_t / rb_callable_method_entry_t, let's make them
mostly immutabe and mark them const.
One exception is rb_export_method(), which destructively modifies
visibilities of method entries. I have left that operation as is
because I suspect that destructiveness is the nature of that
function.
Tired of rb_method_entry_create(..., rb_method_definition_create(
..., &(rb_method_foo_t) {...})) maneuver. Provide a function that
does the thing to reduce copy&paste.
The deleted function was to destructively overwrite existing method
entries, which is now considered to be a bad idea. Delete it, and
assign a newly created method entry instead.
Before this changeset rb_method_definition_create only allocated a
memory region and we had to destructively initialize it later.
That is not a good design so we change the API to return a complete
struct instead.
Most (if not all) of the fields of rb_method_definition_t are never
meant to be modified once after they are stored. Marking them const
makes it possible for compilers to warn on unintended modifications.
If a method accepts no keywords and was called with a keyword, an
ArgumentError was not always issued previously. Force methods that
accept no keywords to go through setup_parameters_complex so that
an ArgumentError is raised if keywords are provided.
This approach uses a flag bit on the final hash object in the regular splat,
as opposed to a previous approach that used a VM frame flag. The hash flag
approach is less invasive, and handles some cases that the VM frame flag
approach does not, such as saving the argument splat array and splatting it
later:
ruby2_keywords def foo(*args)
@args = args
bar
end
def bar
baz(*@args)
end
def baz(*args, **kw)
[args, kw]
end
foo(a:1) #=> [[], {a: 1}]
foo({a: 1}, **{}) #=> [[{a: 1}], {}]
foo({a: 1}) #=> 2.7: [[], {a: 1}] # and warning
foo({a: 1}) #=> 3.0: [[{a: 1}], {}]
It doesn't handle some cases that the VM frame flag handles, such as when
the final hash object is replaced using Hash#merge, but those cases are
probably less common and are unlikely to properly support keyword
argument separation.
Use ruby2_keywords to handle argument delegation in the delegate library.
I noticed that in case of cache misshit, re-calculated cc->me can
be the same method entry than the pevious one. That is an okay
situation but can't we partially reuse the cache, because cc->call
should still be valid then?
One thing that has to be special-cased is when the method entry
gets amended by some refinements. That happens behind-the-scene
of call cache mechanism. We have to check if cc->me->def points to
the previously saved one.
Calculating -------------------------------------
trunk ours
vm2_poly_same_method 1.534M 2.025M i/s - 6.000M times in 3.910203s 2.962752s
Comparison:
vm2_poly_same_method
ours: 2025143.9 i/s
trunk: 1534447.2 i/s - 1.32x slower
Make sure that vm_yield_with_cfunc can correctly set the empty keyword
flag by passing 2 as the kw_splat value when calling it in
vm_invoke_ifunc_block. Make sure calling.kw_splat is set to 1 and not
128 in vm_sendish, so we can safely check for different kw_splat values.
vm_args.c needs to call add_empty_keyword, and to make JIT happy, the
function needs to be exported. Rename the function to
rb_adjust_argv_kw_splat to more accurately reflect what it does, and
mark it as MJIT exported.
Also add keyword argument separation warnings for Class#new and Method#call.
To allow for keyword argument to required positional hash converstion in
cfuncs, add a vm frame flag indicating the cfunc was called with an empty
keyword hash (which was removed before calling the cfunc). The cfunc can
check this frame flag and add back an empty hash if it is passing its
arguments to another Ruby method. Add rb_empty_keyword_given_p function
for checking if called with an empty keyword hash, and
rb_add_empty_keyword for adding back an empty hash to argv.
All of this empty keyword argument support is only for 2.7. It will be
removed in 3.0 as Ruby 3 will not convert empty keyword arguments to
required positional hash arguments. Comment all of the relevent code
to make it obvious this is expected to be removed.
Add rb_funcallv_kw as an public C-API function, just like rb_funcallv
but with a keyword flag. This is used by rb_obj_call_init (internals
of Class#new). This also required expected call_type enum with
CALL_FCALL_KW, similar to the recent addition of CALL_PUBLIC_KW.
Add rb_vm_call_kw as a internal function, used by call_method_data
(internals of Method#call and UnboundMethod#bind_call). Add tests
for UnboundMethod#bind_call keyword handling.
This is the same as the bmethod, sym proc, and send cases,
where we don't remove the keyword splat, so later code can
move it to a required positional parameter and warn.
This is the same as the bmethod and send cases, where we don't
remove the keyword splat, so later code can move it to to a
a required positional parameter and warn.
The lambda case is similar to the attr_writer case, except we have
to determine the number of required parameters from the iseq
instead of being able to assume a single required parameter.
This fixes a lot of lambda tests which were switched to require
warnings for all usage of keyword arguments. Similar to method
handling, we do not warn when passing keyword arguments to
lambdas that do not accept keyword arguments, the argument is
just passed as a positional hash in that case, unless it is empty.
If it is empty and not the final required parameter, then we
ignore it. If it is empty and the final required parameter, then
we pass it for backwards compatibility and emit a warning, as in
Ruby 3 we will not pass it.
The bmethod case is similar to the send case, in that we do not
want to remove empty keyword splats in vm_call_bmethod, as that
prevents later call handling from moving them to required
positional arguments and warning.
In general, we want to ignore empty keyword hashes. The only case
where we want to allow them for backwards compatibility is when
they are necessary to satify the final required positional argument.
In that case, we want to not ignore them, but we do want to warn,
as that will be going away in Ruby 3.
This commit implements this support for regular methods and
attr_writer methods.
In order to allow send to forward arguments correctly, send no
longer removes empty keyword hashes. It is the responsibility of
the final method to remove the empty keyword hashes now. This
change was necessary as otherwise send could remove the empty
keyword hashes before the regular or attr_writer methods could
move them to required positional arguments.
For completeness, add tests for keyword handling regular
methods calls.
This makes rb_warn_keyword_to_last_hash non-static in vm_args.c
so it can be reused in vm_insnhelper.c, and also moves declarations
before statements in the rb_warn_* functions in vm_args.c.
While doing so is not backwards compatible with Ruby 2.6, it is
necessary for generic argument forwarding to work for all methods:
```ruby
def foo(*args, **kw, &block)
bar(*args, **kw, &block)
end
```
If you do not remove empty keyword hashes, and bar does not accept
keyword arguments, then a call to foo without keyword arguments
calls bar with an extra positional empty hash argument.
Actually, the following call is wrongly warned without this change.
```
class C
def method_missing(x, *args, **opt)
end
end
C.new.foo(k: 1)
# warning: The last argument is used as the keyword parameter
# warning: for `method_missing' defined here
```
