Method#super_method crashes for aliased module methods because they are
not defined on a class. This bug was introduced in
c60aaed185 as part of bug #17130.
The most common use case for `bind_call` is to protect from core
methods being redefined, for instance a typical use:
```ruby
UNBOUND_METHOD_MODULE_NAME = Module.instance_method(:name)
def real_mod_name(mod)
UNBOUND_METHOD_MODULE_NAME.bind_call(mod)
end
```
But it's extremely common that the method wasn't actually redefined.
In such case we can avoid creating a new callable method entry,
and simply delegate to the receiver.
This result in a 1.5-2X speed-up for the fast path, and little to
no impact on the slowpath:
```
compare-ruby: ruby 3.1.0dev (2021-02-05T06:33:00Z master b2674c1fd7) [x86_64-darwin19]
built-ruby: ruby 3.1.0dev (2021-02-15T10:35:17Z bind-call-fastpath d687e06615) [x86_64-darwin19]
| |compare-ruby|built-ruby|
|:---------|-----------:|---------:|
|fastpath | 11.325M| 16.393M|
| | -| 1.45x|
|slowpath | 10.488M| 10.242M|
| | 1.02x| -|
```
They are no longer how Object#clone/Object#dup are defined. In fact
DUPSETUP is not used from anywhere. CLONESETUP has only one usage.
Let's not expose them to extension libraries.
cf https://github.com/ruby/ruby/pull/4100#discussion_r563481718
Previously, due to a change to fix bug 15608, Method#inspect output
changed for class methods:
Ruby 2.7
"#<Method: String.prepend(*)>"
Before change:
"#<Method: #<Class:Object>(Module)#prepend(*)>"
This is wrong because the Method object was created from String and
not Object. This is because the fix for bug 15608 assumed it was
being called on the singleton class of a instance, and would skip
the first singleton class until it got to the class itself. For
class methods, this results in always using the superclass. Fix
behavior to not skip until the superclass if the singleton class
is the singleton class of a module or class.
After change:
"#<Method: #<Class:Object>(Module)#prepend(*)>"
Fixes [Bug #17428]
lambda(&b) where b is given block of method (like: def foo(&b))
should warn correctly.
[Feature #17361]
Also labmda(&labmda_block) or lambda(&:to_s) (Symbol#to_proc)
should not warn (but I'm not sure who cares about it).
* `GC.auto_compact=`, `GC.auto_compact` can be used to control when
compaction runs. Setting `auto_compact=` to true will cause
compaction to occurr duing major collections. At the moment,
compaction adds significant overhead to major collections, so please
test first!
[Feature #17176]
Previously, Method#super_method looked at the called_id to
determine the method id to use, but that isn't correct for
aliased methods, because the super target depends on the
original method id, not the called_id.
Additionally, aliases can reference methods defined in other
classes and modules, and super lookup needs to start in the
super of the defined class in such cases.
This adds tests for Method#super_method for both types of
aliases, one that uses VM_METHOD_TYPE_ALIAS and another that
does not. Both check that the results for calling super
methods return the expected values.
To find the defined class for alias methods, add an rb_ prefix
to find_defined_class_by_owner in vm_insnhelper.c and make it
non-static, so that it can be called from method_super_method
in proc.c.
This bug was original discovered while researching [Bug #11189].
Fixes [Bug #17130]
Not every compilers understand that rb_raise does not return. When a
function does not end with a return statement, such compilers can issue
warnings. We would better tell them about reachabilities.
Previously, these were not implemented, and Object#== and #eql?
were used. This tries to check the proc internals to make sure
that procs created from separate blocks are treated as not equal,
but procs created from the same block are treated as equal, even
when the lazy proc allocation optimization is used.
Implements [Feature #14267]
If you look at the code flow (break -> goto), this assignment never
makes any sense. Should just remove.
I _guess_ this behaviour is unintended. Original code at commit
4dc1a21809 did something. It might be
the code flow that is buggy. However rubyspec already includes this
particular edge case at ruby/core/module/undef_method_spec.rb. I don't
think we can change the way it is any longer.
For ZSUPER methods with no defined class for the method entry, start the next lookup at the superclass of the origin class of the method owner, instead of the superclass of the method owner.
Fixes [Bug #16942]
As a semantics, Hash#each yields a 2-element array (pairs of keys and
values). So, `{ a: 1 }.each(&->(k, v) { })` should raise an exception
due to lambda's arity check.
However, the optimization that avoids Array allocation by using
rb_yield_values for blocks whose arity is more than 1 (introduced at
b9d2960337 and some commits), seemed to
overlook the lambda case, and wrongly allowed the code above to work.
This change experimentally attempts to make it strict; now the code
above raises an ArgumentError. This is an incompatible change; if the
compatibility issue is bigger than our expectation, it may be reverted
(until Ruby 3.0 release).
[Bug #12706]
Previously, if an object has a singleton class, and you call
Object#method on the object, the resulting string would include
the object's singleton class, even though the method was not
defined in the singleton class.
Change this so the we only show the singleton class if the method
is defined in the singleton class.
Fixes [Bug #15608]
This removes the warnings added in 2.7, and changes the behavior
so that a final positional hash is not treated as keywords or
vice-versa.
To handle the arg_setup_block splat case correctly with keyword
arguments, we need to check if we are taking a keyword hash.
That case didn't have a test, but it affects real-world code,
so add a test for it.
This removes rb_empty_keyword_given_p() and related code, as
that is not needed in Ruby 3. The empty keyword case is the
same as the no keyword case in Ruby 3.
This changes rb_scan_args to implement keyword argument
separation for C functions when the : character is used.
For backwards compatibility, it returns a duped hash.
This is a bad idea for performance, but not duping the hash
breaks at least Enumerator::ArithmeticSequence#inspect.
Instead of having RB_PASS_CALLED_KEYWORDS be a number,
simplify the code by just making it be rb_keyword_given_p().
Saves comitters' daily life by avoid #include-ing everything from
internal.h to make each file do so instead. This would significantly
speed up incremental builds.
We take the following inclusion order in this changeset:
1. "ruby/config.h", where _GNU_SOURCE is defined (must be the very
first thing among everything).
2. RUBY_EXTCONF_H if any.
3. Standard C headers, sorted alphabetically.
4. Other system headers, maybe guarded by #ifdef
5. Everything else, sorted alphabetically.
Exceptions are those win32-related headers, which tend not be self-
containing (headers have inclusion order dependencies).
Before this commit, Kernel#lambda can't tell the difference between a
directly passed literal block and one passed with an ampersand.
A block passed with an ampersand is semantically speaking already a
non-lambda proc. When Kernel#lambda receives a non-lambda proc, it
should simply return it.
Implementation wise, when the VM calls a method with a literal block, it
places the code for the block on the calling control frame and passes a
pointer (block handler) to the callee. Before this commit, the VM
forwards block arguments by simply forwarding the block handler, which
leaves the slot for block code unused when a control frame forwards its
block argument. I use the vacant space to indicate that a frame has
forwarded its block argument and inspect that in Kernel#lambda to detect
forwarded blocks.
This is a very ad-hoc solution and relies *heavily* on the way block
passing works in the VM. However, it's the most self-contained solution
I have.
[Bug #15620]
This allows passing keywords through a normal argument splat in a
Proc. While needing ruby2_keywords support for methods is more
common, there is code that delegates keywords through normal
argument splats in procs, including code in Rails. For that
reason, it makes sense to expose this for procs as well.
Internally, ruby2_keywords is not tied to methods, but iseqs,
so this just allows for setting the ruby2_keywords for the iseq
related to the proc.
The docs are wrong about the behaviour of `#>>` (looks like it was copied from `#<<`)
In `(prc >> g).call(n)` _prc_ is called first (with _n_), *then* _g_ is called with the result.
Code examples are OK.
This removes the related tests, and puts the related specs behind
version guards. This affects all code in lib, including some
libraries that may want to support older versions of Ruby.
This removes the security features added by $SAFE = 1, and warns for access
or modification of $SAFE from Ruby-level, as well as warning when calling
all public C functions related to $SAFE.
This modifies some internal functions that took a safe level argument
to no longer take the argument.
rb_require_safe now warns, rb_require_string has been added as a
version that takes a VALUE and does not warn.
One public C function that still takes a safe level argument and that
this doesn't warn for is rb_eval_cmd. We may want to consider
adding an alternative method that does not take a safe level argument,
and warn for rb_eval_cmd.
Looking at the list of symbols inside of libruby-static.a, I found
hundreds of functions that are defined, but used from nowhere.
There can be reasons for each of them (e.g. some functions are
specific to some platform, some are useful when debugging, etc).
However it seems the functions deleted here exist for no reason.
This changeset reduces the size of ruby binary from 26,671,456
bytes to 26,592,864 bytes on my machine.
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
```
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.
This fixes instance_exec and similar methods. It also fixes
Enumerator::Yielder#yield, rb_yield_block, and a couple of cases
with Proc#{<<,>>}.
This support requires the addition of rb_yield_values_kw, similar to
rb_yield_values2, for passing the keyword flag.
Unlike earlier attempts at this, this does not modify the rb_block_call_func
type or add a separate function type. The functions of type
rb_block_call_func are called by Ruby with a separate VM frame, and we can
get the keyword flag information from the VM frame flags, so it doesn't need
to be passed as a function argument.
These changes require the following VM functions accept a keyword flag:
* vm_yield_with_cref
* vm_yield
* vm_yield_with_block
It is not safe to set this in C functions that can be called from
other C functions, as in the non argument-delegation case, you
can end up calling a Ruby method with a flag indicating keywords
are set without passing keywords.
Introduce some new *_kw functions that take a kw_splat flag and
use these functions to set RB_PASS_CALLED_KEYWORDS in places where
we know we are delegating methods (e.g. Class#new, Method#call)
Remove rb_add_empty_keyword, and instead of calling that every
place you need to add empty keyword hashes, run that code in
a single static function in vm_eval.c.
Add 4 defines to include/ruby/ruby.h, these are to be used as
int kw_splat values when calling the various rb_*_kw functions:
RB_NO_KEYWORDS :: Do not pass keywords
RB_PASS_KEYWORDS :: Pass final argument (which should be hash) as keywords
RB_PASS_EMPTY_KEYWORDS :: Add an empty hash to arguments and pass as keywords
RB_PASS_CALLED_KEYWORDS :: Passes same keyword type as current method was
called with (for method delegation)
rb_empty_keyword_given_p needs to stay. It is required if argument
delegation is done but delayed to a later point, which Enumerator
does.
Use RB_PASS_CALLED_KEYWORDS in rb_call_super to correctly
delegate keyword arguments to super method.
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.
The kw_splat flag is whether the original call passes keyword or not.
Some types of methods (e.g., bmethod and sym_proc) drops the
information. This change tries to propagate the flag to the final
callee, as far as I can.
This shows locations in places it didn't before, such as for
proc calls, and fixes the location for super calls.
This requires making iseq_location non-static and MJIT exported,
which I hope will not cause problems.
`umethod.bind_call(obj, ...)` is semantically equivalent to
`umethod.bind(obj).call(...)`. This idiom is used in some libraries to
call a method that is overridden. The added method does the same
without allocation of intermediate Method object. [Feature #15955]
```
class Foo
def add_1(x)
x + 1
end
end
class Bar < Foo
def add_1(x) # override
x + 2
end
end
obj = Bar.new
p obj.add_1(1) #=> 3
p Foo.instance_method(:add_1).bind(obj).call(1) #=> 2
p Foo.instance_method(:add_1).bind_call(obj, 1) #=> 2
```
We can check the function pointer passed to rb_define_global_function
like we do so in rb_define_method. It turns out that almost anybody
is misunderstanding the API.
After 5e86b005c0, I now think ANYARGS is
dangerous and should be extinct. This commit deletes ANYARGS from
rb_proc_new / rb_fiber_new, and applies RB_BLOCK_CALL_FUNC_ARGLIST
wherever necessary.
After 5e86b005c0, I now think ANYARGS is
dangerous and should be extinct. This commit deletes ANYARGS from
struct vm_ifunc, but in doing so we also have to decouple the usage
of this struct in compile.c, which (I think) is an abuse of ANYARGS.
When we use `define_method` and `define_singleton_method`,
if we supply block parameters to a block then a generated
method has corresponding parameters.
However, the doc doesn't mention it, so this info has been added.
Now Proc#to_s returns
"#<Proc:0x00000237a0f5f170@t.rb:1>".
However, it is convenient to select a file name by (double-)clicking
on some terminals by separating ' ' instead of '@' like
"#<Proc:0x00000237a0f5f170 t.rb:1>"
[Feature #16101]
Renaming this function. "No pin" leaks some implementation details. We
just want users to know that if they mark this object, the reference may
move and they'll need to update the reference accordingly.
Nobuyoshi Nakada
S.H
Benoit Daloze
Jeremy Evans
Peter Zhu
Jean Boussier
卜部昌平
Marcus Stollsteimer
zverok
Koichi Sasada
Tomás Coêlho
Aaron Patterson
Chris Seaton
Kazuki Tsujimoto
Yusuke Endoh
Florian Heinle
0x005c
Alan Wu
Paweł Przeniczny
ksss
Kazuhiro NISHIYAMA
OKURA Masafumi
git