This commit adds a `capacity` field to shapes, and adds shape
transitions whenever an object's capacity changes. Objects which are
allocated out of a bigger size pool will also make a transition from the
root shape to the shape with the correct capacity for their size pool
when they are allocated.
This commit will allow us to remove numiv from objects completely, and
will also mean we can guarantee that if two objects share shapes, their
IVs are in the same positions (an embedded and extended object cannot
share shapes). This will enable us to implement ivar sets in YJIT using
object shapes.
Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
There is currently a bug ([#19084]) in how extended strings work in
IO Buffers. Object Shapes changes will make the string in this test
extended on 32 bit machines. Since this behavior is currently broken
(unrelated to object shapes) on 32 bit machines, this test will then
fail. We preemptively omit it so that this commit can be reverted
once the bug is fixed.
[Bug #19105]
If no fiber scheduler is registered and the fiber that
owns the lock and the one that try to acquire it
both belong to the same thread, we're in a deadlock case.
Co-authored-by: Jean Boussier <byroot@ruby-lang.org>
`throw TAG_BREAK` instruction makes a jump only if the continuation of
catch of TAG_BREAK exactly matches the instruction immediately following
the "send" instruction that is currently being executed. Otherwise, it
seems to determine break from proc-closure.
Branch coverage may insert some recording instructions after "send"
instruction, which broke the conditions for TAG_BREAK to work properly.
This change forces to set the continuation of catch of TAG_BREAK
immediately after "send" (or "invokesuper") instruction.
[Bug #18991]
conveniently
(https://github.com/ruby/irb/pull/429)
* Create a base TestIRB::TestCase class
* Save/restore encodings for tests that initializes InputMethod classes
Because `RelineInputMethod#initializes` calls `set_encoding`, which
changes stdio/out/err and Encoding's default encoding values, we need to
make sure any test that directly or indirectly (e.g. through Context)
initializes `RelineInputMethod` restores encodings.
`ReadlineInputMethod` also changes encodings but currently no tests
cover it.
* Remove unnecessary TestHelper module
Since we now have a base TestCase, without_rdoc can just live there.
c2874ec121
(https://github.com/ruby/irb/pull/426)
* Allow non-identifier aliases
* Move the configuration to IRB.conf
* Avoid abusing method lookup for symbol aliases
* Add more alias tests
* A small optimization
* Assume non-nil Context
* Load IRB.conf earlier
e23db5132e
(https://github.com/ruby/irb/pull/428)
RubyLex is not designed to be used alone. It's usually used with an IRB
context, which requires workspace. So its tests should have access to
those components too.
608f261da4
(https://github.com/ruby/irb/pull/427)
* Make sure `RubyLex#set_input`'s context is always present in tests
In real-world scenarios, the context should always be non-nil:
https://github.com/ruby/irb/blob/master/lib/irb.rb#L489
So we should make sure our test setup reflects that.
* Make context a required keyword
Since in practice, `set_input`'s context should always be non-nil, its
parameters should reflect that.
And since `RubyLex#check_state` is only called by `#lex` and
`#set_input`, both of which now always require context, we can assume
its context should be non-nil too.
1aeeb86203
Fixes a bug where `spot` was using the wrong local variable.
We want to use error highlight with code that has been eval'd,
specifically ERB templates. We can recover the compiled source code of
the ERB template but we need an API to pass the node into error
highlight's `spot`.
Required Ruby PR: https://github.com/ruby/ruby/pull/65930b1b650a59
Co-authored-by: Aaron Patterson <tenderlove@ruby-lang.org>
We want to use error highlight with eval'd code, specifically ERB
templates. We're able to recover the generated code for eval'd templates
and can get a parse tree for the ERB generated code, but we don't have a
way to get the node id from the backtrace location. So we can't pass the
right node into error highlight.
This patch gives us an API to get the node id from the backtrace
location so we can find the node in the AST.
Error Highlight PR: https://github.com/ruby/error_highlight/pull/26
Co-authored-by: Aaron Patterson <tenderlove@ruby-lang.org>
https://no-color.org has been updated (jcs/no_color#83):
> Command-line software which adds ANSI color to its output by default
should check for a `NO_COLOR` environment variable that, when present
and **not an empty string** (regardless of its value), prevents the
addition of ANSI color.
46e0f7e370
Co-authored-by: Stan Lo <stan001212@gmail.com>
This proc displays rdoc document when the input matches certain symbols
perfectly, like "String". It's commonly triggered with autocompletion
but only has 1 test case. So this commit increases its test coverage.
d85d719313
