We recently found out that `make test-spec` wasn't testing with
`--yjit-call-threshold=1` like the other test suites, and that we need
to use `-T` to pass options to it.
Since we often take the VM lock as the first thing we do when entering
YJIT, and that needs a `src_loc!()`, this removes a allocation from
that. The main trick here is `concat!(file!(), '\0')` to get a C string
statically baked into the binary.
* YJIT: Add specialized codegen function for `TrueClass#===`
TrueClass#=== is currently number 10 in the most frequent C calls list of the lobsters benchmark.
```
require "benchmark/ips"
def wrap
true === true
true === false
true === :x
end
Benchmark.ips do |x|
x.report(:wrap) do
wrap
end
end
```
```
before
Warming up --------------------------------------
wrap 1.791M i/100ms
Calculating -------------------------------------
wrap 17.806M (± 1.0%) i/s - 89.544M in 5.029363s
after
Warming up --------------------------------------
wrap 4.024M i/100ms
Calculating -------------------------------------
wrap 40.149M (± 1.1%) i/s - 201.223M in 5.012527s
```
Co-authored-by: Maxime Chevalier-Boisvert <maxime.chevalierboisvert@shopify.com>
Co-authored-by: Takashi Kokubun (k0kubun) <takashikkbn@gmail.com>
Co-authored-by: Kevin Menard <kevin.menard@shopify.com>
Co-authored-by: Alan Wu <XrXr@users.noreply.github.com>
* Fix the new test for RJIT
---------
Co-authored-by: Maxime Chevalier-Boisvert <maxime.chevalierboisvert@shopify.com>
Co-authored-by: Takashi Kokubun (k0kubun) <takashikkbn@gmail.com>
Co-authored-by: Kevin Menard <kevin.menard@shopify.com>
Co-authored-by: Alan Wu <XrXr@users.noreply.github.com>
ast_alloc uses TypedData_Make_Struct, which allocates a rb_ast_t. But it
is overwritten when we set the DATA_PTR so the original memory is leaked.
For example:
10.times do
100_000.times do
eval("")
end
puts `ps -o rss= -p #{$$}`
end
Before:
17328
20752
23664
28400
30656
34224
37424
40784
43328
46656
After:
14320
14320
14320
14320
14320
14320
14320
14336
14336
14336
To make sure we can always update to the latest resolvable version for
each gem explicitly requested for update, we first run a full update,
and then add explicit exact requirements to the resolved versions. This
may lead into conflicts, but our resolver already automatically parses
those and unlocks additional gems to fix them.
https://github.com/rubygems/rubygems/commit/01c0bf34f0
An old platform related bug fix made some existing lockfiles no longer
work because they included invalid platforms. So to make it backwards
compatible, code was added to remove invalid platforms from the lockfile
before resolution. This is skipped though when Gemfile has changed
dependencies because in that case we will be re-resolving anyways.
However, in the `bundle update` case, the detection of "dependencies
have changed" was not actually working making Bundler remove all
platforms and not be able to resolve.
https://github.com/rubygems/rubygems/commit/6452adfd62
This patch moves `ast->node_buffer->config` to `ast->config` aiming to improve readability and maintainability of the source.
## Background
We could not add the `config` field to the `rb_ast_t *` due to the five-word restriction of the IMEMO object.
But it is now doable by merging https://github.com/ruby/ruby/pull/10618
## About assigning `&rb_global_parser_config` to `ast->config` in `ast_alloc()`
The approach of not setting `ast->config` in `ast_alloc()` means that the client, CRuby in this scenario, that directly calls `ast_alloc()` will be responsible for releasing it if a resource that is passed to AST needs to be released.
However, we have put on hold whether we can guarantee the above so far, thus, this patch looks like that.
```
// ruby_parser.c
static VALUE
ast_alloc(void)
{
rb_ast_t *ast;
VALUE vast = TypedData_Make_Struct(0, rb_ast_t, &ast_data_type, ast);
#ifdef UNIVERSAL_PARSER
ast = (rb_ast_t *)DATA_PTR(vast);
ast->config = &rb_global_parser_config;
#endif
return vast;
}
```
Values defined for OPT_THREADED_CODE are 0,1,2. However, 1,2,3 are set in workflow. It seems that case 3 does not exist, so 0 is specified instead.
Co-authored-by: jinroq <jinroq@users.noreply.github.com>
