Before this commit, we were mixing a lot of concerns with the prism
compile between RubyVM::InstructionSequence and the general entry
points to the prism parser/compiler.
This commit makes all of the various prism-related APIs mirror
their corresponding APIs in the existing parser/compiler. This means
we now have the correct frame naming, and it's much easier to follow
where the logic actually flows. Furthermore this consolidates a lot
of the prism initialization, making it easier to see where we could
potentially be raising errors.
pm_scope_node_destroy frees the scope node after we're done using it to
make sure that the index_lookup_table is not leaked.
For example:
10.times do
100_000.times do
RubyVM::InstructionSequence.compile_prism("begin; 1; rescue; 2; end")
end
puts `ps -o rss= -p #{$$}`
end
Before:
33056
50304
67776
84544
101520
118448
135712
152352
169136
186656
After:
15264
15296
15408
17040
17152
17152
18320
18352
18400
18608
This commit completely restructures how we handle parameters. The
motivation for this commit was the fix compilation of MultiTargetNodes
within parameters, including nested MultiTargetNodes. A subsequent
commit will actually do the compilation for the MultiTargetNodes.
This commit's main accomplishment is restructuring the locals table
and how we account for it on the ScopeNode, specifically with regards
to hidden variables.
It has multiple steps, all commented within the code, to calculate
the locals table correctly and compile the parameters:
- Step 1: Caculate the table size for the locals
- Step 2: Populate iv index table and local table
- Step 3: Fill in parameter names of MultiTargetNodes on local table
- Step 4: Fill in method body locals on local table
- Step 5: Compile any locals
Prior to this commit, we weren't accounting for hidden variables
on the locals table, so we would have inconsistencies on the stack.
This commit fixes params, and introduces a hidden_variable_count
on the scope, both of which fix parameters.
Instead of incrementing the depth using call by reference as we're
recursing up the stack we could instead store an offset for each known
scope where we know the depth is going to represented differently in the
Prism ast.
Kevin Newton
Peter Zhu
Matt Valentine-House
HParker
Jemma Issroff