Skip to content

ZJIT: Optimize NewArray to use rb_ec_ary_new_from_values #896

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
goshanraj-g wants to merge 1 commit into from
Closed

ZJIT: Optimize NewArray to use rb_ec_ary_new_from_values #896

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
20 changes: 11 additions & 9 deletions zjit/src/codegen.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -350,7 +350,7 @@ fn gen_insn(cb: &mut CodeBlock, jit: &mut JITState, asm: &mut Assembler, functio
&Insn::Const { val: Const::CPtr(val) } => gen_const_cptr(val),
&Insn::Const { val: Const::CInt64(val) } => gen_const_long(val),
Insn::Const { .. } => panic!("Unexpected Const in gen_insn: {insn}"),
Insn::NewArray { elements, state } => gen_new_array(asm, opnds!(elements), &function.frame_state(*state)),
Insn::NewArray { elements, state } => gen_new_array(jit, asm, opnds!(elements), &function.frame_state(*state)),
Insn::NewHash { elements, state } => gen_new_hash(jit, asm, opnds!(elements), &function.frame_state(*state)),
Insn::NewRange { low, high, flag, state } => gen_new_range(jit, asm, opnd!(low), opnd!(high), *flag, &function.frame_state(*state)),
Insn::NewRangeFixnum { low, high, flag, state } => gen_new_range_fixnum(asm, opnd!(low), opnd!(high), *flag, &function.frame_state(*state)),
Expand Down Expand Up @@ -1405,21 +1405,23 @@ fn gen_array_dup(

/// Compile a new array instruction
fn gen_new_array(
jit: &JITState,
asm: &mut Assembler,
elements: Vec<Opnd>,
state: &FrameState,
) -> lir::Opnd {
gen_prepare_leaf_call_with_gc(asm, state);

let length: c_long = elements.len().try_into().expect("Unable to fit length of elements into c_long");
gen_prepare_non_leaf_call(jit, asm, state);
Copy link

@tekknolagi tekknolagi Dec 3, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why non-leaf call?

Copy link
Author

@goshanraj-g goshanraj-g Dec 3, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sorry about that! I'll change it to a leaf call (gen_prepare_leaf_call_with_gc). I was referencing the hash insertion functions which were similar and utilized non-leaf calls. Now, the updated code uses leaf calls (submitted a PR to ruby/ruby)


let new_array = asm_ccall!(asm, rb_ary_new_capa, length.into());
let num: c_long = elements.len().try_into().expect("Unable to fit length of elements into c_long");

for val in elements {
asm_ccall!(asm, rb_ary_push, new_array, val);
if elements.is_empty() {
asm_ccall!(asm, rb_ec_ary_new_from_values, EC, 0i64.into(), Opnd::UImm(0))
} else {
let argv = gen_push_opnds(asm, &elements);
let new_array = asm_ccall!(asm, rb_ec_ary_new_from_values, EC, num.into(), argv);
gen_pop_opnds(asm, &elements);
new_array
}

new_array
}

/// Compile array access (`array[index]`)
Expand Down
Loading