Performance degradation of Buffer allocation in NodeJS v24

Version

v24.13.0

Platform

Linux ubuntu24 6.14.0-37-generic #37~24.04.1-Ubuntu SMP PREEMPT_DYNAMIC Thu Nov 20 10:25:38 UTC 2 x86_64 x86_64 x86_64 GNU/Linux

Subsystem

No response

What steps will reproduce the bug?

console.log(process.version);
let table = [], count = 1e6;
for (let chunk_size of [10, 20, 50, 100, 200].map(v=>v*1024))
{
    let ts = Date.now();
    for (let i = 0; i<count; i++)
        Buffer.allocUnsafe(chunk_size);
    let te = Date.now();
    table.push({chunk_size, ms: te-ts});
}
console.table(table, ['chunk_size', 'ms']);

How often does it reproduce? Is there a required condition?

Every time

What is the expected behavior? Why is that the expected behavior?

Buffer allocation in NodeJS v24 should be same or better than in v20.

What do you see instead?

Degradation depended on chunk size:

v20.12.1
┌─────────┬────────────┬──────┐
│ (index) │ chunk_size │ ms   │
├─────────┼────────────┼──────┤
│ 0       │ 10240      │ 1044 │
│ 1       │ 20480      │ 1110 │
│ 2       │ 51200      │ 1230 │
│ 3       │ 102400     │ 1339 │
│ 4       │ 204800     │ 1969 │
└─────────┴────────────┴──────┘


v24.13.0
┌─────────┬────────────┬──────┐
│ (index) │ chunk_size │ ms   │
├─────────┼────────────┼──────┤
│ 0       │ 10240      │ 996  │
│ 1       │ 20480      │ 1223 │
│ 2       │ 51200      │ 1387 │
│ 3       │ 102400     │ 2077 │
│ 4       │ 204800     │ 3349 │
└─────────┴────────────┴──────┘

Additional information

No response

Start by running the supplied Buffer.allocUnsafe benchmark on Node.js v20.12.1 and v24.13.0 under the stated Linux environment, then inspect the Buffer allocation path for changes affecting larger chunks. Done means the regression is explained and a reproducible benchmark or test demonstrates the relevant behavior and any correction.